Method for synchronous space vector modulation with asynchronous sampling at a power converter for operating an electric machine

Abstract

Method for synchronous space vector modulation with asynchronous sampling at a power converter for operating an electric machine, wherein a sequence of space pointers is adjusted by introducing an additional time to freely shift the sequence, wherein this additional time is calculated online from a flow trajectory (FT), where a control unit sets a target voltage v α β * ( k ) calculated using a deadbeat rule formation rule to track the set flow trajectory (FT) and where the magnitude of the flow trajectory (FT) is determined using a minimum function for each target flow angle, characterized by the fact that Then, when a target flow magnitude is less than a maximum, a variable inhex is activated, a change in the flow section is detected with the modulo operator and stored in a variable LR, then, when both variables inhex and LR are zero, 6-step PWM is performed, then, when the variable LR is 1 and the variable inhex is zero, a complete sequence is performed, then, when the variable LR is zero and the variable inhex is 1, the pulse width modulation sequence is partially applied and completed in the next sampling task.

Description

[0001] The invention relates to a method for synchronous space vector modulation with asynchronous sampling at a power converter for operating an electric machine according to the preamble of claim 1 and a vehicle according to the preamble of claim 6.

[0002] US 2017 / 0331410 A1 describes a control device for an alternating current (AC) motor in which a control mode selector compares a modulation rate with a modulation rate threshold and selects between a synchronous control mode to perform synchronous control in a range where the modulation rate is equal to or greater than the modulation rate threshold, and an asynchronous control mode to perform asynchronous control in a range where the modulation rate is less than the modulation rate threshold. The control mode selector uses a different modulation rate threshold in a low-speed range and a high-speed range, such that the modulation rate threshold in the low-speed range is greater than the modulation rate threshold in the high-speed range.A selector chooses between the output of the synchronous pulse signal and the output of the asynchronous pulse signal.

[0003] WO 2014 / 064141 A1 describes an electrical converter comprising several semiconductor switches, wherein the electrical converter is designed to generate a two-stage or multi-stage output voltage from an input voltage by switching the several semiconductor switches.A method for controlling the electrical transducer comprises the steps of: receiving an electrical reference quantity and an actual electrical quantity; determining a sequence of future electrical quantities of the electrical transducer from the actual electrical quantity; determining a maximum cost value based on the sequence of future electrical quantities; iteratively determining an optimal switching sequence for the electrical transducer, wherein a switching sequence comprises a sequence of future switching states for the semiconductor switches of the electrical transducer; and selecting the first switching state of the optimal switching sequence as the next switching state to be applied to the semiconductor switches of the electrical transducer.The optimal switching sequence is determined iteratively by: extending a switching sequence by appending a possible switching state to the switching sequence; determining a cost value for the extended switching sequence using a cost function based on the sequence of future electrical quantities; and discarding the extended switching sequence if the cost value is higher than the maximum cost value.

[0004] WO 2015 / 078656 A1 describes a method for controlling an electrical transducer, comprising the steps of: determining an error value based on a difference between an estimated output value and a reference output value, the estimated output value based on measurements in the electrical transducer; comparing the error value with an error band and controlling the electrical transducer by switching to a different control scheme if the error value exceeds the error band.The converter is controlled by modified pre-calculated switching by: determining a pre-calculated switching sequence for the converter based on an actual state of the electrical converter, wherein the switching sequence comprises a sequence of switching transitions of the converter; modifying the pre-calculated switching sequence by modifying the transition times of switching transitions of the pre-calculated switching sequence so that the error value is minimized; and applying at least part of the modified switching sequence to the electrical converter.

[0005] The generic patent US 2012 / 0161685 A1 discloses a control method for an inverter for an electrical system, which is controlled such that, in an additional step, switching sequences with switching times for the inverter are modified to optimize the control. In this process, resulting flux errors are determined and the switching times are modified to reduce the flux errors.

[0006] The invention is based on the objective of providing a novel method for synchronous space vector modulation with asynchronous sampling at a power converter for operating an electric machine, as well as a novel vehicle.

