VEHICLE DATA ANALYSIS METHOD AND VEHICLE DATA ANALYSIS SYSTEM

专利摘要:
vehicle data analysis method and vehicle data analysis system. The present invention relates to a vehicle data analysis method that allows the quantitative analysis of vehicle data characteristics that indicate the transition in a driver driving operation, and a vehicle data analysis system that uses this method. of analysis. a plurality of vehicle data is collected based on a plurality of types of driving operations. based on an evaluation criterion such as an index to assess the levels of driving operations, the collected vehicle data is grouped into at least two groups. characteristic values of vehicle data that differ between these groups are extracted.
公开号:BR112013029418B1
申请号:R112013029418-3
申请日:2012-05-15
公开日:2021-06-15
发明作者:Hironobu Sugimoto；Shojiro Takeuchi；Satomi Yoshioka；Yoshihiro Suda；Yoichi Sato；Takayuki Hirasawa；Daisuke Yamaguchi；Shuguang Li
申请人:The University Of Tokyo；Toyota Jidosha Kabushiki Kaisha；
IPC主号:

专利说明:

Field of Technique
[001] The present invention relates to a vehicle data analysis method advantageously applicable for analysis of vehicle data obtained from a vehicle, and to a vehicle data analysis system using such data analysis method of vehicle. Background Technique
[002] A driving aid system to assist a driver in driving a vehicle is typically designed to acquire traffic information that relates to intersections, momentary stopping positions, turns, an approaching vehicle from the front, and other information that require the driver to decelerate the vehicle by means of a vehicle-mounted camera, a navigation system or the like. The driving aid system then provides a driving aid to the driver based on the traffic information around the vehicle thus acquired, for example by providing the driver with a voice guidance message to decelerate.
[003] This type of driving aid is generally accomplished using a standard route model obtained by averaging data from various driving behaviors including traffic information perception, judgment, and common driver driving operations, which are measured under a predetermined path pattern such as a false driving course. PTL 1, for example, describes a system that first generates exemplary operating data that indicates a time series transition of an exemplary operating amount of an operating equipment such as an accelerator pedal or a brake pedal based on information that refers to an approach speed to an intersection or a turn, and to the shape of a road such as the radius of curvature of an intersection or a turn (path model). The system then records the generated exemplary operating data in a database as an exemplary driving model (traditional route pattern). The system presents, to a driver of a vehicle to be assisted, the exemplary operating data recorded in the database simultaneously with the data in transition on an operating quantity of various operating equipment by the driver, so that the driving behavior of the driver is evaluated.
[004] A path pattern of a vehicle approaching an intersection or a momentary stop position generally varies according to a road environment that has a variety of elements such as the curvature of a road curve, the width and the slope of a lane, as well as according to the driving technique or driving habits of the driver. It is difficult to adapt standardized travel patterns to such a driver's variable travel pattern. This means that it is unrealistic to generate an exemplary driving model based on the actual road environment, driving technique and driver driving habits, as it takes an immense amount of man-hours to generate such a model.
[005] On the other hand, another method has been studied recently, in which a group of vehicle data collected during the operation of driving by a plurality of drivers is stratified according to their driving techniques, that is, the levels of driving performance of them, analyzing vehicle data that indicate the transition of the driving operation by the drivers. However, even with such a method of stratifying the vehicle data group according to the driving performance levels of drivers, it is still difficult to identify which type of driving operation causes a difference in driving performance level among drivers. Therefore, this method was not successful in identifying the factors for which different drivers have different levels of driving performance. In other words, it was not successful in identifying the driving elements to be assisted in order to improve the driving performance level. This means that the characteristics of vehicle data collected based on drivers' driving operations cannot be quantitatively understood.Citation ListPatent Literature
[006] PTL 1: Publication of Patent Open to Public Inspection in JP 2009-294250 Invention Summary
[007] The present invention was created in view of the circumstances as described above. It is an object of the invention to provide a vehicle data analysis method advantageously applicable for the analysis of vehicle data obtained from vehicles, and a vehicle data analysis system using that vehicle data analysis method.
[008] Means to achieve the objective and its advantages will be described below.
[009] In accordance with an aspect of the present disclosure, a vehicle data analysis method for analyzing vehicle data that reflects a driver driving operation is provided. The method includes:
[0010] collect a plurality of vehicle data fragments based on a plurality of types of driving operations; group these collected fragments of vehicle data into at least two groups based on an evaluation criterion which is an index for evaluating a level of driving operation; and
[0011] extract characteristic values from vehicle data that differ between groups.
[0012] In accordance with another aspect of the present disclosure, a vehicle data analysis system for analyzing vehicle data that reflects a driver driving operation is provided. The system includes a storage device for storing vehicle data based on a plurality of types of driving operations, a vehicle data classification unit for grouping the vehicle data stored in the storage device into at least two groups based on in an evaluation criterion which is an index to evaluate a driving operation level, and a vehicle data analysis unit (250) to extract characteristic values of vehicle data that differ between groups grouped by the data classification unit of vehicle.
[0013] Vehicle data indicating a driver-driving operation indicates an operating mode for various driving elements such as an accelerator, a brake pedal and a steering wheel. The mode of operation for such driving elements significantly affects fuel efficiency and vehicle behavior during vehicle operation. For example, a vehicle data group that is assessed as being of a high driving performance level as it has a low fuel consumption per unit of time, i.e. a high fuel efficiency (fuel conservation), often includes a common characteristic value such as throttle shutdown at a predetermined time. On the other hand, a vehicle data group that is assessed as being of a low driving performance level as it has a high fuel consumption often includes a characteristic value such as a relatively delayed throttle shutdown or a excessive depression of the accelerator pedal. When the characteristic values contained in vehicle data groups that are mutually different in driving performance level are different from each other, the driving operation indicated by these different characteristic values often constitutes a factor causing a difference in the performance level of driving the vehicle data. This means that these different characteristic values, in other words the performance of a particular driving operation indicated by these characteristic values, causes a difference in fuel efficiency as a result of the driving operation or vehicle behavior during vehicle travel.
[0014] According to the above-mentioned configuration or methods, vehicle data indicating the transition in the driving elements such as the amount of throttle operation or the steering angle of the steering wheel are collected from the vehicles. The collected vehicle data is grouped according to an evaluation criterion capable of identifying a driver's driving technique, whereby the group of vehicle data acquired under a plurality of types of driving operations is categorized according to driving performance levels. Characteristic values that differ between these groups categorized according to driving performance levels are extracted, whereby the characteristic value that constitutes a factor causing a difference in driving performance level between the categorized vehicle data is extracted. This makes it possible to extract information that quantitatively indicate the factor causing the difference in the level of driving performance between the vehicle data from the acquired vehicle data group based on a plurality of types of driving operations. In other words, the driving operation characteristic contained in the vehicle data can be quantitatively analyzed.
[0015] The collected vehicle data is acquired from vehicles that made a journey on an actual road under a plurality of types of driving operations by a driver. Therefore, the use of vehicle data further makes it possible to generate a route model that reflects an actual route environment or driving operations performed under the route environment. In that case, a route model that incorporates characteristic values that determine a level of driving performance can be generated by generating an exemplary route model from the characteristic values extracted from the vehicle data.
[0016] The vehicle data analysis method of the present disclosure preferably obtains a degree of influence exerted by the characteristic value extracted from vehicle data on the evaluated vehicle data based on the evaluation criteria.
[0017] In the vehicle data analysis system of the present disclosure, the vehicle data analysis unit preferably further includes an influence calculation unit to obtain a degree of influence exerted by the characteristic value extracted on the evaluated vehicle data based on the evaluation criteria.
[0018] The characteristic values contained in the vehicle data include characteristic values that exert a high influence on an evaluation result of the vehicle data based on an evaluation criterion and characteristic values that exert a low influence on the evaluation result. Characteristic values that exert a high influence on the evaluation result are a major factor causing a difference in the level of driving performance in the vehicle data group.
[0019] Therefore, the method or configuration described above determines an influence exerted by the characteristic value extracted from the vehicle data on the vehicle data evaluated under the evaluation criterion. This makes it possible to specify not only the characteristic values that differ between the grouped vehicle data, but also a degree of influence exerted by the characteristic value on the vehicle data evaluated under the evaluation criterion, in other words, on the evaluation using the criterion of evaluation.
[0020] Preferably, in the vehicle data analysis method of the present disclosure, vehicle data that includes information indicating one of a traffic element, a route section and a route area in which the traffic element and sections are connected in series, and the grouping of the vehicle data and the extraction of characteristic values from the vehicle data is performed by treating the traffic element or the route section or the route area as a unit.
[0021] Preferably, in the vehicle data analysis system of the present disclosure, the vehicle data includes information indicating a traffic element, a route section and a route area in which the traffic element and route section are connected in series. The vehicle data classification unit and the vehicle data analysis unit perform the grouping of the vehicle data and the extraction of characteristic values from the vehicle data regarding the traffic element or the route section or the area of route as a unit.
[0022] A driver's driving operation significantly reflects traffic elements such as intersections with a traffic light or curves, predetermined route sections defined by traffic elements such as junctions or curves, and a route environment surrounding the vehicle such as a route area that includes traffic elements and route sections. The characteristics of driving operations performed under such a travel environment typically vary depending on the travel environment under which the vehicle travels. A driver's driving technique also varies depending on different riding environments. For example, a driver may exhibit high driving technique around a curve as the vehicle's behavior in the curve is small, but may exhibit poor driving technique in a decelerating or stopping position such as a crossing as the consumption of fuel is high in the deceleration or stop position. Consequently, even if data for a single vehicle is acquired based on the driving operations of the same driver, the results of evaluating them under an evaluation criterion often differ depending on the vehicle's journey environment.
[0023] When vehicle data are categorized into groups and characteristic values of the vehicle data are extracted by treating the traffic elements or the route sections or the route areas that make up a route environment as units according to the method or setting described above, vehicle data characteristic values that reflect a series of driving operations that relate to traffic elements, route sections and route areas can be accurately extracted from the vehicle data groups . This makes it possible to identify a factor causing a difference in the level of driving performance by treating traffic elements, route sections and route areas as units. This also makes it possible to extract more characteristic values from the collected vehicle data.
[0024] Preferably, in the vehicle data analysis method of the present disclosure, the vehicle data includes information indicating a waypoint, and the vehicle data analysis method further comprises obtaining a correspondence relationship between the extracted characteristic values of the vehicle data and the waypoint, and a correspondence relationship between the characteristic values extracted from the vehicle data and an evaluation result of the vehicle data based on the evaluation criteria. The correspondence relationship between the characteristic value and the waypoint indicates how the weighted characteristic value of the vehicle data relates to the waypoint at which the driving operation indicated by the characteristic value is performed. The correspondence relationship between the characteristic value and the evaluation criterion indicates how the weighted characteristic value of the vehicle data relates to the waypoint at which the driving operation indicated by the characteristic value is performed.
