Fault-tolerant predictive control method for dual three-phase permanent magnet motor based on virtual vector and common-mode rejection

Abstract

The invention relates to a dual three-phase permanent magnet motor fault-tolerant prediction control method based on virtual vectors and common-mode rejection, and belongs to the technical field of permanent magnet synchronous motor control. In order to solve the problems of fundamental frequency offset and high harmonic content of common-mode voltage after a T-type three-level dual three-phase electric drive system has a single-phase open-circuit fault, a symmetric prediction current model is established by constructing a reduced-order decoupling matrix, a zero-harmonic voltage virtual vector set is generated, and the vector set is alternately input in combination with real-time common-mode voltage offset detection, so that the zero-harmonic voltage is obtained. And a weight-free series-parallel multi-objective optimization cost function is adopted to carry out vector preferential optimization. According to the method, the abc phase common-mode voltage is reduced to Udc / 6 in a high-speed interval, the full-speed domain of the def phase is lower than Udc / 8, the current distortion rate is smaller than 10%, meanwhile, neutral-point potential balance and low switching frequency are considered, and the fault-tolerant capability and the operation efficiency of the system are remarkably improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of permanent magnet synchronous motor control, and relates to a fault-tolerant predictive control method for a dual three-phase permanent magnet motor based on virtual vectors and common-mode suppression. Background Art

[0002] The all-electric technology that uses electric propulsion to replace traditional internal combustion power is gradually being applied to traditional mechanical equipment such as ships and automobiles. This technology requires that the electric drive system has the fault-tolerant ability to continue operating during the period from the occurrence of a fault to shutdown for maintenance, and needs to maintain low common-mode electromagnetic interference and high system efficiency before and after the fault occurs. In this context, due to its high reliability and strong common-mode voltage regulation ability, the T-type three-level dual three-phase permanent magnet electric drive system has attracted much attention in the field of all-electric technology. Given that more than 70% of the faults in multi-phase electric drive systems can be attributed to single-phase open-circuit faults, the research on open-phase operation is of great significance for improving system performance.

[0003] Existing research mainly focuses on three-phase inverters under normal operating conditions, while the topological structure and vector distribution of the T-type three-level dual three-phase electric drive system after a fault are significantly different from them. In view of this, a fault-tolerant predictive current control method with low common-mode voltage based on virtual vectors is proposed to achieve the comprehensive suppression of common-mode voltage and harmonics. Considering the multi-objective optimization requirements of capacitor midpoint potential balance and low switching frequency, a non-weighted multi-objective optimization scheme is also designed. Summary of the Invention

[0004] In view of this, the purpose of the present invention is to provide a fault-tolerant predictive control method for a dual three-phase permanent magnet motor based on virtual vectors and common-mode suppression. Aiming at the problems of high fundamental frequency bias and high harmonic content of the common-mode voltage caused by single-phase open-circuit faults in a three-level feed-type dual three-phase electric drive system, a comprehensive suppression method for common-mode voltage and harmonic current based on virtual vectors is proposed.

[0005] This method first derives a reduced-order decoupling matrix that matches the motor control degrees of freedom based on the principle of magnetic motive force conservation and the current constraints of the fault topology, and establishes a symmetric predictive current model for an asymmetric open-phase dual three-phase motor. Secondly, a virtual vector set with low common-mode voltage that synthesizes zero harmonic voltage components through variable ratio double vectors is obtained, and the large-amplitude vectors with low equivalent harmonic voltage components are selected as alternative vectors to achieve low harmonic operation under high bus utilization. Then, the vector sets are grouped and alternately input according to the real-time bias amplitude of the common-mode voltage to perform peak clipping and valley filling on the bias. In addition, considering the low switching frequency and the requirement of capacitor midpoint potential balance, a non-weighted coefficient series-parallel hybrid multi-objective optimization cost function is proposed, and an integral sorting system is introduced to select the optimal vector.

[0006] After adopting the proposed method, the common-mode voltages of the abc and def phases can be suppressed to Udc / 6 and U dc / 8 or less (U dc is the busbar voltage), and the phase current distortion rate is less than 10%.