[0007] The problem is solved according to the invention by a method for synchronous space vector modulation with asynchronous sampling at a power converter for operating an electric machine with the features of claim 1 and by a vehicle with the features of claim 6.

[0008] Advantageous embodiments of the invention are the subject of the dependent claims.

[0009] A method for synchronous space vector modulation with asynchronous sampling at a power converter for operating an electric machine is proposed. According to the invention, a sequence of space vectors is adapted by introducing an additional time to freely shift the sequence, wherein this additional time is calculated online from a flux trajectory. For this purpose, a control unit calculates a target voltage using a deadbeat rule formula to track the set flux trajectory. The magnitude of the flux trajectory is determined using a minimum function for each target flux angle.

[0010] According to the invention, when a target flux magnitude is less than a maximum, a variable inhex is activated, wherein a change of the flux section is detected with the modulo operator and stored in a variable LR, wherein 6-step PWM is performed when both variables inhex and LR are equal to zero, wherein a complete sequence is performed when the variable LR is equal to 1 and the variable inhex is equal to zero, wherein when the variable LR is equal to zero and the variable inhex is equal to 1, the pulse width modulation sequence is partially applied and completed in the next sampling task.

[0011] In one embodiment, the additional time is a start time up to the beginning of the sequence in a computation period or PWM period, or the start time is extended or shortened by the additional time, with a remaining time being shortened or lengthened accordingly.

[0012] In one embodiment, the proportions of the flow trajectory correspond to the proportions of the computation period.

[0013] In one embodiment, a maximum length of a flux hexagon is calculated as a function of a DC bus voltage and an electric angular velocity, and at least one flux trajectory is calculated based on this maximum length.

[0014] In one embodiment, to calculate a time or additional time until the first switching or until the start of a PWM section, an additional flux setpoint is introduced together with an additional virtual voltage vector, and the time or additional time is calculated using the equation T1=U*Udc3sin(π3−ϕUαβ*) calculated.

[0015] According to one aspect of the present invention, a vehicle is proposed comprising at least one electric machine for propelling the vehicle, at least one power converter for operating the electric machine, and a control unit for controlling and / or regulating the power converter. According to the invention, the control unit is configured to carry out the method described above.

[0016] According to the solution of the invention, the space vector modulation is adapted (sequence) and a time variable is added, which makes it possible to position the pulse completely flexibly during the computation period. This allows the pulse width modulation to be synchronized with the electrical rotation frequency. The computation and switching frequencies are decoupled. The additional time variable at the beginning of the sequence is calculated from the flow trajectory.

[0017] According to the solution of the invention, the space vector modulation is extended to synchronous, optimized pulse patterns. Higher modulation levels can thus be achieved without pulse dropping occurring in the overmodulation range.

[0018] The solution according to the invention allows for higher power output through a higher modulation level. This results in fewer losses due to lower field weakening current.

[0019] Exemplary embodiments of the invention are explained in more detail below with reference to drawings.

[0020] This shows: Fig. 1. A schematic diagram of an asynchronous switching function according to the state of the art, Fig. 2 a schematic diagram of a synchronous switching function according to the state of the art with a pulse count q=7, Fig. 3 a schematic diagram illustrating the decomposition of a pointer into sub-pointers parallel to several axes according to the state of the art, Fig. 4 a schematic diagram for determining the maximum length of the pointer according to the state of the art, Fig. 5 a schematic diagram with possible sequences of the converter states according to the state of the art, Fig. 6 a schematic view of a sequence of stress vectors with associated times, Fig. 7 a schematic view illustrating the division of a target voltage vector, Fig. 8 a schematic diagram of a river trajectory, Fig. 9 a schematic diagram showing the time profiles of the midpoint voltages of three half-bridges of a power converter, Fig. 10 a schematic detail view of a calculation period or PWM period of the midpoint voltage, Fig. 11 a schematic diagram of a flux trajectory and the time course of the midpoint voltages of three half-bridges of a power converter to illustrate an example simulation, Fig. 12 a schematic diagram of a flux trajectory and the time course of the midpoint voltages of three half-bridges of a power converter to illustrate an example simulation with seven pulses, where a long start time is provided, Fig. 13 a schematic diagram of a flux trajectory and the time course of the midpoint voltages of three half-bridges of a power converter to illustrate an example simulation with seven pulses, where a long residual time is provided, Fig. 14 a schematic diagram of a flux trajectory and the time course of the midpoint voltages of three half-bridges of a power converter, Fig. 15 a schematic flowchart that represents a decision process for carrying out sequences based on the conditions of the flow trajectory and two parameters, Fig. 16 a schematic diagram of a detail of an 18-corner flow trajectory at one corner, and Fig. 17 a schematic diagram of an exemplary PWM synchronization based on a flow trajectory.

