• 专利附录
  • 专利说明
  • 权利要求
  • 类似技术
  • 同族专利
  • 引用文献
  • 法律状态
  • 优先权
    • 专利摘要:
    The invention relates to a method for controlling the movement of a load which is suspended at a point of attachment of a crane-type lifting device, said attachment point being designed to be able to move in a rotational movement by lace around a first vertical axis (ZZ '), and / or in a translation movement along a second axis (XX') intersecting said orientation axis (ZZ '), said method comprising a step of acquisition of speed reference (Vi), then a step of filtering said speed reference by means of a virtual model (X = AX + BU), which uses a state vector (X) comprising an angle component swaying of the suspended load and a skewed angular velocity component, and to which a feedback control (Ui = ki0.Vi-Ki.X) is applied, whose correcting gains (ki1, ki2) are expressed as a function of the same typing parameter (Tc) freely adjustable by the driver of the hoist.
    公开号:FR3016872A1
    申请号:FR1450751
    申请日:2014-01-30
    公开日:2015-07-31
    发明作者:Xavier Claeys
    申请人:Manitowoc Crane Group France SAS；
    IPC主号:
    专利说明:

    [0001] The present invention relates to the general field of hoisting machines, of the tower crane type, comprising a mobile attachment point, of the trolley type, to which one of the following types of hoist is attached. can suspend a load to be moved, as well as processes for assisting the piloting of such lifting gear. It is already known to provide a control method which aims to control, and more particularly to limit, the amplitude of pendular oscillations, or "dangling", which is the subject of the load suspended during the movements of the carriage, and this so to improve the accuracy and safety of the transport operations of said load. For this purpose, it is known in particular, for example from document FR-2 704 847, to first develop a carriage position instruction by filtering said instruction through a frequency-cut "hole" filter, intended to eliminate from said setpoint frequencies likely to excite an oscillating mode of the load, then to apply this setpoint to said carriage, and then to fight the ballast by a closed loop servo, in which the actual values of position are measured and of the speed of the carriage, as well as of the swing angle of the load, then a setpoint correction is generated which takes into account the difference found between the actual behavior of the carriage and the setpoint applied. While such a system can effectively mitigate ballooning, it may have some disadvantages. First, the simple filtering of the setpoint through a "hole" filter, which avoids certain resonance phenomena, is generally not sufficient to prevent the appearance of pendular oscillations, which must then be combated by means of a complementary closed-loop servocontrol. However, such a closed-loop servocontrol imposes the implementation of numerous sensors, intended for example to measure the actual angle of ballooning, which increases the complexity, and consequently the cost as well as the risk of failure, of the control system. piloting and more generally of the lifting machine that it equips.
    [0002] In addition, the complexity of the model used by this control system, as well as the amount of data to be measured and processed, tend to mobilize relatively large resources and expensive in terms of computing power, memory, and energy.
    [0003] Furthermore, the assistance offered, or even imposed on the operator of the machine (crane operator), may tend to excessively dampen the responses of the hoist to the instructions of the crane operator, thereby distorting the intuitive perception of the behavior of the crane operator. gear that may have said crane operator, including giving the latter the unpleasant impression that the machine lacks responsiveness and does not faithfully implement his instructions. The objects assigned to the invention therefore seek to overcome the aforementioned drawbacks and to propose a new method of controlling the movement of a suspended load which ensures effective control of the ball, while being particularly simple and economical to implement, and that provides a faithful feeling allowing a reactive and relatively intuitive piloting. The objects assigned to the invention are achieved by means of a method of controlling the movement of a load suspended at a point of attachment of a hoist, said point of attachment being designed to be able to move according to a yaw rotation around a first vertical axis (ZZ '), called "orientation axis", and / or in a translational movement along a second axis (XX') called "distribution axis" secant to said orientation axis (ZZ '), the displacement of the suspended load being controlled by a control setpoint which relates to a quantity, typically a kinematic quantity, characteristic of the movement of said charge, called "servo quantity", such that the rotational speed, and / or respectively the translation speed, of the point of attachment, said method comprising: a step of (a) target acquisition, during which a set value is acquired in real time pilot called "driving instruction te "Vi, which corresponds to the value of the servo-controlled quantity that the driver of the lifting machine aims at the moment considered, then a step (b) of setpoint filtering during which: one models the pendular behavior of the load suspended according to the movement considered by a virtual model = AX + BU which uses a state vector X comprising at least one component called "principal component", which corresponds to the controlled variable, and other components, called "components" additional "representative of characteristic kinematic variables of the swinging movement of the suspended load, such as the swinging angle or the angular speed of the balloon according to the movement considered, applying to said virtual model ii7 = AX + BU a control by return of state 1.1, = which involves, in addition to a Ko.V setpoint term, representative of the raw control setpoint acquired during the acquisition step, a correction term KX corresp to the product of a correction vector K, by the state vector X, using a correction vector K, in which at least some of the gain gains km, ki2 associated with the additional components are expressed as a function of the same parameter of predetermined Tc typing, said typing parameter Tc being freely adjustable by the driver of the lifting machine so as to adjust said gains km correctors, and one extract from the virtual model a control setpoint, called "filtered control setpoint Which corresponds to the main component of the state vector X. The objects assigned to the invention are more particularly achieved by means of a method for controlling the displacement of a load suspended at a point of attachment of a lifting gear, said attachment point 25 being designed to be able to move in a yawing rotation movement about a first vertical axis (ZZ '), called "axis of orientation", and / or according to a slack translational movement along a second axis (XX ') said "distribution axis" intersecting said orientation axis (ZZ'), said method comprising: a step of (a) instruction acquisition, during which a speed reference value V, representative of the speed of rotation, and / or respectively of the speed of translation, which the driver of the hoisting machine wishes to give to the point of attachment at a given point in time, is acquired in real time. the moment considered, then a step (b) of setpoint filtering during which: - the pendulum behavior of the suspended load is modeled according to the movement considered by a virtual model 21. = AX + BU which uses a vector of state X comprising at least one instantaneous speed component representative of the instantaneous speed of the point of attachment according to the movement in question, a component of a swinging angle representative of the swing angle of the load suspended according to the movement considered, anda balloon angular velocity component representative of the angular velocity of the balloon according to the movement considered, applying to said virtual model X = AX + BU a feedback control 1J, = which involves, in addition to a set term Ko. V, representative of the speed reference V, acquired during the acquisition step, a corrective term KX corresponding to the product of a corrector vector K, by the state vector X, using a corrector vector K, in wherein the correcting gains km, ki2 associated respectively with the balloon angular velocity component and the balloon angle component of the state vector X are expressed as a function of the same predetermined typing parameter Tc, said typing parameter Tc being freely adjustable by the driver of the lifting machine so as to be able to adjust said gains km correctors, and one extracts from the virtual model a setpoint of filtered speed, which corresponds to the component e of instantaneous velocity of the state vector X.
    [0004] Advantageously, the invention thus allows the operator of the machine to freely choose, from among a selection of predetermined values, or from a range of possible values set in advance, a typing parameter (coefficient) Tc which will enable him to characterize the filtering that will be applied to its speed setpoints, and more specifically to adjust the behavior of the slave system, defining the trade-off that best suits it between speed of response (reactivity) and stability (damping). In practice, this possibility of particularly simple and rapid adjustment of the filtering of the speed setpoint, advantageously common to the various correcting gains to which the same typing parameter is applied, makes it possible in fact to operate, by means of a single adjustment, a placement of the (complex) poles of the state feedback control.
