• 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 inventive method is used to control an inverter for exchanging electrical energy with an AC system, wherein the inverter has a plurality of bridge branches and each bridge branch has a center (AR), which via an upper switch (S + R) with a positive DC rail (p ) and via a lower switch (S-R) to a negative DC rail (s) can be connected, and in the operation of the inverter from the midpoint (AR) of a bridge branch a bridge branch current (iLR, iLS, iLT) by a filter inductance (LR, LS, LT) of an output filter flows to the AC system. During operation of the inverter for influencing switching frequencies of the bridge branches, at least one of: amplitudes of the bridge branch currents (iLR, iLS, iLT) is varied with respect to a respectively predetermined average value; and zero voltage of the inverter powered AC system.
    公开号:CH711454A2
    申请号:CH01237/15
    申请日:2015-08-27
    公开日:2017-02-28
    发明作者:Walter Kolar Johann；Bortis Dominik；Kaufmann Maurus；Tüysüz Arda；Neumayr Dominik
    申请人:Eth Zürich Eth Transfer Hg E 47-49；
    IPC主号:
    专利说明:

    The invention relates to the field of electronic power converters.
    State of the art
    For speed control of three-phase motors (hereinafter referred to simply as motors) today three-phase pulse inverters (hereinafter simply referred to as inverters) used to transform a feeding DC voltage in a symmetrical three-phase voltage system with predeterminable frequency and amplitude and in the simplest case be formed by three, lying between the positive and negative DC bus bar bridge arms. Each bridge branch is in two-point design of a series connection of two transistors with antiparallel freewheeling diodes wherein the outputs applied to the input terminals of the motor between the transistors of the bridge arms are tapped and an output can be connected by turning one of the two transistors with the positive or negative DC rail.
    In order to avoid a high insulation stress of the motor windings or the occurrence of bearing currents which damage the raceways of the bearings, an LC low-pass filter is advantageously inserted between the inverter and the motor. The motor voltage then does not have a pulse-frequency discontinuous, but a continuous course with only a small switching-frequency fluctuation. For the implementation of the low-pass filter, filter inductances are branched off from the outputs of the bridge branches, and their second ends are connected to the associated motor terminals. Furthermore, filter capacitors are arranged in star or delta connection at the motor input. If also the common mode component of the phase output voltages of the bridge branches - which e.g. leads to the occurrence of bearing currents - are filtered, the filter capacitors are switched starting from the motor terminals against the positive or negative DC voltage rail. In terms of a low-inductance feedback as possible schaltfrequenter current components is a symmetrical arrangement of the filter capacitors advantageous, each of which motor terminal a capacitor of the same capacity is placed against the positive as well as against the negative DC voltage rail. The connection of the filter capacitors with the DC voltage rails further brings the advantage of decoupling the current formation in the filter inductances of the phases (the current sum is no longer forced to zero), further, a relatively small inductance value of the filter inductances or a high current ripple can be chosen the subsequent filter capacitors in any case provide for a small ripple of the motor terminal voltage.
    In order largely to avoid switching losses of the bridge branches, it is now advantageous, the amplitude of the ripple of the currents in the filter inductances slightly higher than the low-frequency current component to be formed, i. to select the local mean over a pulse period and thus, both when switching off, as well as when switching on a power transistor of the associated bridge branch to ensure a zero voltage switching (zero voltage switching). This mode of operation is referred to in the literature as Triangular Current Mode (TCM) operation (eg in "Ultrafiat Interleaved Triangular Current Mode (TCM) Single Phase PFC Rectifier, C. Margut et al., IEEE Transactions on Power Electronics, Vol. 29 , No. 2, pp. 873-882, 2014) and is, for example also described in US 4,947,309 (1990).