[0021] Corresponding parts are marked with the same reference symbols in all figures.

[0022] Overmodulation results in pulse loss and a reduction in switching frequency. Asynchronous pulse patterns (SVPWM, DPWM) are only recommended for pulse counts q of at least 7.

[0023] Fig. Figure 1 is a schematic diagram of an asynchronous switching function according to the state of the art from Felix Jenni and Dieter Wüest, Control Methods for Self-Commutated Converters. If the pulse count q, that is, the ratio q = fs / fel of a switching frequency fs and an electrical frequency fel, is not an integer, then it is called asynchronous switching. Fig. For example, in case 1 the ratio q is approximately 7.5.

[0024] In asynchronous clocking, the fundamental oscillation m is generated with different individual pulses in each period, while in synchronous clocking it is always generated with the same individual pulses.

[0025] Fig. Figure 2 is a schematic diagram of a synchronous switching function according to the state of the art with the pulse number q=7 from Felix Jenni, Dieter Wüest, Control methods for self-commutated power converters.

[0026] The goal is to adjust the pulse width modulation or space vector modulation so that it is set synchronously, even if sampling and calculation are asynchronous.

[0027] Fig. Figure 3 is a schematic diagram illustrating the decomposition of a pointer u into pointers parallel to the axes a, b, c according to the state of the art from Felix Jenni, Dieter Wüest, Control methods for self-commutated converters.

[0028] Fig. Figure 4 is a schematic diagram for determining the maximum length of the pointer u according to the state of the art from Felix Jenni, Dieter Wüest, Control methods for self-commutated converters.

[0029] According to Felix Jenni and Dieter Wüest, the following applies to the relationships between the pointer u and the sub-pointers u: Control methods for self-commutated power converters a , u b and u c : |u_b|=−23|u_|sin(ω1t+π3)+|u_a|,|u_c|=−23|u_|sin(ω1t+π3)+|u_a| |u_a|=|1u_|=|1U_|t1TTast=23|u_|sin(π3−ω1t)|u_b|=|2u_|=|2U_|t21TTast=23|u_|sin(ω1t)

[0030] According to Felix Jenni and Dieter Wüest, Control methods for self-commutated converters, the duty cycle of the individual phasors can now be directly determined from equations (2): t1=|1u_||1U|TTast=3TTast|u_|Udsin(π3−ω1t),t2=|2u_||2U|TTast=3TTast|u_|Udsin(ω1t)

[0031] According to Felix Jenni and Dieter Wüest, "Control methods for self-commutated converters," this decomposition allows for the calculation of any voltage phasor. Str be built in sector S1.

[0032] If the times t1 and t2 are given, then the sum of the remaining time within the sampling period is also determined: t0+t7=TTast−(t1+t2)

[0033] In the state of the art, two active vectors and a zero vector are calculated for space vector modulation.

[0034] Fig. Figure 5 is a schematic diagram with possible sequences of the converter states according to the state of the art from Felix Jenni, Dieter Wüest, Control methods for self-commutated converters.