    [0005] Of course, such a possibility of modifying the setting of the setpoint filter can be used both to change the custom setting from one driver to another, for example between an experienced crane operator who wishes to favor a reactive steering and low amortization and a novice crane operator 5 who will wish to privilege a cushioning operation very safe but slower to the reaction, than to adapt the behavior of the hoist according to the conditions of implementation of the site, for example according to the size, fragility or the weight of the load to be moved or even according to the meteorological conditions, and in particular the wind conditions, prevailing at the site at the moment considered. Advantageously, the use of a virtual pendulum model associated with a feedback control system makes it possible to simulate, in a fast, relatively precise manner, but nevertheless with relatively modest on-board calculation means, the virtual behavior of the load. response to said command, and thus to determine a control that reconciles a good stability of the load, minimizing the ballad, with a relatively fast convergence of the filtered speed setpoint to the speed setpoint expressed by the crane operator. In this respect, it will be noted that the purpose of the filter of the invention is to minimize the ballast, by using a virtual predictive model of the behavior of the pendular system constituted by the hoist with its suspended load. In practice, the use of such a virtual model makes it possible to theoretically define driving conditions, in this case speed conditions, and more particularly conditions of speed evolution over time, and consequently trajectory conditions ( of the carriage), which are such that, when said speed conditions are applied to the actual system, the swing of said real system is damped, i.e. remains below a predetermined amplitude threshold. Thus, for any "raw" control setpoint specified by the crane operator, the filtering will instantly transform said "raw" control setpoint into a filtered control setpoint, whose filter will have adapted, on the basis of the virtual model, the evolution in time, so that, on the one hand, said filtered setpoint can evolve to reach, in a response time which is as short as possible (with regard to the selected typing coefficient), the value of the setpoint "brute" wanted by the crane operator, that is to say more particularly that the speed of the carriage can converge to the speed desired by the crane operator, and so, on the other hand, this convergent evolution of the setpoint filtered to the gross setpoint is nevertheless done without creating the conditions for the appearance of a ballant.
    [0006] In other words, the control setpoint, and more particularly the speed setpoint, filtered by the method according to the invention will advantageously correspond exactly to the steering setpoint, and more particularly the speed, which must be applied to the system. real to converge to the gross steering setpoint desired by the crane operator while simultaneously damping the ballant. Once the parameters of the virtual control determined to satisfy, according to the compromise sought, the dual criterion of speed and stability, that is to say once placed the poles of the transfer function of the virtual model in a closed loop, it is possible to apply said virtual model to return by state control as a dynamic filter, so as to filter the raw control setpoint (gross speed setpoint) actually expressed by the crane operator, then to apply the control setpoint ( speed set) thus filtered to the effectors of the hoist (typically electric motors controlled by drives).
    [0007] Advantageously, the application of said filtered control setpoint to the real system can itself be performed in an open loop, without it being necessary for example to obtain feedback on the actual angle of the ball. This will lighten the structure of the hoist and simplify the servo algorithm.
    [0008] Finally, it is remarkable that the method according to the invention allows acquisition and filtering in real time of the control setpoint, which gives it a great responsiveness by making it able to immediately adapt the filtered control set to fluctuations unpredictability of the gross steering instruction expressed by the crane operator, especially when the crane operator seeks to avoid an unforeseen obstacle, to manually compensate a dangling, or to counterbalance the effects of a wind squall. In other words, the control obtained by the method according to the invention advantageously allows the hoist to continuously monitor, and without ever creating unwanted ballad, the instruction expressed by the crane operator, even if it- it is likely to change at any time.
    [0009] Said control method is therefore particularly versatile, in that it is suitable for a wide variety of driving scenarios, that it does not depend on the path taken by the load or even the existence or not of crossing points. imposed, and that it reacts reliably and faithfully to the multiple and random changes of gross pilot setpoint, and more specifically to changes in speed setpoint, imposed by the crane operator. Other objects, features and advantages of the invention will appear in more detail on reading the description which follows, as well as with the accompanying drawings, provided for purely illustrative and non-limiting purposes, among which: FIG. 1 illustrates, in a schematic view, the parameterization of a model of the pendular system formed by the load suspended on the hoist. FIG. 2 illustrates, in the form of a block diagram, a dynamic filter according to the invention based on a virtual model associated with a feedback control. FIG. 3 illustrates, in the complex plane, the principle of placing the poles of a filter according to the invention, by setting the typing parameter. FIGS. 4 and 5 illustrate, in time marks, respectively for a dispensing movement and an orientation movement, an example of filtered setpoint obtained in response to a step-type speed command (raw control setpoint), for a first typing parameter value corresponding to a "reactive" mode operating setting. FIGS. 6 and 7 illustrate, in temporal references, and respectively for a dispensing movement and an orientation movement, an example of filtered setpoint obtained in response to a step-type speed command (raw control setpoint), for a 30 second typing parameter value lower than that of Figures 4 and 5 and corresponding to a functioning setting in "damped" mode. The present invention relates to a method of controlling the movement of a load 1 suspended at a point of attachment H of a hoist 2, said point of attachment H being designed to be able to move in a rotational movement R in lace around a first vertical axis (ZZ '), called "axis of orientation", and / or in a translational movement T along a second axis (XX') said "distribution axis" intersecting audit orientation axis (ZZ '), as shown in Figure 1. The point of attachment H is preferably formed by a carriage guided in translation along the distribution axis (XX'), said carriage can be 5 driven by any suitable motorized device. The load 1 is suspended at the point of attachment H by a suspension device 3, such as a suspension cable, and which will be likened to such a cable in what follows. Preferably, it is possible to vary the length L of the suspension cable 3, so as to be able to modify the distance of the suspended load 1 at the point of attachment H, and thus to vary the lifting height of the suspended load 1 by relative to the ground (by raising the load 1 by shortening the length L, or lowering said load 1 by an elongation of said length L) according to a third movement called "lifting". The hoist 2 may in particular form a tower crane, whose mast 4 materializes the axis of orientation (ZZ '), and whose arrow 5 materializes the distribution axis (XX'). For simplicity of description, such a tower crane configuration will be considered in the following, it being understood that it is perfectly conceivable to apply the principle of the invention to other lifting gear, and particularly to mobile cranes or luffing jib cranes, simply adapting the model accordingly. Note the intersection of the distribution axis (XX ') and the axis of orientation (ZZ'). Preferably, the distribution axis (XX '), which passes from the rest also by the point of attachment H, will be substantially horizontal, and considered as such in the following, for convenience of description. For convenience of description, and with reference in particular to FIG. 1, the following designations and conventions will also be adopted: m [kg] is the mass of the suspended load 1, M [kg] is the mass of the point of attachment H, and more particularly of the carriage forming said point of attachment, 1-y [kg.m2] is the moment of inertia of the hoist 2 carrying the point of attachment H with respect to the axis d 'orientation (ZZ'), y [rad] is the angular position of the point of attachment H around the axis of orientation (ZZ '), x [m] is the distance from the point of attachment H to the axis of orientation (ZZ '), preferably equal to the length of the segment [OH], L [m] is the length of the suspension cable 3 connecting the suspended load 1 to the point of attachment H, (1) [rad ] is the radial component of the swinging angle of the suspended load 1 in the vertical plane containing the distribution axis (XX), 0 [rad] is the orthoradial component of the swinging angle of the suspended load 1 in the plan vertical tangential to the rotational displacement of the point of attachment H, that is to say in the plane which is perpendicular to the preceding, and preferably normal to the distribution axis (XX '), Ty [Nm] is the motor torque applied to drive the hooking point H in rotation around the axis of orientation (ZZ '), Fx [N] is the force exerted on the point of attachment H to move the latter in translation according to the distribution axis (XX '), F_ = x MMTMF = - M = -, mr I y = Ill y I' r Ir r Note that the projection reference used to define the "radial" and "orthoradial" components of the dangling (pendulum movement of the load), and / or those of the movement of the point of attachment H, here advantageously corresponds to the Frenet mark attached to said point of attachment H, and whose normal vector (or "radial") is preferably worn, at any moment, by the distribution axis (XX '). In a manner known per se, the displacement of the suspended load 1 can advantageously be controlled by a control instruction which relates to a quantity, called "servo quantity", which is characteristic of the movement of said load 1, and which preferably constitutes a kinematic magnitude (speed or acceleration type). This controlled quantity may, for example, be the speed of rotation V 1 and / or respectively the translation speed V d of the point of attachment H. According to the invention, the control method comprises a step (a) of acquisition. setpoint, during which it acquires, in real time, a control setpoint value called "gross steering setpoint" Vi, which corresponds to the value of the controlled variable that the driver of the hoist 2 to the moment t considered. Gross control setpoint V; thus reflects the behavior (typically, the speed of movement) that said crane operator, at the moment considered, want to confer the point of attachment (carriage). More particularly, said method may comprise a step (a) of acquisition of setpoint, during which one acquires, in real time, a speed reference value V; (i.e., a pilot set point value expressed as a speed reference value Vi) which is representative of the rotation speed V ,, and / or respectively of the translation speed Vd, that the driver of the hoist 2 wishes to confer the point of attachment at time t considered. In the following, it will preferably be considered, for convenience of description, that the control is performed on the speed of the point of attachment H (carriage), and therefore "driving instruction" and " speed instruction ", without this however constituting a limitation of the invention. By convention, the index i will be assigned the value "o" to refer to the orientation movement (rotation R) and the value "d" to refer to the distribution movement (translation T). Advantageously, the method according to the invention constitutes an iterative method, which firstly allows a substantially real-time monitoring of the pilot control value, and more particularly of the value of the speed reference, which is by nature fluctuating and unpredictable, which is fixed at each moment by the driver of the machine, and then a permanent refresh of the calculations and, consequently, of the corresponding filtered setpoint, and this independently of the total duration of the journey necessary for the transport of the load suspended 1 from its starting point to its point of arrival.