    For the further embodiments, reference is made to FIG. 1 with respect to a circuit topology and with respect to a time characteristic of a bridge branch current iL (current profile in the filter inductance of one phase) to FIG. 2. It is assumed that over a switching cycle, a positive local average of the current in a filter inductance L is to be formed. For this purpose, by switching on an upper transistor S + of an associated bridge branch with center point or connection point AR, a current iL is built up smoothly in the direction of the filter output, wherein the difference between a DC voltage uDC and a voltage uC measured against the negative DC voltage rail uC of an associated, against n switched filter capacitance C- (as mentioned above, in general, a filter capacitor is switched against the positive DC voltage rail p, over which the difference of uDC and uC occurs) acts as a current-increasing. (It should be noted that the voltage uC iA from a DC component equal to half the DC voltage, VI uDC, and one, this DC component superimposed alternating component uM is composed, which, for example, the formation of a symmetrical AC voltage with the same positive and negative peaks with a maximum Height equal to half the DC voltage, ie a symmetrical Aussteuerbarkeit is given.) When switching off S + at a time tS + off or at a current value iLS + off, also called off current, the parasitic output capacitance CS + of S + takes over the current and acts thus in the sense of a de-energized switching as a discharge capacity. In addition to the charging of CS + by iL, the parasitic output capacitance CS- of the bridge branch opposite transistor S- discharged (iL divides between CS + and CS-, ultimately, the transhipment takes place in the form of an oscillation between L and the parallel circuit of CS + and CS-) until finally the voltage at CS- at time tD-on or at a current value iLD-on the value zero and at CS + reaches the value uDC, whereby the antiparallel freewheeling diode D- begins to conduct from S- and the current iL is reduced against the voltage uC- of the filter capacitor C-. S- can be switched on without voltage. If S- is now in the on state, iL can reverse its direction after complete reduction to zero at time tiL0-, ie. build up in a negative direction or over S-. If S- is then left in the ON state for a time interval DT-, starting from the occurrence of the zero crossing from iL to negative values detected by a current zero detector ZC, and turned off at the time tS-off or at a current value iLS-off, CS - Charged by the negative current iLS-off and correspondingly discharging CS +, ie the voltage u at the output A of the bridge branch in the form of a vibration between L and the parallel circuit of CS + and CS- out against the positive DC voltage rail DC +. The voltage across S + then reaches zero in tD + on and diode D +, which is antiparallel to S +, begins to conduct. Due to the conduction of D +, S + can now be turned off without voltage, so that when at time tiL0 + the current in L coming from negative values decreases to zero (which is detected by the current zero detector ZC), a build-up in the positive direction can take place until again as described above, after a time DT + counted from the current zero crossing in tiL0 + has elapsed, the current value iLS + off is reached at time tS + off and S + is switched off. Thus, a switching cycle of the length Ts of the bridge branch is completed.
    It should be noted that the above-mentioned resonant transition of u between DC and DC + requires a sufficiently high value of the current iLS-off. Such as. in The Design of High Performance Mechatronics, R.M. Schmidt et al., Delft University Press, 2011, ISNB 978-1-60750-826-7 (see page 440), iLS-off, in the sense of simple realization, is independent of the local mean iL (1 ) of iL to a constant, for the swinging sufficiently high value set.
    Overall, therefore, iL approximately triangular shape, where iL (1) by appropriate selection of the switching times tS + off and tS-off and the length of the time intervals DT + and DT- can be defined so that the motor phase current supplied to IM and the proportion of the current in the filter capacitors C- and C- (acting in parallel) with output frequency (it is indeed a, dependent on the set engine speed frequency of the output voltage u) are covered. The (high) switching frequency ripple of iL closes over the filter capacitors C + and C-, and causes a corresponding switching frequency ripple of uC, which, however, can be kept small by appropriate dimensioning of the filter capacitors. A disadvantage, however, is that the duration of a switching cycle Ts except for the current to be formed iL (1) and also the ratio of the filter capacitor voltage uC and the DC voltage uDC, i. depends on the drive level of the inverter, since the current build-up is determined by the difference (uDC-uC) and the current reduction by -uC. Within the operating range of the inverter, high or low motor currents iM or local average values iL (1) of the currents in the filter inductances, respectively occur depending on the torque to be formed in accordance with the speed to be set or low or high values of the motor terminal voltage uM. In any case, due to the sine wave of the motor currents in the vicinity of the zero crossings, there are small current values. The advantage of de-energized switching of the bridge branches given for the triangular shape of iL is therefore due to a wide variation of the switching frequency fs = 1 / Ts, which makes it more difficult to suppress conducted interference emissions by another EMC filter stage inserted between the LC low-pass filter and the motor. to the signal processing and the control of the transistors S + and S- in terms of permissible delay times special requirements.