[0035] According to the state of the art as described in Felix Jenni and Dieter Wüest's "Control Methods for Self-Commutated Converters," sequence 1 contains three shutdown processes per sampling period. In the uppermost sequence... Fig. 5 both zero states 0 Z and 7 Z is used. Only one switching operation occurs during each of the three state transitions. Each of the three bridge branches switches once per sampling interval. The state sequence over a complete fundamental oscillation period corresponds to the standard variant of rotary pointer modulation, as used in most applications. For this reason, it will be used as a reference for the subsequent pulse patterns. The number of switching operations per clock cycle is set to 100%.

[0036] According to the state of the art as described in Felix Jenni and Dieter Wüest's "Control Methods for Self-Commutated Converters," sequence 2 has two shutdown processes per sampling period. This sequence uses only one zero state. 0 Z in the left and 7 Z in the right-hand sequence). Compared to the standard variant, this results in 67% of the switching operations per sampling period. Therefore, for the same average switching frequency, the sampling period can be shortened by 33%, or the clock frequency increased by 50%. This sequence represents a limiting case of the first, in which the duration of one zero state is zero. The absence of a zero state means that one branch of the converter does not switch during the considered sampling period.

[0037] Therefore, different arrangements of the individual hands or times are possible. Either one zero hand or two zero hands are used.

[0038] According to the present invention, a sequence of space vectors, or synonymously, voltage vectors, is adapted. An additional time is introduced to freely shift the sequence and thereby synchronize the asynchronous pulse-width modulation. This additional time is calculated online from a flow trajectory FT.

[0039] According to the present invention, as in space vector modulation according to the prior art, a voltage vector with the associated times T1 and T2 is calculated. The distribution of these times T1, T2 (the sequence) is adjusted. According to the prior art, there are always two active voltage vectors with their times T1 and T2. These two times are then further subdivided, as is known in standard rhyme vector modulation, which is also evident from the Fig. 5 is evident.

[0040] Fig. Figure 6 is a schematic view of a sequence of stress vectors with the associated times T1 to T6.

[0041] The hatching indicates which time periods belong together. In this example, no zero vectors are shown, but rather the two active vectors 1 and 2.

[0042] The following applies: t1=T1+T3+T5;t2=T2+T4+T6 T2=T5;T3=T4 (with QWS−quarter-wave symmetry)

[0043] T1 is the so-called start time until the sequence begins. T6 is the so-called remaining time, which is determined as follows: T6=T1−T2−T3−T4−T5

[0044] The start time T1 is determined from the flow trajectory FT as shown below.

[0045] Depending on the section, several or no switching operations may take place.

[0046] Fig. Figure 7 is a schematic representation of the division of a target voltage vector SSV. In standard space vector modulation, the target voltage vector SSV is divided into two active space vectors (SVPWM, DPWM, etc.). According to the present invention, the target voltage vector SSV is divided into a larger number of active space vectors. With QWS (quarter-wave symmetry), the ratios of the times T2 = T5; T3 = T4 are fixed.

[0047] Fig. Figure 8 is a schematic diagram of a river trajectory FT with the rivers ψ α and ψ β . Fig. Figure 9 is a schematic diagram showing the time profiles of the midpoint voltages U, V, W of three half-bridges of a power converter. Fig. Figure 10 is a schematic detail view of a calculation period RP or PWM period of the midpoint voltage V. Shown are the start time T1, the times T2, T3, T4 and T5, and the remaining time T6, which is determined as described above. A comparison of the Fig. 7, Fig. 8 and Fig. Figure 9 shows that the proportions of the flow trajectory FT correspond to the proportions of the PWM period.

[0048] Fig. Figure 11 is a schematic diagram of a flux trajectory FT and the time course of the midpoint voltages U, V, W of three half-bridges of a power converter to illustrate an example simulation.

[0049] Fig. Figure 12 is a schematic diagram of a flux trajectory FT and the time course of the midpoint voltages U, V, W of three half-bridges of a power converter to illustrate an example simulation with seven pulses, where a long start time T1 is provided.

[0050] Fig. Figure 13 is a schematic diagram of a flux trajectory FT and the time course of the midpoint voltages U, V, W of three half-bridges of a power converter to illustrate an example simulation with seven pulses, where a long residual time T6 is provided.