[0025] Preferably, in the vehicle data analysis system of the present disclosure, the vehicle data includes information indicating a waypoint. The vehicle data analysis unit additionally obtains a correspondence relationship between the characteristic values extracted from the vehicle data and the waypoint, and a correspondence relationship between the characteristic values extracted from the vehicle data and an evaluation result of the data vehicle based on the evaluation criteria.
[0026] For example, after a vehicle enters a curve, various driving operations are performed at waypoints that form the curve according to a lane shape or the like before exiting the curve. This means that even though the vehicle data indicates driving operations on a common curve or similar curves in radius of curvature or the like, the vehicle data that reflects driving operations at different waypoints contain separate characteristic values that correspond to the points of respective route. Therefore, even if the vehicle data is acquired on the basis of the driver driving operations tables, the evaluation results of the vehicle data at different waypoints may differ among the waypoints. For example, a high fuel efficiency (fuel economy) will be displayed from a curve start point to a curve midpoint, while a low fuel efficiency will be displayed from the curve midpoint to the end point of the curve.
[0027] Therefore, in the method or configuration as described above, a correspondence relation between characteristic values for vehicle data and waypoints and a correspondence relation between characteristic values of vehicle data and an evaluation criterion are obtained, by means of the which a characteristic value observed at each of the waypoints and an influence exerted by the characteristic value on the evaluation criterion can be obtained for each of the waypoints. This makes it possible to analyze the vehicle data in a more detailed way, and analyze factors that cause a difference in the level of driving performance based on the evaluation criteria and an influence exerted by the factor on the evaluation criteria in detail for the level of points of route.
[0028] Preferably, the vehicle data analysis method of the present disclosure further includes normalizing the time series data as the vehicle data based on a course position.
[0029] Preferably, in the vehicle data analysis system of the present disclosure, the vehicle data analysis unit further includes a normalization operation unit for normalizing time series data as the vehicle data based on a position of route.
[0030] In general, the travel speed of a vehicle that serves as a source of supply for vehicle data varies depending on different vehicles and different drivers. Therefore, a comparison is performed between time series data, more specifically vehicle data in which a driver's driving operating modes acquired in common or similar traveling environments are recorded in time series. However, if vehicle travel speeds as the data source differ significantly, time series data from driving operations at a given travel position is possibly compared with time series data from driving operations at a different travel position.
[0031] However, according to the method or setting described above, the time series data is normalized based on a course position, so that the time series data acquired from vehicles with different course speeds can be transformed at a level that allows for an accurate comparison. This makes it possible to analyze more vehicle data (time series data), and extract more characteristic values that differ between such vehicle data.
[0032] Preferably, in the vehicle data analysis method of the present disclosure, characteristic values indicate characteristics of one or more driving elements that represent a driver driving mode of operation, and the method of analysis further comprises obtaining a degree of influence exerted by the driving element on the evaluation criteria for each of the driving elements.
[0033] Preferably, in the vehicle data analysis system of the present disclosure, characteristic values are characteristics of one or more driving elements that indicate a driver driving mode of operation, and the vehicle data analysis unit determines a degree of influence exerted by the driving element on the assessment criteria for each of the driving elements.
[0034] In general, the fuel efficiency or vehicle behavior, which indicates the results of driving a driver, varies depending on a plurality of types of operating modes of driving elements such as the steering angle and the amount of throttle operation. Driving performance levels differ between vehicle data due to a difference in the operating modes of the driving elements.
[0035] Therefore, according to the method or configuration described above, an influence exerted by a driving element indicated by the characteristic value on the evaluation criterion is determined for each of the driving elements, so that a plurality of factors causing a difference in the driving performance level of vehicle data can be accurately identified even when a difference occurs in the driving performance level due to the influences of a plurality of types of driving elements. This makes it possible to accurately extract characteristic values contained locally in the vehicle data, and to specify a degree of influence exerted by each of the driving elements on the evaluation criteria.
[0036] Preferably, the vehicle data analysis method of the present disclosure further includes generating a plurality of candidate data as original data to indicate the characteristic values of the vehicle data by means of the frequency solution of the vehicle data.
[0037] Preferably, the vehicle data analysis unit additionally has a frequency solution unit for generating a plurality of candidate data as original data to indicate characteristic values of vehicle data by means of the frequency solution of the vehicle data.
[0038] The characteristic values of vehicle data are often contained as various frequency components in the vehicle data, and the driving performance levels of the vehicle data are differentiated from each other through the influence of the characteristic values contained in the frequency components.
[0039] The method or configuration described above is capable of revealing several characteristic values that reflect driving operations by extracting various frequency components contained in the vehicle data by means of the frequency solution of the vehicle data. Therefore, when characteristic values that are different between the grouped vehicle data need to be extracted, large amount candidate data containing candidate characteristic values can be generated from a limited amount of vehicle data.
[0040] Preferably, the vehicle data analysis method of the present disclosure further includes revealing the characteristic values by applying a window function to the vehicle data prior to extracting the characteristic values from the vehicle data.
[0041] Preferably, in the vehicle data analysis system disclosed herein, the vehicle data analysis unit additionally includes a window function operating unit to reveal a characteristic value by applying a window function to the data of vehicle.
[0042] According to the method or configuration described above, a characteristic value of vehicle data is revealed by applying a window function to the vehicle data, whereby a characteristic minute value can also be extracted so necessary.
[0043] Preferably, in the vehicle data analysis method of the present disclosure, the evaluation criterion is a criterion for grouping that is performed for at least one of the evaluation items that consist of the fuel efficiency defined by the distance of travel of a vehicle by amount of fuel unit, travel time, vehicle behavior and side bump of a vehicle.
[0044] Preferably, in the vehicle data analysis system disclosed herein, the vehicle data classification unit categorizes the vehicle data into groups based on assessment criteria that relate to at least one of the assessment items that consist of fuel efficiency indicated by vehicle travel distance by fuel unit quantity, travel time, vehicle behavior, and vehicle side bump.
[0045] In general, a driving operation capable of performing a low fuel consumption and a smooth driving operation with small vehicle behavior can be evaluated that the driver's driving technique is high, and the vehicle data can be categorized into groups according to driving performance levels for assessment items that include fuel efficiency and vehicle behavior, as well as travel time and vehicle lateral bump. Therefore, according to the method or configuration described above, a factor that causes a difference in fuel efficiency, travel time, smooth driving operation (presence or absence of fast braking) and side jolt between collected vehicle data can be identified by extracting characteristic values from categorized vehicle data using rating item indices.
[0046] Preferably, in the vehicle data analysis method of the present disclosure, grouping includes categorizing the plurality of vehicle data types into a "high" driving performance level vehicle data group and a "high" driving performance level data group. vehicle of "low" driving performance level by grouping based on the evaluation criteria. Extracting the vehicle data includes performing weighting on characteristic values of vehicle data that differ in driving performance level based on a total value of characteristic value differences that are common among the driving performance level data group "high" and the "low" driving performance level vehicle data group, which are grouped based on the evaluation criteria. The difference of common characteristic values is a difference between the characteristic value of the data categorized by the "high" driving performance level group and the characteristic value of the data categorized by the "low" driving performance level group.
[0047] Preferably, in the vehicle data analysis system disclosed herein, the vehicle data classification unit categorizes the plurality of vehicle data types into a "high" driving performance level vehicle data group " and a group of vehicle data of "low" driving performance level by grouping based on the evaluation criteria, and the vehicle data analysis unit extracts the vehicle data by weighting characteristic values of the data of vehicles that have different driving performance levels based on a total value of common differences from the data categorized in the "high" driving performance level vehicle data group and the driving performance level vehicle data group "low" based on the evaluation criteria. The common difference of characteristic values is a difference between a characteristic value of data categorized as "high" driving performance level and a characteristic value of data categorized as "low" driving performance level under the common evaluation criteria.
[0048] In the vehicle data group described above, a characteristic value that exhibits a greater difference between vehicle data that indicates an amount of steering angle operation at a given time from the data categorized as "high" driving performance level and vehicle data corresponding to the amount of steering angle operation from data categorized as "low" driving performance level may be a major factor causing a difference in driving performance level. A characteristic value that exhibits a high rate of difference occurrence between the "high" driving performance level vehicle data and the "low" driving performance level vehicle data may be a major factor causing a difference in the driving performance level. On the other hand, when a characteristic value of vehicle data indicating an amount of accelerator pedal depression at a given time is similar between data categorized as "high" driving performance level and data categorized as "high" driving performance level. "low" driving, the characteristic value cannot be a factor causing a difference in the driving performance level of the vehicle data group.
[0049] Therefore, according to the method or setting described above, characteristic values of vehicle data that differ in the level of driving performance are weighted according to a total difference value of the characteristic values, so that the factors causing differences in the level of driving performance can be identified in order of degree of influence to the evaluation criterion. This makes it possible to identify the elements to be improved in the driving operation indicated by the vehicle data evaluated as "low" in the driving performance level, in order of priority.
[0050] The features of the present invention that are considered innovative are presented with particularity in the appended claims. The invention, together with the objectives and advantages thereof, can be better understood by reference to the following description of the current preferred embodiments together with the attached drawings in which: Brief Description of Drawings
[0051] Figure 1 is a block diagram of a vehicle data analysis system to which a vehicle data analysis method of the invention is applied, according to an embodiment of a vehicle data analysis method and a invention system;
[0052] Figure 2(a) is a graph showing an example of a group of vehicle data as time series data;
[0053] Figure 2(b) is a graph showing an example of the group of vehicle data normalized based on course positions;
[0054] Figure 3 is a flowchart showing an example of vehicle data analysis procedures according to an embodiment of a vehicle data analysis method;
[0055] Figure 4 is a diagram schematically showing an example of a route area in which a vehicle used as a vehicle data source has made a route, along with the traffic elements and route sections contained in the area of course;
[0056] Figure 5(a) is a graph showing an example of time series data indicating the steering angle transition of the vehicle that made a route in the route area shown in Figure 4, where St1 represents data from time series that are evaluated as "high" driving performance level, and St2 represents time series data that are evaluated as "low" driving performance level;
[0057] Figure 5(b) is a graph showing normalized time series data obtained by normalizing the time series data shown in Figure 5(a) based on travel positions;
[0058] Figure 6(a) is a graph that shows an example of vehicle data resolved mainly in low frequency components;
[0059] Figure 6(b) is a graph showing an example of candidate data obtained through frequency solution of the vehicle data shown in Figure 6(a);