    [0010] Preferably, the method will have for this purpose a relatively short sampling period, significantly less than the total journey time. Said sampling period will thus preferably be less than 100 ms, and for example of the order of 40 ms. The speed setpoint V; may of course be fixed by the driver of the hoist 2 by means of any appropriate control member 6, such as a joystick, which may preferably simultaneously define the speed reference in translation Vd and the speed reference in rotation V, 10 that the driver wishes to confer on the point of attachment H. The control method also comprises, after the step (a) of acquisition of setpoint, a step (b) of filtering setpoint. During this step (b) of setpoint filtering, as is illustrated in particular in FIG. 2, the pendulum behavior of the suspended load 1 is modeled according to the movement R, T considered by a virtual model ii7 = AX + BU which uses a state vector X. According to the invention, this state vector X comprises at least one component called "principal component", which corresponds to the controlled variable, and other components, called "additional components" x3, 20 x4, representative of kinematic variables characteristic of the swinging movement of the suspended load, such as, for example, the swinging angle, 0 or the angular rate of swing 0, 0 according to the movement R, T considered. More particularly, said state vector X preferably comprises at least one instantaneous speed component x1 representative of the instantaneous speed of the point of attachment H according to the movement under consideration (here the rotational speed component around the axis d the orientation (ZZ '). ay Y = at according to the rotational movement R, or respectively the linear distribution component linear axis = according to the translational movement T), an angular angle component x4 representative of the angle of the suspended load according to the movement considered (here 0 the orthoradial swing component for the rotational movement R, and 0 the radial component of the balloon for the translational movement T), and an angular velocity component x 3 representative ballant the angular velocity of dangling ae according to the movement considered (here the orthoradial component = - for at ao rotational movement, and respectively the radial component o = - for at the translational movement). In the expression of the above model, "A" represents the state matrix (system evolution matrix), "X" the state vector, "B" the command application matrix, "U" the vector of the commands (inputs). The vector "X" corresponds to the first derivative with respect to the time of the state vector X. In FIG. 2, the letter "p" corresponds to the complex variable used by the Laplace transforms; Thus, formally: X = pX Preferably, the control application matrix B will be a column vector, and the control vector U will be reduced to a matrix of dimension 1x1. In practice, the virtual model ii7 = AX + BU advantageously corresponds to the matrix expression of a system of equations derived from Newtonian mechanics and making it possible to describe here in projection in the radial vertical plane containing the distribution axis (XX ') for the translation movement T, and / or in projection in the orthoradial vertical plane for the rotational movement R, the behavior, and more particularly the motion components, of a virtual pendulum system which has the characteristics of the load 1 suspended at the point of attachment H. To simplify this modeling, we will preferably make the assumption of "small angles", considering as a first approximation that the amplitude of the ballant, and therefore the angle components of dangling, are relatively weak, which makes it possible in particular to simplify the trigonometric expressions by limited first-order developments. During the filtering step (b), the virtual model ii7 = AX + BU is advantageously applied to a feedback control 1J; = kio.Vi-Ki.X which involves, in addition to a term of reference kio.Vi representative of the instruction 30 of raw control (and more particularly of the speed reference) V, acquired during the acquisition step, a corrective term KX corresponding to the product of a corrector vector K, (here a vector-line) by the state vector X (here a vector-column). The use of a (virtual) control by state feedback, which in this case involves proportional returns, of the "gains" type, ki2, advantageously provides a simulation of the theoretical behavior of the suspended mass 1 in response. to said command, according to the equivalent of a closed-loop (virtual) system whose dynamics can be studied and chosen, and in particular the stability and the responsiveness (response time), by proceeding, by an appropriate choice of said gains, to a pole placement of the transfer function which corresponds to the evolution matrix AB.K of the closed-loop system. For the definition of the state-return command, a correction vector K is advantageously used, in which at least some of the correcting gains, among the correcting gains km, ki2 associated with the additional components x3, x4, are expressed as a function of a same predetermined typing parameter Tc, said typing parameter Tc being freely adjustable by the driver of the lifting machine so as to be able to adjust said correcting gains km, More preferentially, using a corrector vector K, in which the correcting gains km, ki2 associated respectively with the angular velocity component x3 and the balloon angle component x4 of the state vector X are expressed as a function of the same predetermined typing parameter Tc, said typing parameter Tc being freely adjustable by the driver of the lifting machine 2 so as to be able to adjust said gains km correctors, Advantageously, formally making If a parameter (coefficient) of typing Tc is used in the definition of the gains of the km corrector vector, which are associated with the balloon angular velocity component x3 and the dangling angle component x4, we make the return of state, and consequently the behavior of the closed-loop system, and more particularly the placement of the poles of the evolution matrix AB.K, of the choice of said typing parameter Tc. In other words, according to a characteristic which may constitute an invention as such, the method according to the invention will make it possible to filter a speed setpoint V by means of a feedback control applied to a virtual model ii7 = AX + BU, by providing a typing parameter Tc, freely adjustable by the driver of the machine, which makes it possible to arbitrarily modify the placement of the poles of the corresponding evolution matrix AB.K, and consequently the dynamics of the filtered system.
    [0011] The driver of the hoist can thus, during a setting step, which may precede the maneuver of the suspended load 1 or intervene during said maneuver, vary at will the typing parameter Tc,. Thus, the driver of the machine can, by means of a single adjustment, quick and easy to implement, modify and adapt the degree of responsiveness and stability of the assistance to the maneuver of the suspended load 1 that he provides the filtering, according to the method according to the invention, its speed instructions V.