[0051] Fig. Figure 14 is a schematic diagram of a flux trajectory FT and the time course of the midpoint voltages U, V, W of three half-bridges of a power converter. The time t = t1 + t2 does not have to coincide with the PWM period or calculation period. This allows the dynamic response to a control error to be adjusted. Correction can thus be spread over a longer period. For example, a voltage vector with a period of 400 µs is calculated, but recalculated every 100 µs. Therefore, if a control deviation occurs in the flux vector, the error is distributed over the voltage time surface of 400 µs. This results in a less aggressive response to noise.

[0052] To clarify the different terms, it should be noted again that a space vector usually refers to a stress vector, a flux vector to a vector in flux, and a stress vector to a vector in stress.

[0053] The following describes one way to calculate the start time t1.

[0054] A control unit calculates a target voltage. vαβ*(k) with a deadbeat rule formation rule to track the set flow trajectory FT as shown in equation (8): vαβ*(k)=ψαβ*(k)−ψ⌢αβ(k)Ts+Rsiαβ(k) with ψ - river, ψαβ*(k) - ? ψ̂ αβ (k) - ? T s - Sampling time R s i αβ (k) - ?

[0055] First, the start time T1 and time T2 are calculated using Uab: T1=U*Udc3sin(π3−ϕUαβ*) T2=U*Udc3sin(ϕUαβ*)

[0056] Then, using UabC, the start time T1C and the time T2C are calculated: T1C=UC*Udc3sin(π3−ϕUαβC*) T2C=UC*Udc3sin(ϕUαβC*)

[0057] Then the entire sequence is assembled. ti=[T1cT2cT1−T1cT2−T2c]

[0058] With the associated stress vectors SV=[V1V2V1V2]

[0059] The online calculation of the flow trajectory FT is performed as follows:

[0060] The basis for controlling the flow weakening and for the simple calculation of the flow trajectory FT is the maximum length ψ R of the flux hexagon. This corresponds to the flux magnitude or volt-seconds that can be supplied to the electrical machine controlled by the converter, for example, a drive motor of an electrically powered vehicle, in particular a passenger car, a commercial vehicle, or a bus, while exactly one complete rotation is performed at this specific speed. Since the sampling time T s cancels itself out, the maximum length ψ Rof the flux hexagon as a function of the DC bus voltage U dc and the electric angular velocity ω e calculated as shown in equation (9): ψR=2π6(23VdcTs)ωeTs=2πVdc9ωe

[0061] Based on this maximum flow ψ R Different flow trajectories FT can be calculated. Since the flow trajectory FT follows an angle of π3 or repeated at 60°, simplifications can be made. The rest Θαβπ3* of the angle for π3 is determined using a modulo function as shown in equation (10): Θαβπ3*=Rem(mod(Θαβ*,π3))

[0062] To increase computational efficiency, the calculations for sine and cosine are performed only once and then reused. a=sin(Θαβπ3*) b=cos(Θαβπ3*)⋅3

[0063] By applying the trigonometric laws and after some rearrangements, the magnitude of the hexagonal flow trajectory FT (6-step - 6-step) can be calculated for each target flow angle as shown in equation (13): |ψmag6step(Θαβπ3*)|=ψR⋅3(a+b)

[0064] This hexagonal flow trajectory FT can be extended to an 18-corner flow trajectory FT with very simple modifications (optimized pulse pattern OPP with pulse count q=3). For faster computation on the control unit, intermediate results can be calculated and stored in variables, which are named c, d, and e here: c=ψR⋅3⋅Kf d=c(b−a) e=ca

[0065] The magnitude of the 18-sided flow trajectory FT (OPP with q=3) can then be determined with a simple minimum function for each target flow angle: |ψmag3q(Θαβπ3*)|=min(d,e|ψmag6step(Θαβπ3*)|)

[0066] Furthermore, modifications to the space vector pulse width modulation (SVPWM) can be performed to achieve an optimized pulse pattern OPP with a pulse count of q=3.