    In order to narrow down the switching frequency variation, the replacement of the bridge branch S + and S- and the filter inductance L by several parallel bridge branches with their own filter inductances is therefore described in US-2004/0 151 010 A1, 2004. For small motor current amplitudes, only one bridge branch is then in operation, at higher currents, another bridge branch is activated, so that a higher current amplitude is split into two parallel (and advantageously phase-shifted) branches, ie not a bridge branch has to handle the entire variation of the current value. However, in particular in the case of more than two bridge branches operating in parallel, the complexity of the system is substantially increased as a result. It should be noted that the decision whether a bridge branch or both bridge branches are in operation depends on the current amplitude and not on the respective current instantaneous value.
    The object of the invention is therefore to provide a control method which retains the advantage of de-energized switching of the bridge arms, and even with the arrangement of only one bridge branch and low-pass filter per motor phase regardless of the amplitude of the voltage applied to the motor voltage, and the motor current amplitude and the phase shift of motor voltage and motor current in TCM operation avoids a wide variation of the switching frequency over the output period (motor voltage period) and on the other hand guarantees the lowest possible switching losses.
    This object is achieved by a method according to the claims.
    The method is used to control an inverter for exchanging electrical energy with an AC system, wherein the inverter has a plurality of bridge arms and each bridge branch has a center, which has an upper switch with a positive DC rail and a lower switch with a negative DC rail can be connected, and flows in the operation of the inverter respectively from the midpoint of a bridge branch a bridge branch current through a filter inductance of an output filter to the AC system. In the operation of the inverter for influencing switching frequencies of the bridge branches at least one of:Amplitudes of the bridge branch currents with respect to a given average value; andZero voltage of the inverter powered AC system.
    Each of the bridge branch currents thus oscillates, in a manner known per se, according to the switching frequency around the respective mean value. This is determined by a higher-level regulation. The average essentially follows an oscillation with a mains frequency or fundamental frequency according to the AC system, with deviations corresponding to a momentary load of the AC system. The switching frequency is a multiple of the fundamental frequency of the AC system. Through an output filter, the bridge branch currents are smoothed and the smoothed bridge branch currents appear at an interface to the AC system. By varying the deviations of the bridge branch currents by the mean value, the frequency of the switching operations in the bridge branch is changed, and thus the switching frequency of the bridge branches. The amplitude of the bridge branch currents thus represents a degree of freedom for influencing the switching frequency.
    Independently or in combination with the variation of these amplitudes, the zero voltage of the fed system can be varied. According to the prior art, it is equal to zero. However, it may be non-zero, for example, if a star point of the AC system is not grounded or not connected to a midpoint of the DC side. If the zero voltage is varied, so that the differential voltages can be influenced, which rest against the filter inductances, and thus in turn the steepness of the current change of the bridge branch currents. If the voltages after the filter inductances are kept as close as possible to the voltage of the positive or negative DC voltage rail (depending on the phase position), the differential voltages are relatively smaller in comparison to a purely sinusoidal profile of the voltages. With a smaller differential voltage, the steepness of the current change is smaller and thus the switching frequency in the corresponding bridge branch. The variation of the zero voltage thus represents a further degree of freedom for influencing the switching frequency.
    In a variant of the method, the amplitude of the corresponding bridge branch current is increased to reduce the switching frequency of a bridge branch.
    In a variant of the method, the zero voltage and thus a voltage at the output of the filter inductance of the bridge arm in the direction of the voltage of the positive DC rail or the negative DC rail is moved to reduce the switching frequency of a bridge branch.
    In a variant of the method are each a plurality of parallel bridge branches for one phase of the AC system, and are in each phase in accordance with a current instantaneous value of this phase and a target voltage at the output of the filter inductance of this phase one or more of the parallel-connected bridge arms operated this phase.