[0060] Figure 6(c) is a graph showing an example of candidate data obtained by solving the frequency of the vehicle data shown in Figure 6(a);
[0061] Figure 6(d) is a graph showing an example of candidate data obtained by solving the frequency of vehicle data shown in Figure 6(a);
[0062] Figure 6(e) is a graph showing an example of candidate data obtained by solving the frequency of vehicle data shown in Figure 6(a);
[0063] Figure 7 illustrates an example in which a window function is applied to frequency solution vehicle data;
[0064] Figure 8(a) is a diagram showing an example of an analysis result of a group of vehicle data;
[0065] Figure 8(b) is a diagram showing a route model in the route area of Figure 4, together with the result of analysis of the vehicle data group;
[0066] Figure 9 is a flowchart showing an example of frequency solution processing performed by a frequency solution unit; and
[0067] Figure 10(a) is a diagram showing an example of an analysis result of a vehicle data group according to another embodiment of a vehicle data analysis method and a vehicle data analysis system of the invention.
[0068] Figure 10(b) is a diagram showing an example of an analysis result of a vehicle data group according to another embodiment of a vehicle data analysis method and a vehicle data analysis system of the invention. Description of Modalities
[0069] A vehicle data analysis method and a vehicle data analysis system according to an embodiment of the present invention will be described with reference to Figures 1 to 9.
[0070] As shown in Figure 1, a vehicle data analysis system to which a vehicle data analysis method of the present embodiment is applied has a vehicle data analysis center 200 for collecting vehicle data reflecting vehicle operations. driving a driver of a vehicle 100.
[0071] The vehicle 100 that serves as a source of supply (acquisition source) of vehicle data is configured to allow detection of modes of operation of various driving elements by a driver, and has an accelerator sensor 101, a brake sensor 102, a steering angle sensor 103, a gyro sensor 104, a vehicle speed sensor 105 and an acceleration sensor 106. These sensors 101 to 106 are connected to a vehicle data storage area 110 via an internal vehicle network such as a CAN (Control Area Network) type network. Vehicle data storage area 110 stores detection results from sensors 101 to 106 in time series as vehicle data that reflect driving operations of a driver of vehicle 100.
[0072] The throttle sensor 101 detects a throttle operating amount, which varies according to the throttle pedal driver operation, and issues a signal corresponding to the detected throttle operating amount to the vehicle data storage area 110. Brake sensor 102 detects a depression amount of a driver operated brake pedal, and outputs a signal corresponding to the detected amount of depression to vehicle data storage area 110. Steering angle sensor 103 detects a steering angle, which varies according to the driver's operation of the steering wheel, and outputs a signal corresponding to the detected steering angle to the vehicle data storage area 110. The gyro sensor 104 detects the direction of vehicle path 100, and outputs a signal corresponding to the detected travel direction to vehicle data storage area 110. Vehicle speed sensor 105 detects a speed. rotational speed of vehicle wheels 100, and outputs a signal corresponding to the detected rotational speed to vehicle data storage area 110. Acceleration sensor 106 detects an acceleration of vehicle 100, and outputs a signal corresponding to the detected acceleration to the vehicle data storage area 110.
[0073] Vehicle 100 additionally has a motor ECU mechanism 107 to control a motor mechanism mounted on vehicle 100, a GPS 108 to detect an absolute position of the vehicle 100, and a car navigation system 109 which has lane map data on the vehicle. same.
[0074] The engine engine ECU 107 generates a fuel injection signal to determine a fuel injection amount based on a detection result of the throttle sensor 101, and outputs the generated fuel injection signal to a fuel injection device. fuel (not shown). Engine engine ECU 107 sends the fuel injection signal to vehicle data storage area 110.
[0075] The GPS 108 receives a GPS satellite signal to detect the absolute position of the vehicle, and finds a latitude and longitude of the vehicle 100 based on the received GPS satellite signal. The GPS 108 outputs information indicating the latitude and longitude of the vehicle 100 to the vehicle data storage area 110.
[0076] The car navigation system 109 provides the driver with guidance on a recommended route from a starting point to a destination point, and has road map data of the driving area for vehicle 100. This road map data is information about a map, including information about road gradients, road alignments such as curves, traffic elements such as railroad intersections and intersections, intersection names, road names, area names, installation of a driving guide and the like . Recorded in this road map data are information relating to latitudes and longitudes and information relating to traffic elements such as continuing lanes, intersections and traffic lights, route sections defined by traffic elements, and route areas that include successive traffic elements and route sections. In other words, information that relates to a vehicle's route environment is recorded in the road map data. Car navigation system 109 refers to road map data for outputting information relating to a travel environment that vehicle 100 has passed, to vehicle data storage area 110.
[0077] The vehicle data storage area 110 records, in time series, the detection results supplied from the sensors 101 to 106 and the GPS 108, the fuel injection signals received from the ECU engine mechanism 107 , and the route environment information received from the car navigation system 109. The vehicle data storage area 110 thus accumulates, as vehicle data from vehicle 100, information indicating driving operations by a driver of vehicle 100 , and information relating to the travel environment in which vehicle 100 has made a journey under driver driving operations. In vehicle data storage area 110, information indicating the driver driving operations of vehicle 100 is associated with information relating to the travel environment in which vehicle 100 has made a tour under the driver driving operations.
[0078] The vehicle 100 additionally has a vehicle internal communication device 120 capable of wireless communication with a vehicle data analysis center 200. The vehicle internal communication device 120 transmits vehicle data accumulated in the data storage area from vehicle 110 to vehicle data analysis center 200, for example, when vehicle 100 travel is complete and an accessory position of vehicle 100 is switched from an ON state to an OFF state.
The vehicle data analysis center 200 has a central communication device 210 for receiving vehicle data transmitted from a plurality of vehicles including the vehicle 100. The central communication device 210 receives vehicle data from a plurality of vehicles including vehicle 100 and outputs the vehicle data to a central storage device 220.
[0080] The central storage device 220 accumulated vehicle based on a plurality of types of driving operations obtained from a plurality of vehicles through wireless communication between the central communication device 210 and the vehicle internal communication devices mounted on the plurality of vehicles.
The vehicle data analysis center 200 has a vehicle data classification unit 230. The vehicle data classification unit 230 categorizes groups of vehicle data accumulated in the central storage device 220 into two groups, namely one "high" driving performance level group and a "low" driving performance level group, based on an evaluation criterion such as an index for evaluating driving operation levels.
[0082] The vehicle data classification unit 230 according to the present embodiment categorizes the vehicle data groups accumulated in the central storage device 220 while considering, as a unit, at least one of such traffic elements such as curves and junctions, predetermined route sections defined by traffic elements, and route areas that include successive traffic elements and route sections. This means that the vehicle data classification unit 230 categorizes the vehicle data groups stored in the central storage device 220 according to driving performance levels, for example by dividing the vehicle data groups into given units each of which indicates a driving operation in a common route area. The vehicle data classification unit 230 categorizes the vehicle data groups stored in the central storage device 220 according to driving performance levels by subdividing the vehicle data groups into data units. Each of the data units indicates, for example, a driving operation on a particular curve that exists in the common travel area, or a driving operation on a travel section defined by successive intersections. A route environment which includes such traffic elements, route section and route area is identified based on information provided by the vehicle data. Information is, for example, latitude and longitude information obtained as a detection result from the GPS 108 or information acquired from the car navigation system 109.
[0083] In the present modality, the evaluation criterion to be used as a basis for grouping vehicle data is established for at least one of the fuel efficiency represented by travel distance per vehicle fuel unit quantity, travel time, vehicle behavior and vehicle side jerk.
[0084] For example, when fuel efficiency is selected as an assessment item, the assessment criteria for categorizing vehicle data groups by level is set to "15 km/l", for example. The fuel efficiency of vehicle 100 is calculated based on a fuel injection signal contained in vehicle data of vehicle 100. Whether the fuel efficiency obtained as a result of detection by the engine engine ECU 107 is equal to or greater than the criterion rating of "15km/l", the vehicle data is rated as "high" in the driving performance level. In contrast, if the fuel efficiency obtained as a detection result by the engine mechanism ECU 107 is less than the evaluation criterion of "15 km/l", the vehicle data is evaluated as "low" in the driving performance level.
[0085] When a vehicle makes a route in a certain traffic element or a predetermined section of route within a legal speed limit, a period of time required for the vehicle to make a route from the entrance of the traffic element or route section pre-determined way out of it is called a "passage period". When this transit period is within a predetermined travel time, the vehicle data is evaluated as "high" in driving performance level. In contrast, when the pass period exceeds the predetermined travel time, the vehicle data is evaluated as "low" in driving performance level.
[0086] A jolt, which is a variation in acceleration in a lateral direction relative to the direction of travel of the vehicle 100, can be obtained from a detection result by the acceleration sensor 106. A predetermined reference value is also specified for the bump for driving performance level assessment. Each of the vehicle data is evaluated as "high" or "low" in driving performance level based on bump and reference value.
[0087] In addition, a reference value is specified for the assessment of the driving performance level based on vehicle behaviors 100. Vehicle behaviors include the occurrence of rapid acceleration or deceleration or rapid braking and the frequency of their occurrence , which can be obtained from the detection results by the throttle sensor 101 and the brake sensor 102. For example, when a fast braking occurrence frequency is less than a predetermined threshold, the vehicle data is evaluated as " high" in the driving performance level, while when the frequency of occurrence of rapid braking is equal to or greater than the predetermined threshold, the vehicle data is evaluated as "low" in the driving performance level.
[0088] According to the present embodiment, the groups of vehicle data accumulated in the central storage device 220 are categorized into two groups in total, one with a driving performance level of "high" and the other with a performance level of "low" driving, for the "fuel efficiency" rating item, for example, and information indicating the rating result is associated with each of the vehicle data.
The vehicle data classification unit 230 outputs the vehicle data groups categorized on the basis of the evaluation criteria to a classification result storage unit 240. The classification result storage unit 240 stores the classification result through vehicle data classification unit 230.
[0090] The classification result storing unit240 stores the vehicle data received from the vehicle data classification unit 230 separately according to the classification result through the vehicle data classification unit 230. classification result storage 240 in accordance with the present embodiment has a first storage area 241 and a second storage area 242. For example, the first storage area 241 stores vehicle data that has been assessed as "high" in the level of driving performance by vehicle data classification unit 230, while the second storage area 242 stores the vehicle data that has been assessed as "low" in the driving performance level by vehicle data classification unit 230. In this case, the first storage area 241 accumulates a group of vehicle data DA, which are evaluated as "high" in the level. of driving performance by the vehicle data classification unit 230, and the second storage area 242 accumulates a group of vehicle data DB, which are evaluated as "low" in the level of driving performance by the data classification unit of vehicle 230.
The vehicle data analysis center 200 has a vehicle data analysis unit 250. The vehicle data analysis unit 250 extracts characteristic values from vehicle data that differ between the vehicle data group DA and the DB vehicle data group separately stored in classification result storage unit 240.
[0092] The vehicle data analysis unit 250 according to the present embodiment has a normalization operation unit 251. The normalization operation unit 251 normalizes the vehicle data, which is time series data based on the course position. The first storage area 241 stores a group of DA vehicle data collected from the vehicle which has made a route in several route areas. During normalization of the vehicle data, the normalization operation unit 251 retrieves, for example, vehicle data relating to a curve having a predetermined radius of curvature from the group of vehicle data DA stored in the first area of storage 241. The normalizing operation unit 251 retrieves, from the vehicle data group DB stored in the second storage area 242, a group of vehicle data acquired under a travel environment that is common with or similar to the travel environment. route indicated by the recovered DA vehicle data group. Thus, the normalizing operation unit 251 retrieves a first group of vehicle data (DA) and a second group of vehicle data (DB) from the classification result storage unit 240. The first group of vehicle data indicates a driving operation that performs a low fuel consumption (fuel economy) is achieved on a given curve that exists on a given route of travel, and the second group of vehicle data indicates a driving operation in which a curve with a shape similar to the aforementioned curve is a factor causing high fuel consumption.