[0067] The calculated flow magnitudes can be used to distinguish between a 6-step section and a pulse-width modulation section as follows. If the target flow magnitude is less than the maximum, then the variable inhex becomes active: inhex=|ψmag3q(Θαβπ3*)|<|ψmag6step(Θαβπ3*)|

[0068] Furthermore, a change in the flow segment is detected using the modulo operator: LR=[mod(Θαβ*,π3)≠0]

[0069] These two variables allow for simple case distinctions: 6-step processing if LR==0 and inhex==0, and a complete sequence if LR==1 and inhex==0. If LR==0 and inhex==1, the pulse-width modulation sequence is partially applied and completed in the next sampling task. In this case, a bit FS is activated, indicating that the sequence is to be completed.

[0070] Fig. Figure 15 schematically shows a flowchart that represents the decision process for carrying out sequences based on the conditions of the flow trajectory FT and the two parameters LR and inhex.

[0071] If the flux lies within the maximum possible circular flux, with the maximum occurring at a modulation degree of 1.21, a normal space vector modulation is applied; otherwise, the modifications are carried out.

[0072] The modifications of the standard space vector pulse-width modulation are described in detail below. Equations (20) to (23) for the standard space vector pulse-width modulation are: T1=U*Udc3sin(π3−ϕUαβ*) T2=U*Udc3 sin(ϕUαβ*) ti=[T04T12T22T02T22T12T04] SV=[V7V1V2V8V2V1V7]

[0073] In each voltage sector, V1 and V2 represent the two active space vectors, with T1 and T2 being their corresponding durations. V7 and V8 denote the two zero space vectors with their duration T0, which, in standard space vector pulse-width modulation, are evenly distributed between the beginning and middle of the pulse-width modulation cycle.

[0074] This sequence of space vectors is typically modified for optimization in asynchronous pulse-width modulation. For example, in discontinuous pulse-width modulation (DPWM), only a zero-voltage vector is used. This principle is applied here in a similar way for a synchronous optimized pulse pattern (OPP). 1) The sequence of the space pointers is modified in accordance with the specific optimized pulse pattern (OPP). 2) Based on the flow trajectory FT, the logic requires a pulse width modulation section or a 6-step section. 3) Calculations of the timing ensure that the initiation of the PWM sequence is precisely adjusted to the correct time within a single sampling period.

[0075] For an optimized pulse pattern (OPP) with q=3, the sequence of space pointers looks like this: SV=[V1V2V1V2]

[0076] Unlike asynchronous PWM, the timing is not constantly adjusted in each sampling period. The point of the first switch, or the time until the start of the PWM segment, varies. To calculate this switching point, an additional flux setpoint C is introduced, along with an additional virtual voltage vector U. αβC . Fig. Figure 16 is a schematic diagram of a detail of an 18-corner flow trajectory FT at a corner (PWM section), where the voltage vector U αβ , the flow trajectory FT from a point ψ αβ (k) up to a point ψαβ*(k+1) via the additional flow trajectory point C, and the stress vector U αβC The following are shown: a flow trajectory FT6S for 6-step and a flow trajectory FTQ3 for pulse count q=3.

[0077] By applying the standard space vector equations (20) and (21), the times T are determined. 1Cand T 2C calculated and the time sequence vector results from equation (25). ti=[T1CT2CT1−T2CT2−T2C]

[0078] As shown in the flowchart in Fig. As shown in Figure 15, a sequence is completed in the next sampling task if it cannot be fully executed in a single sampling task. For example, in the Fig. Case 16 of the stress vector U shown αβ The PWM sequence and the subsequent voltage vector terminate the PWM sequence.

[0079] Fig. Figure 17 is a schematic diagram of an exemplary PWM synchronization based on a flow trajectory FT. The position of the flow vector FVP and the position of the d-axis DAP are shown.

[0080] In period P1, a constant voltage vector is required to follow the flux trajectory FT, which means that no switching event occurs in this sampling period P1. In the subsequent sampling period P2, the PWM sequence starts at the very end and the FS bit is activated. In period P3, the flux trajectory FT is extended with the additional flux setpoint C, and the sequence is completed.