    In one variant of the method, an optimization method is carried out to determine switching times of the switches and / or for determining a setpoint for the zero voltage, whichas objective function has a weighted sum of a measure for a variation width of the switching frequencies of the bridge arms with a measure of losses, in particular switching losses of the inverter, andhas predetermined values for voltages and currents at an interface to the AC system given as boundary conditions by a superordinate control.
    In a variant of the method, the optimization method is carried out offline and thus an online used in the operation of the inverter multi-dimensional table or a mathematically equivalent function generated, this table respectively functionhas setpoint values for mean values of the bridge branch currents as well as nominal values of voltages at the interface to the AC system as input variables,and as output values, time intervals which in each case predetermine a switch-off delay after a corresponding zero crossing for the switches of the bridge branches, and / or has a nominal value for the zero voltage as output value.
    Instead of the setpoint values for mean values of the bridge branch currents as well as setpoint values of voltages at the interface to the AC system, if the AC system is a symmetric polyphase system, values of an equivalent description can also be specified, for example a current amplitude and voltage amplitude common to all phases a phase shift.
    In a variant of the method is done to determine operating values in the form of switching times of the switches and / or a setpoint for the zero voltage scaling of over a period predetermined values for these operating values, depending on an amplitude of the currents and an amplitude of the voltages at an interface to the AC system and a phase shift between these voltages and currents.
    For example, this is followed by a course of predetermined values of the nominal value for the zero voltage essentially a third harmonic of the voltages at an interface to the AC system (also called output voltages), and is the profile of switch-off currents iLS-offR, iLS-offS , iLS offT approximately triangular in the vicinity of the zero crossings of the setpoints for mean values of the bridge branch currents (see Fig. 3).
    In a variant of the method, an optimization is carried out on-line, in particular according to the concept "Perturb & Observe", withDetermination of detected values by measuring the occurring switching frequency on the basis of current zero crossings, measurement of occurring losses by subtraction of measured input and output power or calculation by means of a model, andby adaptation of operating values in the form of switching times of the switches and / or a setpoint value for the zero voltage according to the values thus detected.
    In the following, the subject invention will be explained in more detail with reference to preferred embodiments, which are illustrated in the accompanying drawings. Each show schematically:<Tb> FIG. 1 <SEP> a circuit arrangement with a control structure according to an embodiment of the invention;<Tb> FIG. 2 <SEP> a time profile of a current profile in the filter inductance of a phase;<Tb> FIG. 3 <SEP> time profiles of various signals from the circuit of FIG. 1; and<Tb> FIG. 4 <SEP> a comparison of switching frequencies with and without a regulation according to the invention.
    An embodiment of the invention hereinafter based on the figures Fig. 1 -Fig. 4, the control circuit of a TCM inverter for variable-speed operation of a motor is to be expanded (see dash-dot framed circuit part in FIG. 1) such that a degree of freedom given due to a missing connection of the neutral point N of the motor windings to the feeding DC voltage uDC addition of a zero voltage with setpoint u0 * («*» denotes setpoint values) to the actual motor phase voltages uCR *, uCS *, uCT * to be used for limiting the switching frequency variation. The zero voltage corresponds to the voltage of a zero point N of a three-phase system supplied by the inverter. The influence of a zero voltage u0 on the length of a switching cycle Ts (for a given local average current iL (1)) is determined by the above-described dependence of the rate of change of the current in L on (uDC-uC) on current build-up and on (-uC) understandable at power down. If μC assumes values close to uDC, the current increases only very slowly and the duration of the switching cycle Ts assumes very high values or the switching frequency fs = 1 / Ts very low values. If, instead of the filter capacitor voltage uC, the voltage uC = (uC) + u0 with u0 <0, the current build-up will be faster and Ts will be shortened (despite the then lower speed of the current decrease determined by - ((uC) + u0)). Since the zero voltage u0 is added to all the motor phase voltages uM which are actually to be formed, the choice of the voltage value u0 on