[0093] Based on information indicating traffic elements, information indicating route sections and information indicating route areas contained in various vehicle data, the normalization operation unit 251 according to the present modality thus retrieves the from the classification result storing unit 240, vehicle data indicating driving operations that are common or similar in traffic elements, route sections or route areas.
[0094] Time series data as shown in Figure 2(a) by way of example is thus retrieved from classification result storage unit 240. The time series data indicates the transition of modes of operation of a plurality of types of operating equipment that include an accelerator pedal, a brake pedal, a steering wheel and other driving elements of the vehicle 100. The transition of operating modes of the operating equipment is obtained as detection results by sensors 101 to 106 in a particular traffic element, a particular route section or a particular route area.
[0095] In Figure 2(a), the group of vehicle data DA assessed as "high" in the driving performance level is represented by solid lines, while the group of vehicle data DB assessed as "low" in the performance level conduction is represented by lines formed by a long dash alternating with a short dash. The driving operations performed by the vehicles serving as sources of supply of the vehicle data groups DA and DB are different from each other, and consequently the vehicle data groups DA and DB have characteristic values different from each other. The characteristic values contained in the vehicle data indicate characteristics of operating modes of driving elements that are operating an equipment mounted on vehicle 100. The operating modes of driving elements can be exemplified by turning off the throttle at a predetermined time, time when the brake is turned on, change the amount of brake pedal depression and change the steering angle. In vehicle 100 that serves as the source of vehicle data supply, variations in fuel efficiency, travel time in a predetermined travel section, rapid braking and side jolt of vehicle 100 occur, according to the characteristics of the operating modes of the driving elements.
[0096] Vehicle data according to the present embodiment will be detection results by sensors 101 to 106, which are recorded in time series. As seen from Figure 2(a), even though vehicle data is collected from vehicles passing a route area starting point, route section or common or similar traffic element, vehicle data has a different data length due to the difference in travel speed or the like of the vehicles that serve as the sources of supply. Due to the difference in travel speed or the like of the vehicles that serve as the vehicle data supply sources, the vehicles display different travel periods, each of which is defined as a period of time required by a vehicle to pass through. a particular traffic element, route section or route area after entering it.
[0097] Therefore, the normalization operation unit 251 in accordance with the present embodiment retrieves the vehicle data groups DA and DB acquired under a common or similar path environment of the first storage area 241 and the second storage area 242, respectively. The normalizing operation unit 251 then normalizes the retrieved DA and DB vehicle data groups based on the travel position. As shown in Figure 2(b) by way of example, the data lengths of the vehicle data are thereby matched. As shown in Figure 1, the normalizing operation unit 251 outputs the normalized vehicle data to a frequency solution unit 252 for a frequency solution of the vehicle data.
[0098] Once the normalized vehicle data group is inputted into the frequency solution unit 252 by the normalizing operation unit 251, the frequency solution unit 252 frequency solves the vehicle data group in a plurality of bands for example, by means of a wavelet transform. For example, frequency solution unit 252 generates a plurality of data by solving single vehicle data normalized by normalizing operation unit 251 according to predetermined frequency components. The frequency solution unit 252 performs this frequency solution for each of the vehicle data groups normalized by the normalization operation unit 251. Thus, a massive amount of candidate data is generated as candidates for extraction to be used when a value characteristic is extracted as a factor causing a difference in vehicle data driving performance levels. This frequency solution further reveals a frequency component (characteristic value) contained locally in the vehicle data. Thus, the frequency solution unit 252 outputs a group of candidate data generated by means of the frequency solution to a window function operating unit 253 to apply a window function to a group of candidate data.
[0099] The window function operating unit 253 reveals the characteristic value contained in each of the candidate data groups received from the frequency solution unit 252 by applying a window function to each of these. The window function operating unit 253 outputs the candidate data group in which the characteristic value has been revealed by applying the window function to an influence calculation unit 254. The influence calculation unit 254 obtains a degree of influence exercised by the characteristic value that was revealed in the vehicle data evaluated using the evaluation criterion.
[00100] Since the window function operating unit 253 inputs the candidate data group in which the characteristic value is revealed by the window function operating unit 253 into the influence calculation unit 254, the influence 254 weights the characteristic value of vehicle data with different driving performance levels between vehicle data group DA and vehicle data group DB, based on a total difference value of a common characteristic value between the data data classified as "high" driving performance level data and data classified as "low" driving performance level data. The common characteristic value difference is a difference between a characteristic value of data categorized into a "high" driving performance level and a characteristic value of data categorized into a "low" driving performance level under a common evaluation criterion. The influence calculation unit 254 according to the present embodiment weights the characteristic value of the vehicle data by learning with "AdaBoost" which is a known learning algorithm. By means of this weighting, characteristic values that differ between the DA vehicle data group categorized as "high" driving performance level and the DB vehicle data group categorized as "low" driving performance level are extracted of the characteristic values contained in vehicle data group DA and vehicle data group DB.
[00101] The influence calculation unit 254 according to the present embodiment determines how the weighted characteristic value of the vehicle data relates to the waypoint at which the driving operation indicated by the characteristic value is performed. It is assumed that a correspondence relationship between the characteristic value and the waypoint is defined by how the weighted characteristic value of the vehicle data is related to the waypoint at which the driving operation indicated by the characteristic value is performed. The correspondence relationship between the characteristic value and the waypoint can be determined based on latitude and longitude information contained in the vehicle data or information indicating various traffic elements, route sections and route areas. The influence calculation unit 254 according to the present embodiment further determines how the weighted characteristic value of the vehicle data corresponds to evaluation results of the vehicle data based on the evaluation items. It is assumed that a correspondence relationship between the characteristic value and the valuation result is defined by how the characteristic value of the vehicle data relates to the valuation result of the vehicle data based on the valuation items. The correspondence relationship between the characteristic value and the evaluation result can be determined from an evaluation result based on the evaluation criteria associated with the vehicle data from which the candidate data is generated.
[00102] The influence calculation unit 254 outputs a calculation result obtained by the influence calculation unit 254 to an analysis result storage unit 260 as an analysis result of the vehicle data. The analysis result storage unit 260 thus stores the characteristic value, which differs between the vehicle data group DA and the vehicle data group DB which have different driving performance levels from each other, information that indicate the weight of the characteristic value, information indicating the correspondence relationship between the characteristic value and the waypoint, and information indicating the correspondence relationship between the amount of information and the evaluation result of the vehicle data.
[00103] The operation of vehicle data performed by the vehicle data analysis method and the system according to the present embodiment will be described with reference to Figures 3 to 9.
[00104] As shown in Figure 3, first turning on at step S101, vehicle data is collected from a plurality of vehicles, vehicle data indicating driving operations performed by the vehicles, and route environments corresponding to the respective driving operations. Thus, as shown in Figure 4, vehicle data is collected from a plurality of vehicles that have made a route through a route area Ar1 in which a route section Sec1, a curve Cv1, a route section Sec2, a Cv2 curve and a Sec3 run section are connected in series in that order. Vehicle data indicates the operating modes of various driving elements in time series. The operating modes of the driving elements include steering angles and accelerator pedal or brake pedal operations performed by the drivers of vehicles that have passed through the Ar1 travel area.
[00105] In step S102 of Figure 3, the collected vehicle data is categorized into a vehicle data group DA (driving performance level "high") and a vehicle data group DB (driving performance level " low"). For example, vehicle data group DA indicates a driving operation that contributes to a high fuel efficiency for the "fuel efficiency" rating item, while vehicle data group DB indicates a driving operation that contributes for low fuel efficiency.
[00106] As shown in Figure 5(a), time series data indicating the steering angle transition of vehicles which have made a course through Ar1 course area are categorized into St1 time series data. and St2 time series data. In Figure 5(a), the St1 time series data indicated by the solid line belongs to a "high" driving performance level group, and the St2 time series data indicated by the broken line belongs to a level group of "low" driving performance.
[00107] Since these St1 and St2 time series data are acquired under different conduction operations, a characteristic value contained in the St1 time series data and a characteristic value contained in the St2 time series data are different from each other . These mutually different characteristic values are a factor that causes a difference in the level of driving performance between the St1 time series data and the St2 time series data.
[00108] This St1 and St2 time series data is acquired from the vehicles, which have traversed the common route area Ar1. However, the vehicles traveled at different speeds. Therefore, a period of time required by the vehicle supplying the St1 time series data to pass through the Ar1 route area is different from a period of time required by the vehicle supplying the St2 time series data to pass through the area distance Ar1, and consequently the length of the St1 data indicated by the geometric time axis is different from the length of the St2 data.
[00109] Therefore, in step S103 of Figure 3, vehicle data groups categorized by driving performance levels are normalized based on a course position. Thus, as shown in Figure 5(b) by way of example, the time series data St1 and St2 that indicate the steering angle transition in the path area Ar1 are transformed into data formats comparable to each other and normalized based on at the waypoints from the starting point Ps to the ending point Pg of the wayway area Ar1. Normalization is thus performed on all vehicle data categorized in vehicle data groups DA and DB in step S102.
[00110] Then, in step S104 of Figure 3, the vehicle data contained in the normalized vehicle data groups DA and DB are resolved by frequency, for example, by means of a wavelet transform, whereby the data that contain mostly high frequency components and data that contain mostly low frequency components as shown in Figure 6(a) by way of example is generated. Data which mainly contain low frequency components are further resolved into data A1 and B1 to B3 which contain mutually different frequency components as shown in Figures 6(b) to 6(e) by way of example. As seen from Figures 6(b) to 6(e), data A1 and B1 to B3 that contain predetermined frequency components have characteristic values indicated by predetermined frequency components. The data that is opposite to the data that contains mostly low frequency components shown in Figure 6(a), that is, the data that contains mostly high frequency components, is also resolved to a plurality of data fragments that have frequency components many different. Thus, when a characteristic value that indicates a factor causing a difference in the level of driving performance is extracted, a plurality of candidate data fragments that constitute candidates to be extracted are generated from a single vehicle data fragment. Frequency solving is performed on all vehicle data fragments in the normalized vehicle data groups DA and DB in step S103, whereby a large amount of candidate data is generated.
[00111] Then, in step S105 of Figure 3, a processing to reveal the characteristic value contained in the generated candidate data is performed, for example, by means of a window function operation based on the following equation.
[00112] [Mat. 1]