[0081] The ratio between the length of a PWM sequence and the sampling time Ts decreases with increasing speed, resulting in more complete sequences, but an identical pulse pattern with a higher fundamental frequency.

[0082] Simulation and experimentation have shown that DBFC (Deadbeat Flux Vector Control) is a single control rule that enables control with asynchronous space vector pulse-width modulation (SVPWM), synchronous optimized pulse patterns (OPP) with a pulse count of q=3, and six-step control. No switching or change to another control rule is required.

[0083] DBFC enables the integration of OPP (q=3) for high modulation indices with simple SVPWM adjustments. Unlike many other control rules, it avoids complex calculations and large look-up tables (LUTs) for the pulse pattern. A particularly fast computation or sampling time of < 100 µs is not required.

[0084] In this description, the terms zero phasor, zero vector, zero space phasor and zero voltage phasor are used synonymously and always refer to the same thing. Reference symbol list 1 active vector 2 active vectors a axis b axis c axis C River trajectory point DAP Position of the d-axis FT flow trajectory FTQ3 Flow trajectory for pulse count q=3 FT6S flow trajectory for 6-step FVP position of the flow vector m fundamental vibration P1, P2, P3, P4, P5, P6 period RP calculation period S Magnitude S1, S2, S6 Sector SSV target voltage vector Space pointer t time t0, t1, t2, t7 Time t1 Start time, time T, T Tast sampling period T1, T2, T3, T4, T5, T6 Time T0 duration T1 start time T6 remaining time u a , u b and u c Sub-pointer u pointer U, V, W Midpoint voltage U αβ Stress vector U αβC Stress vector u Str Voltage indicator 0 Z, 7 Z Zero state 0 Z, 1 Z, 2 Z, 7 Z condition ψ α , Ψ β Flow ψ αβ (k),ω angular frequency

Claims

[1] Method for synchronous space vector modulation with asynchronous sampling at a power converter for operating an electric machine, wherein a sequence of space pointers is adjusted by introducing an additional time to freely shift the sequence, wherein this additional time is calculated online from a flow trajectory (FT), where a control unit sets a target voltage vαβ*(k) calculated using a deadbeat rule formation rule to track the set flow trajectory (FT) and where the magnitude of the flow trajectory (FT) is determined using a minimum function for each target flow angle, characterized by , that Then, when a target flow magnitude is less than a maximum, a variable inhex is activated, a change in the flow section is detected with the modulo operator and stored in a variable LR, then, when both variables inhex and LR are zero, 6-step PWM is performed, then, when the variable LR is 1 and the variable inhex is zero, a complete sequence is performed, then, when the variable LR is zero and the variable inhex is 1, the pulse width modulation sequence is partially applied and completed in the next sampling task. [2] Method according to claim 1, characterized by , that as additional time a start time (T1) until the start of the sequence in a computation period (RP) or PWM period is used, or the start time (T1) is lengthened or shortened by the additional time, whereby a remaining time (T6) is shortened or lengthened accordingly. [3] Method according to claim 2, characterized by that the proportions of the flow trajectory (FT) correspond to the proportions of the computation period (RP). [4] Method according to any one of the preceding claims, characterized by , that a maximum length (ψ R ) of a flux hexagon as a function of a DC bus voltage (U dc ) and an electric angular velocity (ω e ) is calculated, based on this maximum length (ψ R ) at least one flow trajectory (FT) is calculated. [5] Method according to any one of the preceding claims, characterized by , that to calculate a time or additional time (T1) until the first switching or until the start of a PWM section, an additional flux setpoint (C) together with an additional virtual voltage vector (U) αβC ) introduced and the time or additional time (T1) by the equation T1 = T1=U*Udc3 sin(π3−ϕUαβ*) is calculated. [6] Vehicle with at least one electric machine for propelling the vehicle, at least one power converter for operating the electric machine and a control unit for controlling and / or regulating the power converter, characterized by that the control unit is configured to carry out the method according to one of the preceding claims.

Images