the one hand must take into account that in no phase the voltage uCR *, uCS *, uCT * extended by the zero value is the value uDC exceeds or falls below the value zero, ie Overmodulation is avoided. On the other hand, u0 on the one hand results in the lowering of the frequency of one phase, but on the other hand u.U. increasing the frequency in another phase. Therefore, for the determination of u0 *, all phases are considered and further, it is taken into account which local mean current value is to be formed in one phase. In this context, or independently thereof, the negative current iLS-off and thus the variation of the amplitude of the current by its predetermined mean value as a further or as the only degree of freedom of TCM operation is included in the considerations. A, in one phase, e.g. R, required local mean current value iLR (1) can indeed be formed even at relatively high negative current values iLS-offR, if a correspondingly high positive current value iLS + offR is set, which compensates the effect of iLS-offR. This makes it immediately clear that iLS-offR can be used to increase Ts or reduce fs. However, this also increases the conductivities in the transistors S + and S-. The choice of u0 * and the current values iLS-offR, iLS-offS, iLS-offT therefore has to be made, on the one hand, with regard to the phase voltages uCR *, uCS *, uCT * which are actually to be formed, which are determined by uDC-max (uCR *, uCS *, uCT *) define an upper limit, and with -min (uCR *, uCS *, uCT *) a lower maximum value of u0 *; On the other hand, the local average currents iLR (1) *, iLS1) *, iLT (1) * to be set must be taken into account. Since an intervention only by adding u0 * i.A. However, in any case all phases are affected and, moreover, the local switching frequencies of the bridge branches fsR, fsS, fsT are influenced by the set values JLS-offR, iLS-offS, iLS-offT, the complex interaction can be effected by an off-line or on-line optimization are solved, the quality criterion on the one hand has the goal to keep the switching frequency of all bridge arms within a frequency band (fsmin, fsmax) and on the other hand to limit the associated increase in the conduction losses of the bridge branches. If there are two bridge branches per phase, this degree of freedom can also be incorporated into the optimization. In order to avoid the occurrence of very high switching frequencies at low currents (with small motor current amplitudes, but also in the vicinity of the current zero crossings), then only one bridge branch is clocked and can be changed at higher current instantaneous values to a timing of the two bridge arms, thus preventing the occurrence of low switching frequencies can be. In particular, this change between the operation of one or both bridge branches then also takes place within a period of the motor current, that is, depending on the current instantaneous value in the respective relation of uC * and uDC.
    In summary, therefore, knowing the circuit parameters, e.g. the inductance of the low-pass filter inductors, the number of parallel bridge branches and the intended limits of the switching frequency, as well as the on-resistance or current dependence of the conduction losses of the power transistors to perform an optimization, which further the current state of the circuit in the form of the DC voltage uDC and on the part of the Motor control required and actually set filter capacitor voltages uCR *, uCS *, uCT * as well as the respectively to be formed local average currents iLR (1 *, iLS (1) *, iLT (1) * into account and finally times DT + R, DT + S, DT + T and DT-R, DT-S, specifies DT-T, which keep the local switching frequencies within predetermined limits, ie fsmin <fsR <fsmax, fsmin <fsS <fsmax, fsmin <fst <fsmax with only a moderate increase in the conduction losses Since the DC voltage typically has a constant value, for each combination of values (uCR *, uCS *, uCT *, iLR (1) *, iLS (1) *, iLT (1) *) an optimum set of operating values (DT + R, DT + S, DT + T, DT-R, DT-S, DT-T, u0 *), e.g. by means of a multidimensional table or functionally equivalent (interpolation) functions or a combination of tables and functions; in the presence of more parallel bridge branches per phase, the dimension of the table increases accordingly, since then each phase, e.g. two values DT + and DT- are to be output or a bridge branch must be permanently locked. The value range of uCR *, uCS *, uCT * is determined by uDC and the value range of iLR (1) *, iLS (1) *, iLT (1) * by the maximum permissible peak value of the motor current, taking into account the maximum output frequency required low-frequency component of the filter capacitor current iCR *, iCS *, iCT * defined taking into account a safety factor and discretized accordingly. The operating values (DT + R, DT + S, DT + T, DT-R, DT-S, DT-T, u0 *) for intermediate values are then to be calculated by known approximation methods.