[00113] Figure 7 is an enlarged view of the A1 data in region 5b of Figure 6(b) described above. In this processing, a range from the starting point Ps (distance "0") to a predetermined travel point (distance "30") of the travel area Ar1 is selected from the data A1. Window function operation is performed on the selected range of data. As a result of this, a difference between the data in the distance range "0" to "15" and the data in the distance range "16" to "30" of the A1 data is obtained, whereby the characteristic value contained in each data track is revealed. The window function operation that applies a window function is performed on the data in the entire range of data A1, that is, the data corresponding to the travel area Ar1 from the starting point Ps to the endpoint Pg shown in Figure 4, whereby all characteristic values contained in the A1 data are revealed. Similarly, a window function operation is performed on data B1 to B3 obtained through the frequency solution, whereby the characteristic values contained in data B1 to B3 are revealed. Thus, characteristic values are revealed for all vehicle data in vehicle data group DA and DB obtained through the frequency solution in step S104.
[00114] In step S106 of Figure 3, a characteristic value that differs between a group of candidate data generated from the DA vehicle data group that belong to the "high" driving performance level and a group of candidate data generated to From the DB vehicle data group that belong to the "low" driving performance level is selected through learning with the use of the learning algorithm. The weighting of the characteristic value is performed by calculating the influence of the characteristic value on the evaluation result based on the evaluation criteria through the aforementioned learning (step S107).
[00115] Thus, as shown in Figure 8(a) by way of example, when the evaluation item is "fuel efficiency", for example, the characteristic value that exerts the greatest influence on fuel efficiency is a characteristic value of "steering angle A1" which indicates a characteristic of a steering mode of operation at travel point Pa3 on curve Cv2 contained in travel area Ar1, as shown in Figure 8(b) corresponding to Figure 4. characteristic value of "steering angle A1" is identified. This means that this steering angle A1 is identified as the characteristic value that exerts the most significant influence on the vehicle data evaluated based on fuel efficiency among the characteristic values contained in the generated candidate data group. The waypoint Pa3 is a point located at a predetermined distance from the starting point of the Cv2 curve.
[00116] A characteristic value that exerts the second greatest influence on fuel efficiency after steering angle A1 is a characteristic value of "steering angle B3" which indicates a steering operating mode at a waypoint Pa2 present at a predetermined distance from the starting point of the route section Sec2 in the route area Ar1 shown in Figure 8(b). Therefore, the characteristic value "steering angle B3" is identified.
[00117] A characteristic value that exerts the third major influence on fuel efficiency after steering angle B3 is "throttle B3" which indicates a mode of accelerator pedal operation at a waypoint Pa1 that extends over a run section Sec1 and a curve Cv1 that follows run section Sec1 in run area Ar1 shown in Figure 8(b). Therefore, "accelerator B3" is identified.
[00118] Such an operation is performed for each of the route areas Ar1 to Arn in which the vehicles that serve as the vehicle data supply sources have made a route, through which, as shown in Figure 8(a), the characteristic values that influence the fuel efficiency of the traffic elements or route sections present in the route areas Ar1 to Arn are weighted. As shown in Figure 8(a) described above, a correspondence relationship between the weighted characteristic values and the waypoints is obtained.
[00119] According to the present modality as described above, when vehicle data groups are categorized by driving performance levels for the "fuel efficiency" evaluation item, for example, a characteristic value that indicates a mode Of operation of a driving element that constitutes a factor causing a difference in fuel efficiency between the vehicle data groups is identified for each of the waypoints contained in the waypoints. At the same time, a traffic element or route section containing each of the route points is also identified.
[00120] In the following, the frequency solution processing by the frequency solution unit 252 will be described in detail with reference to Figure 9.
[00121] In this process, as shown in Figure 9, one of a plurality of types of driving elements that include an accelerator pedal, a brake pedal, a steering wheel and the like is selected as a driving element to be solved by frequency (step S201). Thus, a group of vehicle data which indicates, for example, time series transition in amount of depression of the accelerator pedal is selected from the groups of vehicle data normalized based on the various types of driving elements.
[00122] Subsequently, when variable i is to count N pieces of selected vehicle data that indicate the transition of time series in amount of throttle pedal depression, "0" is assigned to variable i (step S202). Next, a vehicle data fragment is selected from the N vehicle data fragments that indicate time series transition in amount of throttle pedal depression (step S203).
[00123] The selected vehicle data fragment is solved by frequency (step S204), and one is added to variable i (step S205). Since the selected vehicle data fragment is selected by frequency, the other vehicle data fragments that indicate a time series transition in amount of accelerator pedal depression that was not resolved by frequency are resolved by frequency sequentially, and one is added to variable ia each time (NOT in step S206; steps S203 to S205).
[00124] All vehicle data fragments N that indicate time series transition in amount of accelerator pedal depression were solved by frequency and variable i achieves "N" (YES in step S206), a different driving element the accelerator pedal for which the frequency solution has been completed is newly selected (step S201). Frequency solution processing sequentially of the data in the vehicle data group for the selected driving element is repeated until all vehicle data groups based on all types of driving elements stored in central storage device 220 are resolved by frequency (steps S202 to S206).
[00125] The vehicle data analysis method and the system according to the present embodiment described above provide advantages as follows.
[00126] The analysis system and method collects a plurality of vehicle data fragments based on a plurality of driving operation types, and categorizes the collected vehicle data into two groups according to an evaluation criterion as one index to assess a level of driving operation. The analysis system and method then extract a characteristic value from vehicle data, which differs between the categorized groups. The analysis system and method are thus enabled to quantitatively analyze the driving operation characteristic contained in the vehicle data.
[00127] The collected vehicle data group is acquired from vehicles that made a journey on a real road under a plurality of types of driving operations by one driver. Therefore, using the vehicle data group makes it possible to generate route models that reflect actual route environments and the driving operations performed under the route environment. In that case, a route model that incorporates a characteristic value to determine a level of driving performance can be generated by generating the route model that serves as a standard based on the characteristic value extracted from the vehicle data. In other words, the vehicle data analysis result can be applied to generate a route model according to a vehicle route environment, the driver's driving technique or a driving operation pattern peculiar to the driver.
[00128] The system and method of analysis obtain a degree of influence exerted by the characteristic value extracted from vehicle data on vehicle data to be evaluated based on the evaluation criteria. The analysis system and method are thus not only able to specify a characteristic value that differs between the categorized vehicle data groups, but also able to specify a degree of influence exerted by the characteristic value for the vehicle data evaluated with based on an evaluation criterion, in other words, for the evaluation based on the evaluation criterion.
[00129] The analysis system and method categorizes vehicle data into groups and extracts a characteristic value from vehicle data by considering any one of traffic elements, route sections and route areas as units. Thus, the system and the analysis method are able to extract a characteristic value from vehicle data that reflects a series of driving operations performed on traffic elements, route sections, and route areas, by each traffic element, by each course section and for each course area. This allows the system and analysis method to identify a factor that causes a difference in the level of driving performance between the vehicle data by considering traffic elements, route sections and route areas as units. In addition, the system and the analysis method are able to extract a characteristic value from a single piece of vehicle data for each traffic element, for each route section and for each route area, and are therefore capable of extract more characteristic values from a group of vehicle data collected by vehicle data analysis center 200.
[00130] The system and the analysis method detect the driving operations of a driver through sensors 101 to 106. The system and the analysis method treat, as vehicle data from vehicle 100, these detection results and the information latitude and longitude that allow the identification of a waypoint where the driving operation is performed by the driver and information that is related to various road maps. In the embodiment shown in Figures 8(a) and 8(b), the system and the analysis method further obtain a correspondence relationship between the characteristic value of conduction operation and waypoint, and a correspondence relationship between the characteristic value and the evaluation criteria. The analysis system and method are thus able to analyze the vehicle data in more detail, and are able to analyze a factor that causes a difference in the level of driving performance based on an evaluation criterion and an exerted influence by this factor on the evaluation criterion thoroughly for a level of waypoints.
[00131] The system and analysis method normalizes time series data as vehicle data based on a course position. This allows the system and analysis method to transform the time series data acquired from vehicles that have different travel speeds to levels that can be compared to a high degree of accuracy. Thus, the system and the analysis method are able to analyze more vehicle data (time series data), and are able to extract more characteristic values that differ between the vehicle data.
[00132] The system and method of analysis use, as characteristic values, data that indicate characteristics of a plurality of driving elements that indicate modes of operation of driving a driver. The system and method of analysis obtain a degree of influence exerted on an evaluation criterion by a plurality of types of driving elements that include an accelerator pedal, a brake pedal, a steering wheel and the like, for each of the driving elements. driving. Thus, the analysis system and method are able to accurately identify a plurality of factors that cause a difference in the level of driving performance between the vehicle data even though there are a plurality of driving elements that affect fuel efficiency, vehicle behavior and vehicle lateral jerk 100. The system and the analysis method are able to accurately extract the characteristic values that are locally contained in the vehicle data, and are able to specify a degree of influence exerted by each one of the driving elements on the assessment criterion.
[00133] The system and the analysis method generate a plurality of candidate data that serve as original data that indicate the characteristic values of the vehicle data by means of the frequency solution of the vehicle data. Therefore, the analysis system and method are able to extract various frequency components contained in the vehicle data, and are able to reveal various characteristic values that reflect driving operations. During the extraction of characteristic values that differ between the grouped vehicle data, the system and the analysis method are able to generate a large amount of candidate data that contain the characteristic value as candidates for those to be extracted, from a quantity. limited vehicle data.
[00134] Before extracting a characteristic value from the vehicle data, the system and analysis method reveals the characteristic value by applying a window function to the vehicle data. This allows the system and analysis method to accurately extract a characteristic minute value contained in the vehicle data, and accurately extract the characteristic values that differ between the vehicle data group DA and the vehicle data group DB obtained by grouping vehicle data by driving performance levels.
[00135] As evaluation items to be evaluated based on the evaluation criteria, the system and method of analysis use the fuel efficiency, which is defined by the vehicle travel distance per unit quantity of fuel, the travel time , the vehicle behavior and the vehicle's lateral bump. This allows the system and analysis method to identify a factor causing a difference between vehicle data in fuel efficiency, travel time, smooth driving operation (fast braking) or jolt by extracting characteristic values from the vehicle data categorized with an index of each assessment item.
[00136] The system and analysis method uses a grouping based on an evaluation criterion to categorize a plurality of vehicle data types into two vehicle data groups, namely a performance level DA vehicle data group driving performance level "high" and a driving performance level DB vehicle data group "low", and weight the characteristic value of the vehicle data, which differs in driving performance level between the DA vehicle data groups and DB. This allows the system and analysis method to identify the factors causing a difference in the level of driving performance between the DA and DB vehicle data groups in order of degree of influence for the evaluation criteria. Thus, the analysis system and method are able to determine the driving operation elements to be improved, which are shown in the vehicle data evaluated as "low" in driving performance level, in the order of priority for improvement.
[00137] The modality described above can be incorporated in the forms as described below.
[00138] The transmission of vehicle data from the vehicle 100 to the vehicle data analysis center 200 is carried out on the condition that the accessory position of the vehicle 100 can be switched from the ON state to the OFF state upon completion of driving of vehicle 100. Meanwhile, the transmission of vehicle data from vehicle 100 to vehicle data analysis center 200 can be performed, for example, at a time when the accessory position of vehicle 100 is switched from the OFF state to the ON state, or at a time when vehicle 100 passes through a predetermined traffic element or a predetermined route section. Similarly, transmission of vehicle data from vehicle 100 to vehicle data analysis center 200 can be performed at predetermined intervals. In addition, transmission of vehicle data from vehicle 100 to vehicle data analysis center 200 can be performed when vehicle data analysis center 200 makes a request to that effect to vehicle 100.
[00139] The transmission of vehicle data from the vehicle 100 to the vehicle data analysis center 200 is performed via wireless communication between the vehicle internal communication device 120 and the central communication device 210. The data transmission from vehicle 100 to vehicle data analysis center 200 can be performed via wired communication or the like with the use of an external storage medium such as a USB memory. In other words, any medium capable of transmitting vehicle data acquired by vehicle 100 to vehicle data analysis center 200 can be used as a vehicle data transfer unit.
[00140] The weighting of characteristic values of vehicle data is accomplished through learning using "AdaBoost" as a learning algorithm. The setting used for the weighting is not limited to this, but it can be any setting that can perform weighting on characteristic values that differ between vehicle data group DA and vehicle data group DB, which are different from each other. at the driving performance level, and various other learning or operations algorithms can be used.