    [0012] More particularly, the driver can opt according to his choice either for a mode of assistance "damped", presenting graphically (see Figure 3) poles relatively far from the oscillating mode of the imaginary axis, and according to which the filtered speed reference converge towards the speed setpoint V; relatively slowly, but in a particularly stable manner, without exceeding, as is illustrated in FIGS. 6 and 7, or, if he feels the damped mode being too "soft", for a "reactive" mode of assistance, according to which the The poles are closer to the imaginary axis and the damping coefficient is lower than in the damped mode, so that the filtered speed reference converges more rapidly towards the setpoint V; (for example with a response time of 5% which is lower than that of the damped mode, as is the case in FIGS. 4 and 5 with respect to FIGS. 6 and 7 respectively), but possibly tolerating a slight overshoot and / or some damped oscillations of the filtered speed reference (see Figure 5, for example).
    [0013] Of course, it is conceivable to provide more than two or three adjustment values, and in particular a continuous range of settings of the typing parameter Tc, preferably, with the increasing values of Tc, of a first damped assistance mode. a second mode of assistance (more) reactive, through different modes of assistance intermediate.
    [0014] It will be noted that the choice (setting) of the typing parameter Tc is advantageously free, that is to say arbitrary, in that the fixing of said typing parameter Tc depends on the (only) will of the driver of the machine, and that, in fact, said typing parameter Tc, which remains preferably constant between two successive modifications made by the driver of the machine, has a proper influence on the setting of the correct gains km, kj2, independently of the configuration of the lifting gear 2 or speed reference values V; applied by the driver of the machine. In particular, the fixing of said typing parameter Tc will advantageously be dissociated from the speed reference value V.
    [0015] Said typing parameter Tc will also be distinct and independent of the mass m of the suspended load 1 or that M of the point of attachment H, of the moment of inertia of the machine 2 around the axis of orientation (ZZ '), and the length L of the suspension cable 3. Of course, the definition (or selection) of the value of the typing parameter Tc 10 by the driver of the hoist 2 can be made by any selector, potentiometer or appropriate electronic or electromechanical programmer. According to one possibility, the method may be implemented by a computer having a non-volatile memory, preferably reprogrammable by the operator of the machine or by a maintenance technician, intended to store several predefined settings of the typing parameter. Tc, for example associated with different drivers and / or working conditions (including different weather conditions). It should also be noted that the invention preferably allows the operator of the machine to modify and adjust the typing parameter Tc as often as necessary, if necessary several times, before and / or during the maneuver of the machine. suspended load 1. This availability and this accessibility, if necessary in real time, of the setting of the typing parameter Tc gives a great versatility and a great flexibility of use to the hoist 2 thus equipped. Finally, during the filtering step (b), the virtual model is extracted from a control setpoint, called "filtered control setpoint Y". , z, which corresponds to the main component x1 of the state vector X, that is to say to the controlled variable. More particularly, the virtual model is extracted from a filtered speed setpoint Y 1. , z, which corresponds to the instantaneous speed component x1 of the state vector X. For the sake of convenience of description, it will be understood in what follows, without this constituting a limitation of the invention, the filtered control setpoint Y at a filtered speed setting) , z.
    [0016] It is this filtered set value Y, "), z (here preferably corresponding to the single coefficient x1 of the vector of the outputs Y), represented by the curves in solid lines in FIGS. 4 to 7, which is applies to the drive means (not shown) of the point of attachment H, designed to drive said point of attachment according to the movement R, T. Typically, the filtered control setpoint may correspond to the translation speed setpoint z applied to a first frequency converter driving a first electric motor designed to translate the carriage traveling on the arrow 5 in translation, respectively to the rotation speed instruction "). applied to a second frequency converter driving a second electric motor designed to drive the boom 5 in rotation around the mast 4. Preferably, the method uses, to model the pendular behavior of the suspended load 1, a dynamic (virtual) model which involves the mass M of the point of attachment H (and more particularly that of the corresponding carriage) and the mass m of the suspended load 1. Advantageously, such a model makes it possible to describe, in an approximate but relatively faithful and precise manner, the pendulum behavior of the load by means of relatively simple equations allowing rapid calculations and not very greedy in material and energy resources. The mass M of the point of attachment H may advantageously be provided by the manufacturer of the hoisting machine 2. The mass m of the suspended load 1 may be measured or estimated by any appropriate means, and for example by a measurement of the torque that must be provided by the hoist motor to be able to move said load vertically. If necessary, a fixed average value representative of the mass of a suspended load 30 1 "average" may be considered, as a first approximation, or on the contrary measure said mass m on a case-by-case basis, at each loading, in order to adapt the model of the filter as finely as possible to this parameter. In a particularly preferred manner, the dynamic model is defined as follows, by a model called "without coupling" expressed by: = AX + BU Y = CX where "Y" represents the vector of the outputs, and "C" the observation matrix , and X = xi A = (0 0 0 a 1 B = b1 C = 1 0 0 0 0 0 0 0 0 0 1 0 0 0 0 1 X2 1 0 0 0 0 X3 _x4 0 0 0 a2 b2 0 0 1 0) 0 with: in distribution orientation in, gx ay al = -mtg a. = g + gmt x = i = 1. a1 x1 = y = at X2 = y at 1+ M rx2 Ô X2 = X g 1+ M rx2 ± MrX2 = Ô = - ao = a2 L b1 = 1 b = -1 2 L a2 L 2 L 1+ M X2 Y 1 b x3 = 0 = at .x4 = 0 U = F, x3 at x4 = e U = T y 1- 1+ mr x2 1 xb = 2 L 1+ MrX2 As mentioned above, the equations will be simplified here considering the hypothesis of small dangling angles. Advantageously, it should be noted that here the general expression of the model remains the same for each of the movements, whether said model is applied to the command in orientation (rotation R) or in distribution (translation T), only the coefficients of the matrix of state A and of the state vector X being adapted according to the movement to which said model applies. It will also be noted that it is perfectly conceivable to apply the method to the control of a single movement (for example the distribution only), especially if the hoist 2 is a loading gantry, of the traveling crane type, comprising a linear displacement movement in translation of the suspended load, and no rotational movement.
    [0017] Conversely, it is advantageous to simultaneously control both the distribution movement (translation T) and the orientation movement (rotation R), when the lifting machine offers these two movements, and this with the aid of the same model, and therefore with a relatively reduced onboard computing power. As such, it will be noted that the typing parameter Tc may advantageously be the same, that is to say present an identical value, for the application of the model in distribution and for the application of the model in orientation, which further simplifies filter adjustment for the crane operator.