[0007] To achieve the above object, the present invention provides the following technical solutions:

[0008] A fault-tolerant predictive control method for a dual three-phase permanent magnet motor based on virtual vectors and common-mode suppression, comprising the following steps:

[0009] S1: According to the topological constraints after a single-phase open-circuit fault, construct a reduced-order decoupling matrix, map the control degrees of freedom of the dual three-phase motor to the αβ stationary coordinate system, and establish a symmetric predictive current model in the asymmetric open-phase state;

[0010] S2: Based on the reduced-order decoupling matrix, generate a set of virtual vectors with multiple zero-harmonic voltage components. The set of virtual vectors is synthesized by variable-ratio dual vectors and satisfies that the equivalent harmonic voltage component is lower than a preset threshold;

[0011] S3: Real-time detect the common-mode voltage offset of the open-phase winding, and alternately select different sets of virtual vectors to be input into the control according to the amplitude range of the offset to suppress the common-mode voltage and harmonics period by period;

[0012] S4: Design a series-parallel hybrid multi-objective optimization cost function, perform multi-objective optimization of the dq-axis current tracking error, capacitor midpoint potential balance, and switching frequency with unweighted priorities, and select the optimal virtual vector through an integral sorting algorithm;

[0013] S5: According to the optimal virtual vector, output the switching signal for the next control period to achieve torque control and common-mode voltage suppression under fault-tolerant operation.

[0014] Further, in the S1, the construction of the reduced-order decoupling matrix is based on the principle of magnetic motive force conservation and the current constraints after the fault, and maps the five healthy-phase currents of the dual three-phase motor to the αβ sub-plane and the z harmonic plane, where the αβ sub-plane is used for torque and flux linkage control, and the z plane is used for harmonic suppression.

[0015] Further, in the S2, the generation of the set of virtual vectors includes:

[0016] S21: Define the switching states of two basic voltage vectors and synthesize virtual vectors proportionally;

[0017] S22: With the goal that the average harmonic voltage component of the virtual vector is zero, screen the synthesis paths that meet the equivalent harmonic voltage threshold conditions;

[0018] S23: Group the virtual vector sets into VV1-1, VV1-2, VV1-3 and VV2 according to the common-mode voltage characteristics of the open-phase winding, where the grouping basis is the combination mode of the switching functions.

[0019] Furthermore, in the above-mentioned S3, the logic of alternately selecting the virtual vector sets is implemented by a hysteresis controller, specifically including:

[0020] S31: When the common-mode voltage offset exceeds the positive threshold, select the first group of virtual vector sets;

[0021] S32: When the common-mode voltage offset is lower than the negative threshold, select the second group of virtual vector sets;

[0022] S33: When the common-mode voltage offset is within the threshold range, select the third group of virtual vector sets.

[0023] Furthermore, in the above-mentioned S4, the series-parallel hybrid multi-objective optimization cost function includes:

[0024] S41: The first parallel branch, successively connecting in series the dq-axis current tracking error cost function and the capacitor midpoint potential balance cost function;

[0025] S42: The second parallel branch, successively connecting in series the dq-axis current tracking error cost function and the switching frequency cost function;

[0026] S43: Calculate the vector intersection of the two branches, and determine the final optimal vector based on the integral sorting algorithm.

[0027] Furthermore, the specific steps of the integral sorting algorithm include:

[0028] (a) Sort the candidate vectors selected from each branch in ascending order of the cost function value, and assign decreasing integral values;

[0029] (b) Accumulate the integral values of the vectors within the intersection, and select the vector with the highest integral;

[0030] (c) If the intersection is empty, select the optimal vector of the dq-axis current tracking error cost function.

[0031] Furthermore, the method further includes S6: Compensate for the control delay through a two-step prediction method, specifically including:

[0032] S61: Predict the current state of the next cycle using the measured values at the current moment;

[0033] S62: Combine the optimal virtual vector to correct the delay error in the prediction model.

[0034] Furthermore, in the above-mentioned S22, the equivalent harmonic voltage threshold is 10% of the bus voltage.

[0035] Further, in S3, the suppression target of the common-mode voltage is as follows:

[0036] (a) The common-mode voltages of the abc phases are suppressed to zero in the low-speed range and their amplitudes are reduced to 1 / 6 of the bus voltage in the high-speed range;

[0037] (b) The common-mode voltages of the def phases are lower than 1 / 8 of the bus voltage within the full speed range.

[0038] Further, the method is applied to a T-type three-level dual three-phase permanent magnet electric drive system, and the single-phase open-circuit fault is the open-circuit fault of phase F.

[0039] The beneficial effects of the present invention are as follows: By reconstructing the reduced-order decoupling matrix, the motor model after the fault still has symmetry similar to that in the normal state, and the voltage vector is not affected by the midpoint potential fluctuation after the fault. At the same time, the current control dimension is consistent with the three-dimensional degrees of freedom of the motor, thereby realizing the effective decoupling control of torque current, flux current, and harmonic current. By constructing multiple sets of zero-harmonic voltage virtual vectors, and observing the common-mode voltage offset in real time, and alternately applying the vector sets, the harmonic and common-mode voltages are suppressed cycle by cycle. In addition, by adopting the series-parallel cost function topology and combining the integral vector selection algorithm, multi-objective optimization without weights and importance grading is realized. Specifically, the common-mode voltages of the abc phases are successfully suppressed to zero in the low-speed range and their amplitudes are reduced to Udc / 6 in the high-speed range; the common-mode voltages of the def phases are lower than Udc / 8 within the full speed range, and the current distortion rate is reduced to less than 10%.