    In suboptimal form, the TCM performance can also be improved thereby, i. the switching frequency variation is limited by adding to the actual motor phase voltages uM * a zero quantity u0 * in the form of a third harmonic with 1/6 of the amplitude of the fundamental voltage oscillations uMR *, uMS *, uMT * actually to be formed at the motor terminals; Current values iLS-off of the phases in the vicinity of the zero crossings of the associated phase currents towards the zero crossing are increasingly raised. This increase can then be in normalized form and scaled according to the respective operating state. Overall, this provides a significant reduction in complexity. The addition of u0 * then decreases the instantaneous values of uMR *, uMS *, uMT * in the vicinity of the amplitudes and an increase in the edge region, which results in a relatively constant voltage separation of uCR *, uCS over a wider range *, UCT * of uDC and zero and thus an approximately constant switching frequency is given. Advantageously, this also maximizes the voltage controllability of the inverter. Alternatively, the superimposed zero size u0 * can also be calculated by K (max (uCR *, uCS *, uCT *) - mm (uCR *, uCS *, uCT *)) («max» denotes the formation of the maximum value, «min »The formation of the minimum value) and then has stationary approximately a triangular course with three times the inverter output frequency. Advantageously, the voltage u 0 * is then formed in normalized form by the motor control and according to the invention only optimally scaled.
    It is important to note that the control method described above for the three-phase case, mutatis mutandis, the same for others, i. E. lower or higher phase numbers, e.g. can be used for two phases or six phases. Further, the inverter could be used for other applications, such as the output stage of an uninterruptible power supply except for the supply of a motor.
    Fig. 1: Schematic block diagram of the control circuit of an inverter in TCM operation for variable-speed operation of an engine based on a field-oriented control with dash-dotted bordered extension according to an embodiment of the invention, by means of the switching frequency variation of the inverter bridge arms by adding a zero size u0 * to the capacitor voltages uCR *, uCS *, uCT * actually to be generated for the generation of the motor voltages uMR *, uMS *, uMT * and by varying the switching times DT + R, DT + S, DT + T and DT-R, DT-S , DT-T is limited to the frequency range (fsmin, fsmax) in consideration of the formation of the required local average currents iLR (1) *, iLS (1) *, iLT1) *. The resulting increase in the conduction losses of the bridge branches is in the context of an off-line optimization, which is the mapping of (uCR *, uCS *, uCT *, iLR (1) *, iLS (1) *, iLT (1) *) on (DT + R, DT + S, DT + T, DT-R, DT-S, DT-T, u0 *) in the form of a multi-dimensional table with optimized entries, limited to moderate values. The quality function underlying the optimization on the one hand aims at maintaining fsR, fsS, fsT in (fmin, fmax) and, on the other hand, evaluates the increase in switching losses; The influence of the two goals can be changed in a conventional manner by weight factors relative to each other. The times DT + R, DT + S, DT + T, DT-R, DT-S, DT-T are synchronized by drive stages to the times t0-R, t0-S, t0 detected by current zero detectors ZCR, ZCS, ZCT -T, t0 + R, t0 + S, t0 + T of the zero crossings of the respectively associated phase current used for the control of the transistors S + R, S + S, S + T, S-R, S-S, S-T.
    Fig. 2: For TCM operation characteristic approximately triangular shape of occurring within a pulse period Ts current curve in the filter inductance of a phase (in the interest of clarity, the index of the phase designation is omitted) with designation of the switch-off of S + and S- (tS + off and tS-off), the times in which D- and D + reach the conduction state (tD-on, tD + on) and the zero crossings of the current iL (tiL0 + and tiL0-) and the respective present current values. In order to ensure the lowest possible conduction losses, S + and S- are advantageously switched through directly after the antiparallel free-wheeling diode D + (in tD + on) and D- (in tD-on), but at the latest shortly before tiL0 + or tiL0-. The switch-on times which are directly present in tiL0 + or tiL0- when S + and S- are switched on are indicated by DT + and DT-. Furthermore, the local mean value iL (1) formed within the switching period is entered, which on the one hand covers the output frequency transhipment of the (parallel) filter capacitors C + and C- and on the other hand the motor phase current iM. Furthermore, the course of the output of the associated detector ZC for the zero crossings of the current iL is shown.