    [0018] Moreover, it will be noted that, in the equations of the above model, it has been chosen, as a first approximation, not to take into account the coupling phenomena between the axes, that is to say the centrifugal acceleration. This being so, it would be perfectly conceivable to add a coupling term involving centrifugal acceleration, while retaining the typing principle proper to the invention. Thus, it is possible, in an alternative manner to the "without coupling" model above, to use mutatis mutandis a so-called "coupled" model, expressed by: 20 = AX + Biui + B2u2 Y = CX where "Y" represents the vector of the outputs, and "C" the observation matrix, and b3 b1 0 X 'X2 X3 X4 X = B2 = B1 =, C =, A = b2 0 b4 0 with: (0 0 0 al 1 0 0 0 0 0 0 0 a2 0 0 1 0 1 0 0 0 0 0 0 0 0 0 1 0 0 0 0 1 30 in distribution orientation mr gx xi =). where x = x x2 = x x2 = y + gm, X3 = x3, e __ X4 = X4 = 0 ul = Fx u2fy 111 = iÿ u2 = Fx al = 1+ M rx2 g 1+ M rx2 + mrx2 = a2 = L b1 = 1 b2 = 1 L b3 = 0 b4 = 0 a2 L 1 + M X2 r 1 // I = 1 ± MrX2 1 x - LO b2 = L 1 + M rx2 b3 = 0 b4 = 0 ( 1 + Mrx2) Advantageously, whether one considers the model with coupling or the model without coupling, one applies to the aforementioned model control 5 by state feedback: 1.1; = kio.Vi - Ki X with, for vector (multiplier) corrector, Ki = [ki0,0, km, ki2], and with, by convention i = d for distribution regulation, and i = o for regulation in orientation. Here again, whether we consider the distribution (translation) or orientation (rotation) movement, we can use mutatis mutandis the same type of virtual model and the same control principle by state feedback, with placement of poles. Preferably, for simplicity of calculation, it will be considered, as is illustrated in FIG. 2, that the gain ki0 is the same for the speed reference V; and for the state return (in the correction vector K). In a particularly preferred manner, also for simplicity of calculation, it will be considered that ki0 = 1, that is to say that the control has a unit gain state feedback for the instantaneous speed component. Finally, the dynamic filter applied to the speed setpoint V; during the step (a) of filtering, and corresponding to the control law illustrated in Figure 2, can be written, for the model without coupling: = (A- BK) X + BV, Y = C'tX with C't = 0 0 0] For the model with coupling, one will have the same: = -131KW + + B2 Ty Y = C'tX with C't ooo] The matrix A-BK (respectively A-B1K) form here the matrix of evolution of the virtual model in closed loop. As such, it will be noted that, advantageously, the control application matrices "B" of the model without coupling and B1 of the model with coupling being identical (B = B1), the evolution matrix A-BK is the same, whether we consider the model with coupling or the model without coupling. The method of resolution (and filter definition) according to the invention can therefore advantageously be applied equally to both models. In the case of the dynamic model proposed above, said evolution matrix of the closed-loop virtual model can be written as: (0 0 0 a1 "b1 -b1 0 -blka blki2 1 0 0 0 1 0 0 0 A-BK = 0 0 0 a2 b2 [1 0 ka ki2] = - b2 0 -b2ka a2-b2ki2 0 0 1 0 0 0 0 0 1 0 20 Preferably, during the set-point filtering step, the correcting gains km, ki2 are expressed as a function of the typing parameter Tc so that one can extract from the evolution matrix A-BK of the closed-loop virtual model a sub-matrix Ar of dimension 2 x 2 which, on the one hand, connects said correcting gains km, ki2 to two additional components of the state vector, which components preferably correspond respectively to the angular velocity component x3 and the balloon angle component x4, and on the other hand the eigenvalues λ have a non-zero real part entirely determined by said typing parameter Tc.
    [0019] With reference to the above-mentioned model, the sub-matrix Ar = [(A-BK) i, j] with i = (3, 4) and j = (3, 4) can be extracted from the closed-loop evolution matrix. ), which is in practice sufficient to describe the dynamics of the system: A = Lb21 (ii a2 b2k / 21 r 1 The inventors have indeed found that, according to a characteristic which may be a whole invention in itself, and this, whether or not the correcting gains are expressed as a function of the same typing parameter Tc, it is possible to operate, as a first approximation, a setting of the dynamics of the closed-loop feedback control model, a model which possesses for matrix (complete) evolution A-BK, from a placement of the poles of a subarray Ar of said evolution matrix A-BK, said subarray Ar representing indeed a subsystem enough to describe, in an approximate manner, the dynamics of the complete system. The purpose of this invention is to simplify and accelerate the computation of the eigenvalues, and consequently the implementation of the filtering of the speed reference. In the present case, said subsystem is a non-deterministic system which relates the correcting gains kir, ki2 to the two components of the state vector X respectively corresponding to the angular velocity component 20 of swaying x3 and to the component belly angle x4. The validity of the approximation may depend in particular on the fact that the above-mentioned corrective gains kir, ki2 have, in the complete system corresponding to the evolution matrix A-BK, a relatively small or negligible influence in determining the component representative of the instantaneous speed, in particular in comparison with the influence of the coefficients of the state matrix A (and more particularly of the coefficient a1, in the preceding example, which depends in turn on the structural characteristics of the machine 2 and the mass m of the suspended load 1). In practice, this approximation can in particular be verified as long as the typing parameter Tc remains below a predetermined threshold. In a particularly preferred manner, the eigenvalues de of the sub-matrix Ar have a non-zero real part Re (λ) directly proportional, or even equal, to said typing parameter Tc or to the inverse 1 / Tc of the typing parameter.
    [0020] Preferably, the real part Re (λ) of said eigenvalues will even be determined exclusively by the typing parameter Tc, and for example exactly equal to the opposite of the inverse -1 / Tc of the typing parameter, as this is illustrated in Figure 3.
    [0021] It will thus be possible to operate very simply an immediate positioning of the poles, by only adjusting the value of the typing parameter Tc, which is sufficient to characterize said real part Re (A). More particularly, an increase in the typing parameter Tc will bring the poles of the imaginary axis closer together (by reducing the absolute value of their real part) and thus confer on the filtering a more reactive character. Conversely, a decrease in the typing parameter Tc will distance the poles from the imaginary axis, by increasing the absolute value of their real part, and thus confer on the filtering a less reactive, more depreciated character. Furthermore, the sub-matrix Ar and the expression of the gain gains km, ki2 will preferably be chosen such that the imaginary part of said eigenvalues Im (λ) is independent of the typing parameter Tc. Thus, for a given hardware configuration of the hoist, and more particularly for an imaginary part Im (A) given constant, the dynamics of the system may advantageously be entirely defined by the choice of the value of the typing parameter Tc. In practice, this imaginary part Im (A) may depend on the material configuration of the hoist, and in particular, in the aforementioned dynamic model, the length L of the suspension cable 3, the distance x to the axis of rotation of the point of attachment H, and ratios between the mass M of the point of attachment H, the mass m of the suspended load 1, and the moment of inertia 1-y of the hoist. Preferably, the choice of correcting gains (km, ki2) is as follows: 2 T 2b2 1 = where i = d for the control in distribution, and i = o for the command in orientation. The sub-matrix Ar can then be written: 2 1 T a2 - T 1 0 so that its eigenvalues are: = - 1 - ± i - V- a2 Tc Thus we find, as indicated above, and such that is illustrated in FIG. 3, a placement of the poles (that is to say eigenvalues represented by crosses in FIG. 3) whose dynamics, and more particularly the damping, is determined exclusively by the choice of typing parameter Tc. Indeed, the real part Re (A) of said poles (eigenvalues) is equal to 1 / Tc, while the imaginary part Im (A) of said poles (eigenvalues), which is here .iC / 2, is advantageously independent of typing parameter.
    [0022] Advantageously, said imaginary part is here constant for a given hardware configuration of the hoist, if one considers a constant value of the length L of the suspension cable, in the case of the control according to the translation movement T, or if we consider constant values of the length L of the suspension cable and the distance x of the point of attachment H to the axis of rotation, in the case of a control according to the rotational movement R. By Moreover, it will be noted that the method, and more particularly the proposed models, advantageously make it possible to take into consideration the length L of the suspension cable 3 which connects the suspended load 1 to the point of attachment H. In particular, the correcting gains km , ki2 are preferably also expressed as a function of the length L of the suspension cable 3 which connects the suspended load 1 to the point of attachment H. In this way the correcting gains km, ki2 of the return control 30 of state, and therefore the filtering of the setpoint speed, may vary in real time according to variations of the length L of the suspension cable 3.