[0040] Other advantages, objectives, and features of the present invention will be described to some extent in the subsequent description, and to some extent, will be obvious to those skilled in the art based on the study of the following text, or can be taught from the practice of the present invention. The objectives and other advantages of the present invention can be realized and obtained through the following description. Description of the Drawings

[0041] In order to make the objectives, technical solutions, and advantages of the present invention clearer, the present invention will be described in detail preferably with reference to the accompanying drawings, where:

[0042] Figure 1 For the virtual vector set VV1 and its equivalent harmonic voltage u ez ;

[0043] Figure 2 For the virtual vector set VV2 and its equivalent harmonic voltage u ez ;

[0044] Figure 3 For the virtual vector set VV2 and its u ez ;

[0045] Figure 4 is the cost function topology of the series-parallel hybrid structure;

[0046] Figure 5 are the steady-state experimental waveforms when operating at 200 rpm and 500 rpm respectively under a load of 4 Nm;

[0047] Figure 6 is the dynamic response waveform during the process of the load stepping from no-load to the rated 4 Nm load at 500 rpm;

[0048] Figure 7 is the dynamic response waveform during the process of stepping from stationary to 500 rpm under a load of 4 Nm. Specific embodiments

[0049] The following uses specific specific examples to illustrate the implementation manners of the present invention. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific implementation manners. Various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that the drawings provided in the following embodiments only illustrate the basic concept of the present invention in a schematic manner. Without conflict, the following embodiments and the features in the embodiments can be combined with each other.

[0050] Among them, the drawings are only for illustrative purposes, showing only schematic diagrams, rather than physical diagrams, and should not be construed as a limitation on the present invention; in order to better illustrate the embodiments of the present invention, some components in the drawings will be omitted, enlarged or reduced, and do not represent the dimensions of actual products; for those skilled in the art, it is understandable that some well-known structures and their descriptions in the drawings may be omitted.

[0051] In the drawings of the embodiments of the present invention, the same or similar reference numerals correspond to the same or similar components; in the description of the present invention, it should be understood that if there are terms such as "upper", "lower", "left", "right", "front", "rear", etc. indicating the orientation or positional relationship, they are based on the orientation or positional relationship shown in the drawings. It is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation. Therefore, the terms describing the positional relationship in the drawings are only for illustrative purposes and should not be construed as a limitation on the present invention. For those of ordinary skill in the art, the specific meanings of the above terms can be understood according to specific circumstances.

[0052] First, a reduced-order decoupling matrix based on the actual topological constraints was designed, and multiple sets of zero-harmonic voltage virtual vector sets with low common-mode voltage were constructed. By online observing the common-mode voltage bias signal, different vector sets were alternately input to suppress the common-mode voltage and harmonics in real time. In addition, a series-parallel topological cost function that does not distinguish the priority of secondary objectives was adopted to achieve unweighted multi-object tracking. The proposed algorithm realizes the coordinated control of the common-mode voltage and torque, effectively suppresses the harmonic current, and reduces the switching frequency. Taking the open circuit of phase F as an example, the specific implementation steps are described as follows:

[0053] 1. Virtual vector construction

[0054] Assume that the open phase is phase F. After the open-circuit fault occurs, the control freedom of the dual three-phase motor decreases, and the remaining five healthy phases operate asymmetrically. To achieve fault-tolerant operation of the system, the reduced-order decoupling orthogonal matrix T mapping from the stator coordinate system to the two-phase αβ stationary coordinate system is reconstructed 5s2s , and there is the following mapping equation:

[0055]

[0056] In the formula, f represents current, voltage, or torque. The αβ sub-plane after being mapped by equation (2) is the control plane of torque and flux linkage; the z-plane is the monopole plane of harmonic mapping; the o1o2 sub-plane is the generalized zero-sequence plane, and the current in this plane is uncontrollable.

[0057] Using the Park transformation, the mathematical model in the rotating coordinate system is obtained. After discretization using the forward Euler method and then using the two-step prediction method for delay compensation, the current prediction model is obtained:

[0058]

[0059] In the formula, u dqs , i dqs , L dqs are the voltage, current, and inductance components of the dq axes respectively, R s is the phase resistance, ψ f is the permanent magnet flux linkage. T s is the calculation period, and the superscripts "k + 1" and "k + 2" represent the values of the variables in the next period and the next next period respectively.