[00141] A degree of influence exerted by a characteristic value of vehicle data on the vehicle data to be evaluated with an evaluation criterion is determined by weighting the characteristic value of vehicle data, which differ between groups of DA and DB categorized vehicle data. The degree of influence can be determined by categorizing, for example, into four levels: "high degree of influence", "medium degree of influence", "low degree of influence", and "no influence". In this case, among the characteristic values of groups of vehicle data collected in a route section, a route area or a common or similar traffic element, those characteristic values which differ by a relatively high frequency between the St1 data and the data St2 shown in Figures 5(a) and 5(b), and those characteristic values which exhibit a noticeable difference between the data are determined to be of "high degree of influence". On the other hand, among the characteristic values of groups of vehicle data collected in a route section, a route area or a common or similar traffic element, those characteristic values which differ at a relatively low frequency between the St1 data and the St2 data shown in Figure 5, or those characteristic values which exhibit a minute difference between the data are determined to be of "low degree of influence". Furthermore, when the characteristic values are common between the DA vehicle data groups and the grouped DB vehicle data groups, they are determined to be "no influence" as they are not a factor causing a difference in the driving performance level between vehicle data.
[00142] The group of vehicle data collected by the vehicle data analysis center 200 is categorized into two groups of "high" driving performance level and "low" driving performance level based on the evaluation criteria. Instead, the vehicle data group can be categorized into three or more groups based on the evaluation criteria. For example, when the vehicle data group is categorized into three, first to third groups in descending order of driving performance level, a characteristic vehicle data value that differs, for example, between the first and second groups is extracted. In this case, the extracted characteristic value is identified as a factor causing a difference in the driving performance level between the first group of the highest driving performance level and the second group of the second highest driving performance level. Similarly, a characteristic value of the vehicle data that differs between the second and third groups is identified as a factor causing a difference in driving performance level between the second and third groups. This makes it possible to identify a factor causing a difference in driving performance level between groups even when the driving performance level of the vehicle data can be categorized into a plurality of groups with different driving performance levels.
[00143] The weighting of a vehicle data characteristic value that differs between the DA and DB vehicle data groups is performed, as shown in Figure 8(a), for the "fuel efficiency" evaluation item. As shown in Figures 10(a) and 10(b), a group of vehicle data is grouped for each of the evaluation items among bump, vehicle behavior and travel time, and the weighting can be performed on a characteristic value that differs between these grouped vehicle data. In this case, it is possible to identify, for each waypoint, a factor that causes a difference in the level of driving performance that is evaluated not only based on fuel efficiency, but also based on each of the evaluation items within bumps , vehicle behavior and travel time.
[00144] The grouping of the vehicle data groups described above is performed based on one of the evaluation items among fuel efficiency, travel time, vehicle behavior and vehicle lateral bump. Instead, grouping of vehicle data groups can be performed based on assessment criteria for two or more assessment items. In this case, when the assessment items are fuel efficiency and vehicle behavior, for example, vehicle data indicating low fuel consumption and small vehicle behavior are assessed as "high" driving performance level, while the others vehicle data is evaluated as "low" driving performance level. Extracting characteristic values that differ between the "high" driving performance level and the "low" driving performance level makes it possible to extract a characteristic value from vehicle data that indicate a driving operation that is capable of simultaneously achieving established assessment criteria for a plurality of assessment items.
[00145] The modalities described above employ, as the evaluation items, fuel efficiency, travel time, vehicle behavior and the vehicle's lateral bump. In addition to these, other evaluation items can be used to group vehicle data as long as they reflect a driver's driving technique such as driving operation stability, distance between vehicles and the like. When the distance between vehicles is used as an evaluation item, it can be established as an evaluation criterion, for example, if a distance between vehicles between a vehicle that serves as a source of vehicle data supply and another vehicle ahead in the direction of travel of the aforementioned vehicle is kept equal to or greater than a predetermined distance or not. It is also possible to establish, as an evaluation criterion, whether a variation in distance between vehicles between a vehicle that serves as a source of vehicle data supply and another vehicle path in front of the aforementioned vehicle travel direction is within a certain limit.
[00146] In the embodiments described above, the revelation of a characteristic value of the vehicle data is performed by means of an operation using a window function. The invention is not limited thereto, and any operation capable of revealing a characteristic value contained in the vehicle data can be employed to reveal the vehicle data. Any configuration can be employed as long as it can extract a characteristic value from vehicle data that differs between groups of data grouped based on an evaluation criterion. The configuration in which the window function operating unit 253 is omitted need not perform an operation to reveal the characteristic value of the vehicle data.
[00147] In the embodiments described above, the vehicle data frequency solution is performed by means of a wavelet transform. Instead of the wavelet transform, a discrete cosine transform or a Fourier transform, for example, can be used to perform the frequency solution of the vehicle data. In this case, a preferred transform technique is an invertible transformable technique. A configuration that is capable of extracting frequency components from a plurality of frequency bands by resolving the frequency of the vehicle data can be used. What is required for the configuration is to be able to extract characteristic values from vehicle data that differ between groups obtained by grouping them based on an evaluation criterion. The configuration in which the frequency solution unit 252 is omitted does not need to perform a frequency solution on the vehicle data.
[00148] The classification result storage unit 240 described above is configured to have the first storage area 241 and the second storage area 242 in which vehicle data categorized into groups according to the level of driving performance are stored. Configuration is not limited to this. The configuration may omit classification result storage unit 240 and may be such that distinguishable information is assigned to each piece of vehicle data stored in central storage device 220. Distinguishable information is information by which the classification unit of 230 vehicle data is allowed to distinguish driving performance levels. In that case, the vehicle data classification unit 230 issues to the vehicle data analysis unit 250 assigned vehicle data groups with information indicating opposite driving performance levels. The vehicle data analysis unit 250 then extracts characteristic values that differ between vehicle data that have opposite driving performance levels based on the distinguishable information assigned to the vehicle data.
[00149] Characteristic values that indicate the characteristics of a plurality of driving elements that include an accelerator pedal, a brake pedal and a steering wheel that indicate a driver driving mode of operation are extracted from data groups of vehicle collected by the vehicle data analysis center 200. The invention is not limited thereto and, for example, a characteristic value that indicates a characteristic of one of the driving elements which includes an accelerator pedal, a brake pedal and a steering wheel can be extracted from the vehicle data groups.
[00150] In the modalities described above, vehicle data reflecting a driver driving operation is acquired based on detection results by the throttle sensor 101, by the brake sensor 102, by the steering angle sensor 103, by the sensor of spin 104, by vehicle speed sensor 105 and by acceleration sensor 106. Instead, vehicle data can be acquired based on a detection result through a steer rate sensor to detect a steer rate, which indicates a rate of rotation angle change in the vehicle turning direction 100. In this case, a factor causing a difference in the level of driving performance due to the vehicle's steering rate is extracted as the characteristic value of the data of vehicle. Furthermore, as the vehicle data acquisition unit a configuration can be employed which is capable of acquiring a driver driving mode of operation from a signal that reflects the driving operation.
[00151] The vehicle data analysis unit 250 described above is configured to have the normalizing operation unit 251. The normalizing operation unit 251 normalizes the time series data as the vehicle data based on a position of course. The invention is not limited thereto, and when time series data indicating modes of operation of driving in the same travel environments or similar travel environments have data lengths that are comparably closer together, the operating unit of normalization 251 can be omitted and normalization of the time series data with the vehicle data can be omitted.
[00152] In the above modalities, both the correspondence relation between the characteristic value extracted from the vehicle data and the waypoint and the correspondence relation between the characteristic value and the evaluation criterion are determined. The invention is not limited to this, and only the correspondence relation between the characteristic value extracted from the vehicle data and the waypoint can be determined, or only the correspondence relation between the characteristic value and the evaluation criterion can be determined .
[00153] In the above embodiments, vehicle data having information based on which a waypoint can be identified is collected from vehicle 100. The present invention is not limited thereto, and vehicle data having information on the basis of which a traffic element, a route section and a route area can be identified instead of a route point can be collected from vehicle 100. Then a correspondence relationship between the traffic element, the route section, and the route area and the characteristic value of the vehicle data thus collected can be determined. Similarly, a correspondence relationship between the traffic element, the route section and the route area thus collected and an evaluation criterion can be determined.
[00154] In the above embodiments, the weighting of characteristic values of the vehicle data is performed for each waypoint contained in the waypoints Ar1 to Arn as shown in Figure 8(a). However, the invention is not limited to this, and as shown in Figures 10(a) and 10(b) which correspond to Figure 8(a), the weighting of characteristic values of the vehicle data can be performed for each of the elements such as bends or intersections that have common or similar lane alignments, or similar. In addition, as shown in Figures 10(a) and 10(b), weighting of characteristic values of vehicle data can be performed for each of the track sections that have common or similar track alignments, or the like. In such cases, characteristic values that have high degrees of influence are classified for each of the traffic elements or each of the route sections, whereby the characteristic values can be extracted in more detail for each of the traffic elements. traffic or each of the route sections. In that case, vehicle data collected from vehicle routes in different route areas can be an object to be grouped when characteristic values that differ between groups are extracted, provided that the road alignments or similar of the vehicle data are similar to each other. This makes it possible to extract more characteristic values from vehicle data collected from a wider route area.
[00155] Traffic elements, route sections and route areas are treated as units, the vehicle data is grouped and the characteristic values of the vehicle data are extracted. The invention is not limited thereto, and at least one of the traffic elements, the route sections and the route area can be treated as units, the vehicle data is grouped and the characteristic values of the vehicle data are extracted.
[00156] In the modalities above, the grouping based on the evaluation criteria is performed on data that indicate operating modes of driving in a common route environment or similar. The invention is not limited thereto, and grouping based on the evaluation criteria can be performed, for example, only on data indicating driving operating modes in a common route environment.
[00157] In the above modalities, information that relates to traffic elements such as curves and intersections, route sections and route areas is established as information that relates to vehicle 100 route environments that serve as a source of supply of vehicle data. Also, information indicating travel areas in which traffic congestion frequently occurs, or information indicating travel times in which traffic congestion occurs frequently, time zones in which a common traffic volume is observed, and weather conditions at the time the driving operations indicated by the vehicle data are carried out can be set as the information relating to the travel environments of the vehicle 100. The vehicle data can be grouped and can be extracted according to a state of congestion traffic or weather that can be specified based on such information. Vehicle data groups that are common in traffic element, route section or route area in which the driving operation indicated by the vehicle data is performed, and also common and traffic congestion state or weather in the route area , are defined as a common vehicle data group. In this case, the common vehicle data group is specified as vehicle data to be an object for grouping and extracting characteristic values. In this case, grouping and extraction of characteristic values is performed on common vehicle data groups based on an evaluation criterion. Thus, a difference in characteristic value between vehicle data caused by a difference in travel environment can be accurately distinguished from a difference in characteristic value between vehicle data caused by a difference in driver driving technique. This makes it possible to extract a factor from the vehicle data group more accurately, the factor causing a difference in characteristic value caused only by the driver's driving technique, in other words, a factor causing a difference in the performance level of driving between vehicle data even though driving operations are performed in the same driving environment.
[00158] In the above embodiments, information indicating a vehicle travel environment 100 from which vehicle data is collected is contained in the vehicle data. The invention is not limited to that. When vehicle data collected under a common or similar route environment can be identified based on a vehicle data transition to extract characteristic values, the GPS 108 or car navigation system 109 can be omitted, and only the data that indicate the transition of the driver driving operation can be collected as the vehicle data.
[00159] In the above modalities, a degree of influence exerted by the characteristic value extracted from vehicle data on the vehicle data evaluated based on the evaluation criteria is determined. The invention is not limited thereto, and only characteristic values that differ between groups that have been grouped based on the evaluation criteria can be extracted from the vehicle data groups. Also in this case, a factor causing a difference in the level of driving performance between the groups can be quantitatively specified according to the characteristic values extracted from the vehicle data.
[00160] In the above embodiments, vehicle data characteristic values are extracted from vehicle data groups collected from a large number of vehicles without specifying vehicle types. The invention is not limited to this, and only a group of vehicle data collected from the same type of vehicles can be used for the analysis when a difference in characteristic value of vehicle data caused only by the driver driving operation is extracted from the vehicle data group. In this case, individual variability among vehicles is removed, and characteristic values attributable only to a difference in the driver's driving technique can be accurately extracted. This allows for more accurate analysis.
[00161] In the above embodiments, vehicle data is acquired from a plurality of vehicles. The invention is not limited to this, and the vehicle data to be analyzed can be acquired from a single vehicle. Vehicle data to be analyzed can be acquired based on driving operations of the same driver if the vehicle data is different in driving operation. In other words, any vehicle data that can be grouped based on the evaluation criteria and reflects a plurality of types of driving operations can be used as an object of analysis.