    [0023] In the above example, the length L is used in the calculation of the gains by the coefficient b2. It is thus possible to continuously adjust the virtual model to the actual instantaneous configuration of the hoist 2, and consequently to obtain a model that faithfully and accurately reflects the behavior of said hoist, and the suspended load 1, to any time considered, regardless of the lifting height of said load. In this way, it increases the accuracy and reliability of the control by filtering the speed reference.
    [0024] The estimation of the length L of the suspension cable may, for example, result from the counting of revolutions made by a winding / unwinding winding of the suspension cable 3, driven by the hoisting motor. Moreover, it will be noted that the voluntary manual adjustment of the filtering dynamics, by the choice of the value of the typing parameter Tc, is advantageously distinct and de-correlated from the automatic adjustment which consists of adapting the model, and more particularly the one or other of the coefficients of the state matrix A, the instantaneous hardware configuration of the hoist 2, and more particularly the length L of the suspension cable 3.
    [0025] In other words, the typing parameter Tc thus makes it possible to set the desired type of dynamic behavior independently of the material configuration of the machine, and, more particularly, to choose freely, with material configuration (masses, inertia, height of lifting) given, the dynamics of the system among a plurality of modes available.
    [0026] Moreover, it will be noted that the correcting gains km, ki2, as well as the poles characterizing the dynamics of the modeled system, are advantageously defined in an explicit and deterministic manner by formulas which depend exclusively on the typing parameter Tc on the one hand, and input data related to the hardware configuration of the system on the other hand (mass m of the suspended load, mass M of the truck, length L of the suspension cable, inertia module 1-y of the machine, distance x to the axis of rotation of the point of attachment). In this way, said correcting gains km, ki2, as well as the location of the poles, can be calculated directly and instantaneously from these data, without it being necessary to pre-establish, store in memory and then interrogate. periodically mappings (charts or databases) which would associate to the various foreseeable life situations of the hoist, for example in the form of scatter plots, different corrective gain values adapted to each loading situation and / or to each spatial configuration of the hoist.
    [0027] Here again, the method used makes it possible to reduce the data storage capacities necessary for the control of the hoist. As an indication, the typing parameter Tc may be chosen in a range between 0.2 and 2, preferably between 0.3 (FIGS. 6 and 7), and 1.8 (FIGS. 4 and 5). In this case, the low values of Tc correspond to a "damped" assistance mode, relatively slow and stable, and the high values of Tc to a "reactive" assistance mode, faster and a little less stable than the damped mode. . Preferably, the method comprises an open-loop control step which consists in applying the filtered control setpoint (the filtered speed setpoint) Y, that is to say x1 =) for the rotation R, and x1 = for the translation T, to a control system of the lifting gear 2 in an open loop, that is to say to a control system which applies said control setpoint (speed ) filtered Y,) , z to driving means 20 of the point of attachment H, designed to drive the point of attachment H according to the movement R, T considered, without using a measured or calculated return of the angle, or the angular velocity , actual ballooning of the actual suspended load, and, preferably, measured or calculated return of the effective speed of the actual hook point displacement. In other words, the invention makes it possible to define virtually, during the filtering, a speed setpoint x1 which enables the virtual system to satisfy the speed and stability criteria determined by the choice of the typing coefficient Tc, then transpose this velocity setpoint resulting from a virtual modeling, to the actual hoist 2, as an effective setpoint filtered speed, and this in open loop, that is to say blind, without slaving aimed at to fight then the real dangling which would result possibly from the application of this filtered instruction or else which would result from external disturbances. As such, it will be noted that any actual dangling, which would appear as a result of the application of the filtered setpoint to the actual lifting apparatus, would in any case be intrinsically reduced, by the very fact that said filtered setpoint is precisely elaborated so as to minimize or even prevent the appearance of such a ballant. In any case, the invention thus advantageously makes it possible to simplify the structure of the machine 2, since it is in particular not necessary to provide sensors (or corresponding wiring) dedicated to the measurement and control. monitoring the actual values of dangling. In addition, the quantity of measurements and information processing operations to be performed is limited, which makes it possible to reduce the calculations, and consequently to reduce the dimensions and the energy consumption of the on-board electronic control device. 2. The reliability of such an open-loop application assumes, of course, that the actual system of the load suspended from the hoist 2 has a behavior similar to that modeled, which is the case here. The invention also relates as such to a computer, or to a computer readable data medium, which receives or contains computer program code elements for implementing a servo control method according to the invention. when said code elements are read by said computer. The invention finally relates as such to a hoist 2, such as a tower crane, comprising a mast 4 extending substantially along a first vertical axis, said axis of orientation (ZZ '), an arrow Secant auditing mast extending along a second axis, said distribution axis (XX '), and which carries a point of attachment H to which a load 1 can be suspended, and drive members for moving said point of attachment H in rotation R about the axis of orientation (ZZ ') and in translation T along the distribution axis (XX'). According to the invention, the driving member or members of said lifting gear 2, associated with at least one of said two rotational movements R and of translation T, and preferably at each of these two movements, are controlled by a control system provided with calculation and programming means designed to filter in real time, by a method according to the invention, a raw control setpoint (speed reference) V; defined by the driver of the hoist, and to apply the filtered control setpoint (filtered speed setpoint) Y,) resultant to the corresponding drive member 35.
    [0028] Of course, the invention is not limited to the only embodiments described, the person skilled in the art being able to isolate or combine freely between them one or other of the characteristics mentioned in the foregoing. , or to substitute equivalents for them. In particular, it would be perfectly possible to use any other model and / or any other appropriate spatial coordinate system, even if it means adapting the number of components of the state vector X and / or the formulation of the correcting gains as a function of the typing parameter Tc.
    [0029] Advantageously, the invention makes it possible to reliably and relatively simply limit the swing of a suspended load 1 during the displacement of the latter, according to one or two movements R, T, isolated or combined, while conferring on the apparatus of Lifting an adjustable, safe, and predictable behavior, which offers a loyal and comfortable driver feeling to the crane operator, according to the custom Tc typing defined by the latter. Advantageously, the simplicity of the method and of the model used makes it possible to obtain a sufficiently precise approximation of the pendular behavior of the suspended load in order firstly to appreciably attenuate the dangling and, on the other hand, to maintain the desired reactivity vis-a-vis velocity setpoint fluctuations, without having to be burdened with very heavy calculations based on complex algorithms that aim to obtain an optimized speed and / or trajectory (s).