[0060] Based on equation (2), the calculation expression of the voltage component under the αβz plane components is obtained

[0061]

[0062] In the formula, S x (x = a~e) represents the three switching states P, O, N of the healthy phase bridge arm output, and are represented by S x = 1, 0, -1 respectively. Udc is the bus voltage.

[0063] Define the switching functions of two basic voltage vectors u1 and u2 as S1 = [S a1 S b1 S c1 S d1 S e1 and S2 = [S a2 S b2 S c2 S d2 S e2 . Synthesize the virtual vector by mixing u1 and u2 in the ratio of a:1-a, and take the average harmonic voltage component of the virtual vector equal to zero as the target to obtain the following constraints:

[0064]

[0065] To reduce the switching harmonics, the constraint in Equation (6) is proposed to screen the synthesis path of the smaller equivalent harmonic voltage u ez :

[0066]

[0067] where u threshold is the threshold of the equivalent harmonic voltage, and here 0.1U dc is taken.

[0068] In addition, the common-mode voltage suppression strategy after SOPF also needs to be redesigned. For the common-mode suppression strategy of the abc phases, it should be consistent with the normal operating conditions. Usually, it is required that the sub-common-mode u n1o of the abc phases ≤ 1 / 6U dc , and the switching functions of the abc phases satisfy:

[0069]

[0070] Combining Equations (5)-(7), there are Figure 1 and Figure 2 in the two sets of virtual vector sets.

[0071] The synthesis paths of the alternative virtual vectors are shown in Tables 1-3.

[0072] Table 1 Synthesis path of VV1-1

[0073]

[0074] Table 2 Synthesis paths of VV1-2 and VV1-3

[0075]

[0076] Note: S d +Se =-1 represents the virtual vector set VV1-2, S d +S e =1 represents the virtual vector set VV1-3. VV1-2 and VV1-3 at the same position on the αβ sub-plane only differ in the de-phase potential.

[0077] Table 3 Synthesis Path of VV2

[0078]

[0079] 2. Active Common-Mode Voltage Suppression

[0080] The common-mode voltage u of the winding in the def phase where the phase is missing n2o has a sinusoidal offset, and it is necessary to alternately use the vector sets proposed in 1 to actively suppress u n2o . The ideal u after the vector is applied n2o and the input logic are as Figure 3 shown.

[0081] Convert Figure 3 the vector input logic in into a mathematical formula, and there is

[0082]

[0083] E f is a sinusoidal offset of u n2o , E fMax is the maximum value of Ef, and the flag bit P is output by determining the instantaneous value range of E f . When P = 1, 0, -1, the vector sets VV1-2, VV1-3, and VV1-1 are selected respectively. The above operating logic is actually implemented by a hysteresis controller.

[0084] 3. Unweighted Multi-Objective Optimization Design

[0085] The control objectives of the system include i dqs tracking, harmonic current suppression, neutral point potential balance of the bus capacitor, and switching frequency reduction. Among them, the objective of harmonic current suppression has been achieved through the virtual vector set of 0 harmonic voltage. After introducing the two-step prediction method for delay compensation, the sub-cost functions of the remaining objectives are designed as

[0086] (1) dq-axis current tracking cost sub-function

[0087]

[0088] (2) Capacitor neutral point potential balance cost sub-function

[0089]

[0090] Among them, the predicted value The calculation formula is

[0091]

[0092] (3) Sub-cost function of switching frequency

[0093]

[0094] f x represents the number of switching times of arm x in the next cycle.

[0095] The above optimization objective is Figure 4 vector traversed using the cost function topology of

[0096] First, the cost function g dq selects n1 vectors, and then further selects through a parallel topology structure. Finally, two sets containing n3 vectors are output from the two parallel branches. After taking the intersection of these two sets, the remaining vectors are scored according to their sorting in the set: the vector at the front is scored n3 points, and the scores of the remaining vectors decrease successively. The score of the same vector is equal to the sum of the scores of the two sets, and finally the vector with the highest score is output. If the intersection is empty, then the optimal vector selected by g dq is output. If the vectors with the highest scores are not unique, then the vector with the highest sorting in the set selected by g dq is selected.

[0097] 4. Algorithm steps

[0098] The control flow of the algorithm can be briefly described as follows:

[0099] 1) One-beat-ahead prediction: Measure the stator voltage, stator current, and rotor speed at time k, substitute the optimal vector at time k + 1, and then estimate the dq-axis currents at time k + 1 according to the reduced-order decoupling mathematical model in Equation (3).