    Fig. 3: from top to bottom is shown:Time course of when using the inventive control in consideration of DT + R, DT + S, DT + T, DT-R, DT-S, DT-T occurring currents iLR, iLS, iLT within an output period.Absolute value of the respectively achieved current values at switch-off (switch-off currents) iLS-offR, iLS-offS, iLS-offT for positive current average value iLR (1), iLS (1), iLT (1), or absolute value of the respectively achieved current values at switch-off (switch-off currents iLS + offR, iLS + offS, iLS + offT for negative average current iLR (1), iLS (1), iLT (1).The course of the three filter capacitor voltages uCR, uCS, uCT and the underlying course of uCR *, uCS *, uCT *.The course u0 * of the actually to be formed motor voltages superimposed optimal zero size.
    Fig. 4: Time course of the motor phase currents iMR, iMS, iMT and the local switching frequencies fsR = 1 / TsR, fsS = 1 / TsS, fsT = 1 / TsT, the associated bridge branches over a period of the output voltage in implementing the control of Inverter in TCM mode according to the prior art (a) and in inventive design (b), for which the variation of fsR, fsS, fsT limited to the frequency range (fsmin, fsmax), and thus in particular the occurrence of very high switching frequencies in the vicinity of the zero crossings of the motor phase currents is avoided.
    权利要求:
    Claims (8)
    [1]
    1. A method for controlling an inverter for the exchange of electrical energy with an AC system, wherein the inverter has a plurality of bridge branches and each bridge branch has a center (AR, ...), which via an upper switch (S + R, ... ) with a positive DC rail (p) and a lower switch (S-R, ...) can be connected to a negative DC rail (s), and in the operation of the inverter respectively from the center (AR, ...) of a bridge branch a bridge branch current (iLR, iLS, iLT) flows through a filter inductance (LR; LS; LT) of an output filter (LR, C + R, C-R, LS, ...) to the AC system, characterized in that in operation of the inverter for influencing switching frequencies of the bridge branches at least one is varied by:- Amplitudes of the bridge branch currents (iLR, iLS, iLT) with respect to a given average value; and- Zero voltage of the inverter powered AC system.
    [2]
    2. The method according to claim 1, wherein for reducing the switching frequency of a bridge branch, the amplitude of the corresponding bridge branch current (iLR, iLS, iLT) is increased.
    [3]
    3. The method according to claim 1 or 2, wherein to reduce the switching frequency of a bridge branch, the zero voltage and thus a voltage (UCR, UCS, UCT) at the output of the filter inductance (LR; LS; LT) of the bridge branch in the direction of the voltage of the positive DC rail ( p) or the negative DC rail (s) is moved.
    [4]
    4. Method according to one of the preceding claims, wherein there are in each case a plurality of bridge branches connected in parallel for one phase of the AC system, and wherein in each phase, in accordance with a current instantaneous value of this phase and a nominal voltage at the output of the filter inductance (LR; LS; Phase of one or more of the parallel-connected bridge branches of this phase are operated.
    [5]
    5. The method according to any one of the preceding claims, wherein for the determination of switching times (tS + off, tS-off, ...) of the switch (S + R, S-R, ...) and / or for determining a desired value ( u0 *) for the zero voltage (u0) an optimization method is performed, which- has as target function a weighted sum of a measure for a variation width of the switching frequencies of the bridge arms with a measure of losses, in particular switching losses of the inverter, and- Has as boundary conditions specified by a higher-level control setpoints for voltages and currents at an interface to the AC system.
    [6]
    6. The method according to claim 5, wherein the optimization method is carried out offline and thus an online used in the operation of the inverter multi-dimensional table or a mathematically equivalent function is generated, this table or functionSetpoint values for mean values (iLR (1) *, iLS (1) *, iLT (1) *) of the bridge branch currents (iLR, iLS, iLT) as input quantities as well as setpoints of voltages (uCR *, uCS *, uCT *) at the Has an interface to the AC system,And as output values time intervals (DT + R, DT + S, DT + T, DT-R, DT-S, DT-T), which respectively give the switches of the bridge branches a delay of switching off after a corresponding zero crossing, and / or has as output value a setpoint (u0 *) for the zero voltage (u0).