    权利要求:
    Claims (11)
    [0001]
    REVENDICATIONS1. A method of controlling the movement of a load (1) suspended at a point of attachment (H) of a hoist (2), said point of attachment (H) being designed to be able to move in a movement of 5 rotation (R) in lace around a first vertical axis (ZZ '), called "axis of orientation", and / or in a translation movement (T) along a second axis (XX') said "Distribution axis" intersecting said orientation axis (ZZ '), the movement of the suspended load (1) being controlled by a control setpoint which relates to a characteristic quantity of the movement of said load (1), referred to as " controlled variable ", such as the rotational speed (V0), and / or respectively the translation speed (Vd) of the point of attachment (H), said method comprising: a step of (a) acquisition of setpoint, at during which a pilot control value known as "gross piloting setpoint" (V;) is acquired in real time, which corresponds to the value of the large slave which is aimed at the driver of the lifting gear (2) at the instant (t), then a step (b) of setpoint filtering during which: the pendulum behavior of the suspended load 20 is modeled ( 1) according to the movement (R, T) considered by a virtual model (1 (= AX + BU) which uses a state vector (X) comprising at least one component called "principal component" (xi), which corresponds to the controlled size, and other components, referred to as "additional components" (x3, x4), representative of characteristic kinematic magnitudes of the swinging motion of the suspended load, such as the swinging angle (,) or the speed. . angular dangling (0,) according to the movement considered, is applied to said virtual model (I (= AX + BU) control 30 by state feedback (1J; = kio.Vi-Ki.X) which involves, in addition to a setpoint term (kio.Vi) representative of the raw control setpoint acquired during the acquisition step, a corrective term (Ki.X) corresponding to the product of a correction vector (K) by the vector of state (X), using a correction vector (K) in which at least some of the gain gains (km, ki2) associated with the additional components are expressed according to the same predetermined typing parameter (Tc), said typing parameter (Tc) being freely adjustable by the driver of the lifting machine so as to be able to adjust said correcting gains (kii, ki2), and the virtual model is extracted from a control setpoint, called "filtered control setpoint" (Y,) , Z), which corresponds to the main component (xi) of the state vector (X).
    [0002]
    2. Method according to claim 1 characterized in that during the step (a) of target acquisition, is acquired in real time, a speed reference value (Vi) representative of the speed of rotation ( V0), and / or respectively of the translational speed (Vd), which the driver of the lifting machine (2) wishes to give to the point of attachment (H) at the instant (t) in question, then, at the during the step (b) of setpoint filtering: - the pendulum behavior of the suspended load (1) is modeled according to the movement (R, T) considered by a virtual model (I (= AX + BU) which uses a state vector (X) comprising at least one instantaneous speed component (xi) representative of the instantaneous speed (Y) of the point of attachment (H) according to the movement considered, a representative balloon angle component (x4) the swinging angle (,) of the suspended load (1) according to the movement considered, and a component of angular velocity of ballant (x3) representative of the angular velocity of dangling (0,) according to the movement considered, - one applies to said virtual model (I (= AX + BU) a control by state feedback (1J; = kio.Vi-Ki.X) which involves, in addition to a setpoint term (kio.Vi) representative of the speed reference (Vi) acquired during the acquisition step, a corrective term (Ki.X) corresponding to the product of a corrector vector (K) by the state vector (X), using a corrector vector (K) in which the correcting gains (km, ki2) associated respectively with the angular velocity component of dangling ( x3) and the swing angle component (x4) of the state vector (X) are expressed as a function of the same predetermined typing parameter (Tc), said typing parameter (Tc) being freely adjustable by the driver of the lifting apparatus so as to be able to adjust said correcting gains (km, ki2), and the virtual model is extracted a filtered speed setpoint (Y,) ,), which corresponds to the instantaneous speed component (xi) of the state vector (X).
    [0003]
    3. Method according to claim 1 or 2, characterized in that, during the setpoint filtering step, the correcting gains (km, ki2) are expressed as a function of the typing parameter (Tc) so that one can extract from the evolution matrix (A-BK) of the virtual model in a closed loop a sub-matrix (Ar) of dimension 2 x 2 which on the one hand connects said correcting gains (km, ki2) to two additional components of the state vector, components which preferably correspond respectively to the balloon angular velocity component (x3) and to the dangling angle component (x4), and of which the eigenvalues (λ) have a part real non-zero entirely determined by said typing parameter (Tc).
    [0004]
    4. Method according to claim 3 characterized in that the eigenvalues (A) of the sub-matrix (Ar) have a real part (Re (A)) non-zero directly proportional, or even equal, to said typing parameter 20 (Tc ) or conversely (1 / Tc) of the typing parameter.
    [0005]
    5. Method according to one of the preceding claims characterized in that the corrective gains (km, ki2) are also expressed as a function of the length (L) of the suspension cable (3) which connects the suspended load (1) to the point hook (H). 25
    [0006]
    6. Method according to one of the preceding claims characterized in that it uses a dynamic model involving the mass (M) of the point of attachment (H) and that (m) of the suspended load (1), said model dynamic being defined as follows: either by a model called "without coupling" expressed by: = AX + BU 30 Y = CX xi (0 0 0 a1 1 0 0 0 X = X2 A = 1 0 0 0 B = 0 C = 0 0 0 0 X3 0 0 0 a2 b2 0 0 1 0 X4 0 0 1 0) 0 0 0 0 1with: in distribution orientation a = mrgx. ay = -mtg _._ a 1 xi = y = g + gmt Xi = A = Tt X2 = x - ao .x3 = 0 = 1 + M rx2 at g 1 + M rx2 + mrx2 x2 = y. ae = 0 = a2 = L b1 = 1 b2 1 u2 = - L = a2 L 1 + M X2 r 1 b at X4 = 0 U = F, X3 - at x4 = 0 U = F y 1 = 1 + Mrx2 1 X b2 = L1 + M X2 r either by a so-called "coupled" model expressed by: = + + B2u2 Y = CX (0 0 0 al 1 0 0 0 0 0 0 a2 0 0 1 0 b1 b3 0 0 b2 b4 0 0 xl 1 0 0 0 0 0 0 0 0 0 1 0 0 0 0 1 X2 A, B1 = B2, C = X3 X4_ with: in distribution orientation mrgx Xi = ';% al = -mtg X1 = .5C X2 = y g + gm, X2 = X X3 = e = X3 = 0 X4 = 0 X4 = 0 ni = Fy u2 = Fx if, = Fx u2 = Fy al = 1 + Mrx2 g 1 + Mrx2 + mrx2 = a2 L b1 = 1 1 b2 = - L b3 = 0 0 a2 L 1 + M X2 r Ill = 1 1 ± MrX2 1 xL ¢ b2 = L 1 + Mrx2 b3 = 0 b4 = 0 b4 = (1 + Mrx2) models in which: m [kg] is the mass of the suspended load (1) M [kg] is the mass of the point of attachment (H) 1-y [kg.m2] is the moment of inertia of the hoist ( 2) carrying the point of attachment (H) with respect to the axis of orientation (ZZ ') y [rad] is the angular position of the point of attachment (H) around the axis of orientation (ZZ ') x [m] is the distance from the point of attachment (H) to the axis of orientation (ZZ ') L [m] is the length of the suspension cable (3) connecting the suspended load (1) to the point of attachment (H) (1) [rad] is the radial component of the swing angle of the suspended load (1) in the vertical plane containing the distribution axis (XX ') 0 [rad] is the orthoradial component of the swing angle of the suspended load (1) in the plane vertical tangential to the rotational movement of the point of attachment (H) Ty [Nm] is the driving torque applied to drive the point of attachment (H) in rotation around the axis of orientation (ZZ ') Fx [N ] is the force exerted on the point of attachment (H) to move it in translation along the distribution axis (XX ') F_ = x TF = - M m /' / r / 7, yy and in that that we apply to the model considered the command by state feedback: = Ko.V, - K, X with KF [K0,0, k, 2], with, by convention i = d for the regulation in distribution, and i = o for orientation regulation.
    [0007]
    7. The method as claimed in claim 6, characterized in that the choice of correcting gains (km, k, 2) is as follows: mk 2 1 -T b2 1 T2b2 where i = d for the control in distribution, and i = o for the control in orientation.
    [0008]
    8. Method according to one of the preceding claims, characterized in that the typing parameter (Tc) is chosen in a range between 0.2 and 2, preferably between 0.3 and 1.8.