权利要求:
Claims (14)
[0001]
1. Vehicle data analysis method for analyzing vehicle data reflecting a driver driving operation comprising: collecting (S101) a plurality of vehicle data fragments based on a plurality of driving operation types; (S102) these collected fragments of vehicle data in at least two groups based on an evaluation criterion which is an index to evaluate a level of driving operation; and extracting (S105) characteristic values from different vehicle data between groups, the method characterized by the fact that it further comprises: normalizing (S103) the time series data as vehicle data based on a track position.
[0002]
2. Vehicle data analysis method, according to claim 1, characterized in that it further comprises: obtaining (S106) a degree of influence exerted by the characteristic value extracted from vehicle data on the vehicle data evaluated based on in the evaluation criteria.
[0003]
3. Vehicle data analysis method according to claim 1 or 2, characterized in that the vehicle data includes information indicating at least one of a traffic element, a route section and a route area where the traffic element and the route sections are connected in series, and the grouping (S102) of the vehicle data and the extraction (S105) of the characteristic values of the vehicle data is performed by processing the traffic element of the route section or the route area as a unit.
[0004]
4. Vehicle data analysis method according to any one of claims 1 to 3, characterized in that the vehicle data includes information indicating a waypoint and the vehicle data analysis method further comprises: obtain a correspondence relationship between the characteristic values extracted from the vehicle data and the waypoint and a correspondence relationship between the characteristic values extracted from the vehicle data and an evaluation result of the vehicle data based on the evaluation criterion.
[0005]
5. Vehicle data analysis method, according to any one of claims 1 to 4, characterized in that the characteristic values indicate characteristics of one or more driving elements that represent a driver driving operating mode, and the method Analysis further comprises obtaining (S106) a degree of influence exerted by the driving element in the evaluation criterion for each of the driving elements.
[0006]
6. Vehicle data analysis method according to any one of claims 1 to 5, characterized in that it further comprises: generating a plurality of candidate data as original data to indicate the characteristic values of the vehicle data by means of the vehicle data frequency solution.
[0007]
7. Vehicle data analysis method according to any one of claims 1 to 6, characterized in that it further comprises: revealing characteristic values by applying a window function to vehicle data before extracting the values characteristics of vehicle data.
[0008]
8. Vehicle data analysis method, according to any one of claims 1 to 7, characterized by the fact that the evaluation criterion is a criterion for grouping that is performed for at least one among the evaluation items that consist of fuel efficiency defined by a vehicle's travel distance per fuel unit amount, travel time, vehicle behavior, and a vehicle's lateral bump.
[0009]
9. Vehicle data analysis method according to any one of claims 1 to 8, characterized in that the grouping (S102) comprises categorizing the plurality of vehicle data types into a vehicle data group of level of "high" driving performance and a "low" driving performance level vehicle data group by grouping based on the evaluation criterion, the extraction (S105) of the vehicle data comprises performing weighting on characteristic values of vehicle data that differs in driving performance level based on a total value of characteristic value differences that are common between the "high" driving performance level data group and the performance level vehicle data group "low" conduction values, which are grouped based on the evaluation criteria, and the difference of the common characteristic values is a difference between the characteristic value of the data categorized by the d level group. and "high" driving performance and the characteristic value of the data categorized by the "low" driving performance level group.
[0010]
10. Vehicle data analysis system for analyzing vehicle data reflecting a driver driving operation comprising: a storage device (220) for storing vehicle data based on a plurality of types of driving operations; a vehicle data classification unit (230) for grouping the vehicle data stored in the storage device (220) into at least two groups based on an evaluation criterion which is an index for evaluating a level of driving operation; and a vehicle data analysis unit (250) for extracting characteristic values of vehicle data that differ between the groups grouped by the vehicle data classification unit (230), characterized by the fact that the vehicle data analysis unit (250) further comprises a normalizing operation unit (251) for normalizing the time series data as the vehicle data based on a course position.
[0011]
11. Vehicle data analysis system according to claim 10, characterized in that the vehicle data analysis unit (250) further comprises an influence calculation unit (254) to obtain a degree of influence exerted by the extracted characteristic value on the vehicle data evaluated based on the evaluation criteria.
[0012]
12. Vehicle data analysis system according to claim 10 or 11, characterized in that the vehicle data includes information indicating a traffic element, a route section and a route area where the traffic element and the route section are connected in series, the vehicle data classification unit and the vehicle data analysis unit perform the grouping of the vehicle data and the extraction of characteristic values from the vehicle data by processing the element of traffic, route section or route area as a unit.
[0013]
13. Vehicle data analysis system, according to any one of claims 10 to 12, characterized in that the vehicle data includes information that indicates a waypoint; and the vehicle data analysis unit (250) further obtains a correspondence relationship between the characteristic values extracted from the vehicle data and the waypoint and a correspondence relationship between the characteristic values extracted from the vehicle data and an evaluation result vehicle data based on the evaluation criteria.
[0014]
14. Vehicle data analysis system, according to any one of claims 10 to 13, characterized in that the characteristic values are characteristics of one or more driving elements that indicate a driver driving mode of operation, and the unit The vehicle data analysis tool (250) determines a degree of influence exerted by the driving element on the evaluation criterion for each of the driving elements.