    [7]
    7. The method according to any one of claims 1 to 4, wherein for the determination of operating values in the form of switching times (tS + off, tS-off, ...) of the switch (S + R, S-R, ...) and / or a setpoint value (u0 *) for the zero voltage (u0) a scaling of over a period predetermined values for these operating values occurs, depending on an amplitude of the currents and an amplitude of the voltages at an interface to the AC system and a phase shift between these voltages and streams.
    [8]
    8. The method according to any one of claims 1 to 4, wherein an optimization is carried out on-line, in particular according to the concept «Perturb & Observe», with- Determination of detected values by measuring the switching frequency occurring on the basis of current zero crossings, measurement of occurring losses by subtraction of measured input and output power or calculation by means of a model, andBy adapting operating values in the form of switching times (tS + off, tS-off,...) The switch (S + R, S-R,...) And / or a setpoint value (u0 *) for the zero voltage ( u0) according to the values thus acquired.
    类似技术:
    公开号 | 公开日 | 专利标题
    DE102010007184B4|2013-01-24|PWM Rectifier
    EP2661806B1|2014-11-26|Electrical circuit and method for operation thereof
    DE60120664T2|2007-06-06|PWM FREQUENCY CONVERTER
    DE102005028945A1|2006-04-27|Motor driving device
    WO2006000111A1|2006-01-05|Low harmonic multiphase converter circuit
    EP2831989B1|2021-05-05|Rectifier circuit with current injection
    WO2009003959A2|2009-01-08|Inverter system and control method
    EP2678933B1|2017-08-30|Method for producing an output voltage at a moduler multilevel converter and assembly for performing the method
    EP3183809A1|2017-06-28|Mains-coupled inverter, inverter arrangement, and operating method for an inverter arrangement
    EP2992595A1|2016-03-09|Converter assembly having multi-step converters connected in parallel and method for controlling said multi-step converters
    EP2156545A1|2010-02-24|Pwm rectifier having target voltage corrector by zero system voltage for reducing switching losses and preventing noise generation
    DE102017115639A1|2019-01-17|Reduction of ripple current during switching operations of a bridge circuit
    EP2732544A1|2014-05-21|Method for regulating a centre voltage
    CH711454A2|2017-02-28|Control methods for minimizing EM interference and loss of multi-phase AC / DC converters with TCM operation of the bridging branch.
    EP3563475B1|2022-01-19|Power converter circuit and method for controlling same
    EP3300242B1|2020-10-28|Regulation of a three-level inverter
    CH715005B1|2021-10-29|Device for converting a DC voltage that varies within wide limits into a multi-phase AC voltage with variable frequency and amplitude.
    EP2985896B1|2017-01-25|System for increasing the network-side power factor of three phase ec motors
    CH714180B1|2021-11-15|Converter for transferring electrical energy between a DC and an AC system.
    EP3353885A1|2018-08-01|Method for operating a modular multi-level power converter, modular multi-level power converter, and computer program
    EP3576284A1|2019-12-04|Electrical coupling of a first electrical network with a second electric network
    DE112018007147T5|2020-11-05|Rectifier circuit device
    DE102020208381A1|2022-01-05|Power converter and method for operating a power converter
    WO2010088979A1|2010-08-12|Method for regulating a self-commutated mains power converter of an intermediate voltage circuit converter
    EP0234040A1|1987-09-02|Control process for a quick reactive current compensation of parallel thyristor-controlled inductances
    同族专利:
    公开号 | 公开日
    CH711454B1|2019-07-31|
    引用文献:
    公开号 | 申请日 | 公开日 | 申请人 | 专利标题
    法律状态:
    2018-11-15| PCAR| Change of the address of the representative|Free format text: NEW ADDRESS: POSTFACH, 8032 ZUERICH (CH) |
    2021-03-31| PL| Patent ceased|
    优先权:
    申请号 | 申请日 | 专利标题
    CH01237/15A|CH711454B1|2015-08-27|2015-08-27|Control method to minimize EM noise and losses of multi-phase AC / DC converters with TCM operation of bridge arms.|CH01237/15A| CH711454B1|2015-08-27|2015-08-27|Control method to minimize EM noise and losses of multi-phase AC / DC converters with TCM operation of bridge arms.|
    [返回顶部]