    [0009]
    9. Method according to one of the preceding claims characterized in that it comprises an open loop control step which consists in applying the filtered control setpoint (Y, ") ) To a control system 10 of the hoist in open loop, that is to say to a control system which applies said filtered piloting instruction (Y, ") ,) to point-of-sight driving means (H), designed to drive the point of attachment according to the movement (R, T) under consideration, without using a measured or calculated return of the angle or speed angular of the actual swing of the actual suspended load, and preferably no measured or calculated return of the effective speed of displacement of the actual hooking point.
    [0010]
    10. Calculator or data carrier readable by a computer characterized in that it contains computer program code elements allowing the implementation of a servo-control method according to one of the preceding claims when said code elements. are read by said calculator.
    [0011]
    11. Lifting device (2), such as a tower crane, comprising a mast (4) extending substantially along a first vertical axis, said orientation axis (ZZ '), an arrow (5) secant auditing mast which extends along a second axis, said distribution axis (XX '), and which carries a point of attachment (H) to which a load (1) can be suspended, as well as drive members allowing moving said hooking point (H) in rotation (R) about the axis of orientation (ZZ ') and in translation (T) along the distribution axis (XX'), said lifting gear being characterized in that the driving member (s) associated with at least one of said two rotational (R) and translational movements (T), and preferably with each of these two movements, are controlled by a control system. controller provided with calculating and programming means adapted to filter in real time, by a method according to one of claims 1 to 9, a gross piloting setpoint (V;), preferably a speed setpoint, defined by the driver of the hoist, and to apply the filtered control setpoint (Y, ") ,) to the corresponding drive member.
    类似技术:
    公开号 | 公开日 | 专利标题
    FR3016872A1|2015-07-31|ANTI-BALLING CONTROL METHOD WITH ADJUSTABLE ASSISTANCE FOR TRANSPORTING A SUSPENDED LOAD
    US9676464B2|2017-06-13|Stabilized directional control systems and methods
    EP3305710B1|2019-05-29|Method for controlling an anti-oscillatory crane with a third-order filter
    EP2497555B1|2013-08-28|Method for piloting a rotary-wing drone with multiple rotors with curved steering.
    EP2957871B1|2017-05-03|Method of prediction of an aircraft short term flight path, computer program product, associated prediction device, guidance method, guidance system and aircraft
    EP2420909A1|2012-02-22|Method for managing a yaw movement of a taxiing aircraft.
    FR2891051A1|2007-03-23|METHOD AND DEVICE FOR REALIZING AT LEAST ONE FLIGHT TEST ON AN AIRCRAFT AND APPLICATIONS
    EP2219988B1|2013-04-10|Device and method for regulating the displacement of a suspended load
    FR2976686A1|2012-12-21|METHOD AND SYSTEM FOR ATTITUDE CHANGE CONTROL AND PROGRAM RECORDING MEDIUM.
    EP3309642A1|2018-04-18|Method of assisted piloting of a rotary wing aircraft having at least one propulsion propeller, an assisted piloting device, and an aircraft
    JP6334567B2|2018-05-30|Mechanical torque control system for vehicles with two drive wheels operated by different motors
    FR2964206A1|2012-03-02|CONTROL DEVICE SYSTEM AND DEVICE FOR THE PRODUCTION OF A VIRTUAL MODEL IN REAL-TIME
    EP3048503A1|2016-07-27|Method for one or more obstacles avoidance by an aircraft, and related computer program product, electronic system and aircraft
    FR2881849A1|2006-08-11|METHOD AND DEVICE FOR CONTROLLING A PLANE IN BLANK
    FR3018364A1|2015-09-11|METHOD OF DETERMINING AN OBSTACLE AVIATION GUIDANCE LAW BY AN AIRCRAFT, COMPUTER PROGRAM PRODUCT, ELECTRONIC SYSTEM AND AIRCRAFT
    CA2987097C|2020-09-08|Flight control process for a rotorcraft, and rotorcraft
    JP6863011B2|2021-04-21|Steering control device
    WO2013178899A2|2013-12-05|Device for controlling the course of a vehicle
    EP0394147B1|1993-09-22|Device for controlling automatically the position and the oscillations of a suspended load during its transportation by a lifting device
    FR3031959A1|2016-07-29|METHOD AND DEVICE FOR CONJUGATING CONTROL RODS
    FR3038585A1|2017-01-13|AUTOMATIC AIRCRAFT CONTROL SYSTEM AND METHOD THEREOF
    FR2704847A1|1994-11-10|Process and device for limiting the swing of a load suspended from a motorised support
    EP3321758A1|2018-05-16|A method of controlling an electrical taxiing system
    e Silva et al.2013|Design and implementation of a PID control system for a coaxial two-wheeled mobile robot
    RU2454693C1|2012-06-27|Method of generating director control based on reference signals of object model
    同族专利:
    公开号 | 公开日
    CN104817016A|2015-08-05|
    DE102015100669A1|2015-07-30|
    CN104817016B|2018-01-09|
    FR3016872B1|2019-04-05|
    引用文献:
    公开号 | 申请日 | 公开日 | 申请人 | 专利标题
    FR2704847A1|1993-05-05|1994-11-10|Bertin & Cie|Process and device for limiting the swing of a load suspended from a motorised support|
    WO2001034511A1|1999-11-05|2001-05-17|Virginia Tech Intellectual Properties, Inc.|Nonlinear active control of dynamical systems|
    WO2006115912A2|2005-04-22|2006-11-02|Georgia Tech Research Corporation|Combined feedback and command shaping controller for multistate control with application to improving positioning and reducing cable sway in cranes|FR3056976A1|2016-10-05|2018-04-06|Manitowoc Crane Group France|METHOD OF CONTROLLING ANTI-BALLING CRANE WITH FILTER OF THE THIRD ORDER|
    CN106365043B|2016-09-12|2018-07-03|同济大学|Bridge crane semi-loop constant speed prevents shaking control method|
    CN106348172B|2016-09-28|2018-03-30|上海海事大学|Crane of the family based on positive and negative POSICAST input shapers method is anti-to shake control method|
    FR3071240B1|2017-09-21|2019-09-06|Manitowoc Crane Group France|DYNAMIC OPTIMIZATION OF A CRANE LOAD CURVE|
    TWI676088B|2019-01-11|2019-11-01|國立高雄科技大學|Backup system for detecting failure of a power-assisted device|
    法律状态:
    2016-01-21| PLFP| Fee payment|Year of fee payment: 3 |
    2017-01-12| PLFP| Fee payment|Year of fee payment: 4 |
    2018-01-19| PLFP| Fee payment|Year of fee payment: 5 |
    2020-01-21| PLFP| Fee payment|Year of fee payment: 7 |
    2021-01-21| PLFP| Fee payment|Year of fee payment: 8 |
    2022-01-19| PLFP| Fee payment|Year of fee payment: 9 |
    优先权:
    申请号 | 申请日 | 专利标题
    FR1450751A|FR3016872B1|2014-01-30|2014-01-30|ANTI-BALLING CONTROL METHOD WITH ADJUSTABLE ASSISTANCE FOR TRANSPORTING A SUSPENDED LOAD|
    FR1450751|2014-01-30|FR1450751A| FR3016872B1|2014-01-30|2014-01-30|ANTI-BALLING CONTROL METHOD WITH ADJUSTABLE ASSISTANCE FOR TRANSPORTING A SUSPENDED LOAD|
    DE102015100669.3A| DE102015100669A1|2014-01-30|2015-01-19|ANTI-PENDULUM CONTROL PROCEDURE WITH ADJUSTABLE SUPPORT FOR THE TRANSPORT OF AN ANCHORED LOAD|
    CN201510051412.2A| CN104817016B|2014-01-30|2015-01-30|Adjustable the anti-of assisted suspension load transport waves control method|
    [返回顶部]