类似技术:

---

公开号 | 公开日 | 专利标题

---

BR112013029418B1|2021-06-15|VEHICLE DATA ANALYSIS METHOD AND VEHICLE DATA ANALYSIS SYSTEM

---

Eboli et al.2016|Combining speed and acceleration to define car users’ safe or unsafe driving behaviour

---

JP6069650B2|2017-02-01|Traffic sign recognition method

---

JP5315363B2|2013-10-16|Road network analysis system

---

CN103632540B|2015-11-25|Based on the major urban arterial highway traffic circulation information processing method of floating car data

---

CN101958043A|2011-01-26|Highway section calculation element hourage and method thereof

---

US10775187B2|2020-09-15|Method, apparatus, and computer program product for determining lane level traffic information

---

Budhkar et al.2017|Characteristics of lateral vehicular interactions in heterogeneous traffic with weak lane discipline

---

CN102951147A|2013-03-06|Method for carrying out velocity warning for motor vehicle and device thereof

---

US20190137289A1|2019-05-09|Detection of Map Anomalies

---

CN107111942A|2017-08-29|For detecting motor vehicle by the current method and apparatus of traffic sign door

---

JP6136607B2|2017-05-31|Road condition identification device

---

Kusano et al.2014|Methodology for identifying car following events from naturalistic data

---

CN106463050A|2017-02-22|Method for processing measurement data of a vehicle in order to determine the start of a search for a parking space and computer program product

---

CN110264725B|2021-08-03|Method and device for determining road section flow

---

JP2014106675A|2014-06-09|Required time calculation device and required time calculation method

---

CN106323318A|2017-01-11|Method and device for processing transit time of navigation route

---

CN111351499A|2020-06-30|Path identification method and device, computer equipment and computer readable storage medium

---

Brunauer et al.2014|Deriving driver-centric travel information by mining delay patterns from single GPS trajectories

---

WO2018045974A1|2018-03-15|Method, apparatus, and system for determining traffic light timing

---

Muñoz-Organero et al.2017|Detecting different road infrastructural elements based on the stochastic characterization of speed patterns

---

Jomrich et al.2019|Lane Accurate Detection of Map Changes based on Low Cost Smartphone Data.

---

JP2018190297A|2018-11-29|Evaluation program, evaluation method and evaluation device

---

Wolshon et al.2015|Identifying high-risk roadways for infrastructure investment using naturalistic driving data

---

US10820166B1|2020-10-27|Systems and methods for obtaining location intelligence

同族专利:

---

公开号 | 公开日

---

CN103562978B|2015-11-25|

---

ES2559824T3|2016-02-16|

---

WO2012157255A1|2012-11-22|

---

US9053593B2|2015-06-09|

---

JP5856387B2|2016-02-09|

---

JP2012242894A|2012-12-10|

---

BR112013029418A2|2017-01-31|

---

EP2710572A1|2014-03-26|

---

EP2710572B1|2015-11-04|

---

CN103562978A|2014-02-05|

---

US20140365029A1|2014-12-11|

引用文献:

---

公开号 | 申请日 | 公开日 | 申请人 | 专利标题

---

---

US6879969B2|2001-01-21|2005-04-12|Volvo Technological Development Corporation|System and method for real-time recognition of driving patterns|

---

CA2925145A1|2003-07-07|2005-01-13|Insurance Services Office, Inc.|Traffic information system|

---

JP4682714B2|2005-06-14|2011-05-11|トヨタ自動車株式会社|Dialog system|

---

EP1780090A1|2005-10-28|2007-05-02|CRF Societa'Consortile per Azioni|Method for defining the driving style of a vehicle driver|

---

WO2007077867A1|2005-12-28|2007-07-12|National University Corporation Nagoya University|Drive behavior estimating device, drive supporting device, vehicle evaluating system, driver model making device, and drive behavior judging device|

---

JP2008140118A|2006-12-01|2008-06-19|Advanced Telecommunication Research Institute International|Hazard motion detection device and hazard motion detection method|

---

JP4491472B2|2007-03-27|2010-06-30|日立オートモティブシステムズ株式会社|Traffic information system|

---

JP2008298979A|2007-05-30|2008-12-11|Pioneer Electronic Corp|Evaluation device for driving skill and method for evaluating driving skill and program for evaluating driving skill|

---

JP4997011B2|2007-07-25|2012-08-08|日立オートモティブシステムズ株式会社|Automotive fuel consumption estimation system, route search system, and driving guidance system|

---

JP5005492B2|2007-10-17|2012-08-22|富士重工業株式会社|Safe driving support system|

---

WO2009104256A1|2008-02-20|2009-08-27|パイオニア株式会社|Operational evaluation device of vehicle, method, and computer program|

---

JP5133784B2|2008-06-02|2013-01-30|トヨタ自動車株式会社|Driving operation evaluation device|

---

JP5185220B2|2008-07-18|2013-04-17|みずほ情報総研株式会社|Environmental impact assessment support system, environmental impact assessment support method, and environmental impact assessment support program|

---

WO2010023802A1|2008-08-25|2010-03-04|本田技研工業株式会社|Navigation server|

---

JP5152865B2|2008-12-22|2013-02-27|Ｕｄトラックス株式会社|Fuel efficient driving evaluation system|

---

US8170725B2|2009-02-18|2012-05-01|GM Global Technology Operations LLC|Vehicle stability enhancement control adaptation to driving skill based on highway on/off ramp maneuver|

---

JP2011003076A|2009-06-19|2011-01-06|Fuji Heavy Ind Ltd|Risk recognition system|

---

JP2011065527A|2009-09-18|2011-03-31|Toyota Motor Corp|Driving evaluation system, vehicle-mounted machine, and information processing center|

---

CN101694744B|2009-10-28|2012-01-11|北京交通大学|Method and system for evaluating road emergency evacuation capacity|

---

IL201810A|2009-10-29|2015-06-30|Greenroad Driving Technologies Ltd|Method and device for evaluating a vehicle's fuel consumption efficiency|

---

US8793064B2|2009-12-10|2014-07-29|International Business Machines Corporation|Vehicle fuel efficiency optimization based on vehicle usage patterns|

---

US8825281B2|2010-04-09|2014-09-02|Jacques DeLarochelière|Vehicle telemetry system and method for evaluating and training drivers|

---

EP2431244A1|2010-09-20|2012-03-21|Givenpath Ltd|Driver profiling system and method|

---

JP5375805B2|2010-11-26|2013-12-25|トヨタ自動車株式会社|Driving support system and driving support management center|CN103247091B|2012-02-07|2016-01-20|厦门金龙联合汽车工业有限公司|A kind of driving evaluation system and method|

---

JP5907249B2|2012-03-16|2016-04-27|日産自動車株式会社|Unexpected prediction sensitivity judgment device|

---

JP6160172B2|2013-03-28|2017-07-12|富士通株式会社|Driving diagnosis system, driving diagnosis program, and driving diagnosis device|

---

WO2015131351A1|2014-03-05|2015-09-11|华为终端有限公司|Data processing method, server and terminal for internet of vehicles|

---

CN103914985B|2014-04-25|2015-10-28|大连理工大学|A kind of hybrid power passenger car following speed of a motor vehicle trajectory predictions method|

---

CN104494601A|2014-12-18|2015-04-08|清华大学苏州汽车研究院（吴江）|Driving behavior analysis and driving assistance system based on OBD|

---

CN105774810B|2014-12-24|2019-05-07|中国电信股份有限公司|Vehicle drive behavior treating method and apparatus|

---

CN104640074A|2015-02-17|2015-05-20|北京九五智驾信息技术股份有限公司|Method for providing taxi pooling service|

---

JP6304104B2|2015-04-01|2018-04-04|トヨタ自動車株式会社|Driving environment evaluation system|

---

CN104935566A|2015-04-17|2015-09-23|北京大学深圳研究院|Method of updating numerical value system and device of updating numerical value system|

---

DE102015010270B4|2015-08-08|2021-10-28|Audi Ag|Method for operating driver assistance systems in a motor vehicle and motor vehicle|

---

JP6485306B2|2015-09-25|2019-03-20|株式会社デンソー|Control system|

---

EP3219567A1|2016-03-14|2017-09-20|Honda Research Institute Europe GmbH|Method, system and vehicle for analyzing a rider performance|

---

JP2017220166A|2016-06-10|2017-12-14|トヨタ自動車株式会社|Vehicle control system|

---

US10061314B2|2016-08-15|2018-08-28|Ford Global Technologies, Llc|Enhanced vehicle operation|

---

DE102017000608A1|2017-01-24|2018-07-26|Man Truck & Bus Ag|Method for operating a vehicle, in particular a utility vehicle|

---

CN107123173A|2017-04-25|2017-09-01|百度在线网络技术（北京）有限公司|For the method and apparatus for the transport condition for monitoring unmanned vehicle|

---

CN108132907B|2017-12-14|2021-03-02|一汽-大众汽车有限公司|Method and system for vehicle-mounted software input by combining vehicle information control system|

---

CN109584546B|2018-07-04|2021-06-01|东南大学|Method for determining departure flow threshold of quick-conventional bus sharing lane|

---

US10922970B2|2018-09-14|2021-02-16|ANI Technologies Private Ltd.|Methods and systems for facilitating driving-assistance to drivers of vehicles|

---

KR102169791B1|2018-11-19|2020-10-26|주식회사 글로비즈|Railway vehicle major component and system diagnosis apparatus|

---

CN112550298A|2019-09-26|2021-03-26|上海汽车集团股份有限公司|Driving behavior evaluation method, device and storage medium|

---

CN111688713B|2020-06-16|2021-10-22|东软睿驰汽车技术（沈阳）有限公司|Driving behavior analysis method and device|

---

CN112026679B|2020-08-21|2021-12-17|华人运通技术有限公司|Vehicle use condition monitoring method and device, medium, edge processor and vehicle|

法律状态:
2018-12-11| B06F| Objections, documents and/or translations needed after an examination request according [chapter 6.6 patent gazette]|

---

2019-10-29| B06U| Preliminary requirement: requests with searches performed by other patent offices: procedure suspended [chapter 6.21 patent gazette]|

---

2021-04-20| B09A| Decision: intention to grant [chapter 9.1 patent gazette]|

---

2021-06-15| B16A| Patent or certificate of addition of invention granted|Free format text: PRAZO DE VALIDADE: 20 (VINTE) ANOS CONTADOS A PARTIR DE 15/05/2012, OBSERVADAS AS CONDICOES LEGAIS. |

优先权:

---

申请号 | 申请日 | 专利标题

---

JP2011-109453|2011-05-16|

---

JP2011109453A|JP5856387B2|2011-05-16|2011-05-16|Vehicle data analysis method and vehicle data analysis system|

---

PCT/JP2012/003161|WO2012157255A1|2011-05-16|2012-05-15|Vehicle data analysis method and vehicle data analysis system|

---