[0100] 2) Vector set alternating control: Substitute the F back electromotive force E f directly into the designed hysteresis logic for u n2o to select the alternative vector set at time k + 2.

[0101] 3) Vector optimization based on the unweighted sequence cost function: Substitute the speed loop PI output, the alternative vector set, the reference current, and the reference capacitor midpoint potential into the sequence cost function. First, g dq calculates 7 sub-vector sets with smaller calculated values. Then, respectively find the optimal 5-vector sets of the series sequences of g c , g f and the optimal 5-vector sets of the series sequences of g f , g cOptimal set of 5 vectors for the series sequence. After taking the intersection of two sets of vectors, the vector with the highest output integral is taken as the optimal vector VV k+2 .

[0102] The fault-tolerant operation verification experiments under rated steady-state and dynamic conditions are carried out on the T-type three-level dual three-phase motor pair test platform with a rated load (4 Nm). Finally, the experimental effect diagrams under the following various conditions are obtained, as Figures 5 to 7 shown, indicating that the proposed method can effectively achieve the operation after the open-phase fault and the optimization of its sub-goals. Finally, the common-mode voltage of the abc phase is successfully suppressed to zero in the low-speed range, and the amplitude drops to U dc / 6 in the high-speed range. The common-mode voltage of the def phase is lower than U dc / 8 in the full speed range, and the current distortion rate drops below 10%.

[0103] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the purpose and scope of the present technical solution, and they should all be covered within the scope of the claims of the present invention.

Claims

1. A fault-tolerant predictive control method for a dual three-phase permanent magnet motor based on virtual vectors and common-mode suppression, characterized in that: It includes the following steps: S1: According to the topological constraints after single-phase open-circuit fault, construct a reduced-order decoupling matrix, map the control degrees of freedom of the dual three-phase motor to the αβ stationary coordinate system, and establish a symmetric predictive current model under the asymmetric open-phase state; S2: Based on the reduced-order decoupling matrix, generate multiple sets of virtual vector sets of zero harmonic voltage components. The virtual vector sets are synthesized by variable ratio dual vectors and satisfy that the equivalent harmonic voltage components are lower than a preset threshold; S3: Real-time detect the common-mode voltage offset of the open-phase winding. According to the amplitude range of the offset, alternately select different virtual vector sets to be input for control to suppress the common-mode voltage and harmonics per cycle; S4: Design a series-parallel hybrid multi-objective optimization cost function to perform multi-objective optimization of the dq-axis current tracking error, capacitor midpoint potential balance, and switching frequency without weighted priorities, and select the optimal virtual vector through an integral sorting algorithm; S5: According to the optimal virtual vector, output the switching signal of the next control cycle to achieve torque control and common-mode voltage suppression under fault-tolerant operation.

2. The fault-tolerant predictive control method for a dual-three-phase permanent magnet motor based on virtual vectors and common-mode suppression according to claim 1, wherein: In the S1, the construction of the reduced-order decoupling matrix is based on the principle of magnetic motive force conservation and the current constraints after the fault, and maps the five healthy-phase currents of the dual three-phase motor to the αβ sub-plane and the z harmonic plane. The αβ sub-plane is used for torque and flux linkage control, and the z plane is used for harmonic suppression.

3. The fault-tolerant predictive control method for a dual-three-phase permanent magnet motor based on virtual vectors and common-mode suppression according to claim 1, wherein: In the S2, the generation of the virtual vector set includes: S21: Define the switching states of two basic voltage vectors and synthesize virtual vectors proportionally; S22: With the goal of the average harmonic voltage component of the virtual vector being zero, screen the synthesis paths that meet the equivalent harmonic voltage threshold conditions; S23: According to the common-mode voltage characteristics of the open-phase winding, group the virtual vector sets into VV1-1, VV1-2, VV1-3, and VV2, where the grouping basis is the combination mode of the switching functions.

4. The fault-tolerant predictive control method for a dual three-phase permanent magnet motor based on virtual vectors and common-mode suppression according to claim 3, characterized in that: In the S3, the logic of alternately selecting the virtual vector set is implemented by a hysteresis controller, specifically including: S31: When the common-mode voltage offset exceeds the positive threshold, select the first group of virtual vector sets; S32: When the common-mode voltage offset is lower than the negative threshold, select the second group of virtual vector sets; S33: When the common-mode voltage offset is within the threshold range, select the third group of virtual vector sets.

5. The fault-tolerant predictive control method for a dual-three-phase permanent magnet motor based on virtual vectors and common-mode suppression according to claim 1, characterized in that: In the S4, the series-parallel hybrid multi-objective optimization cost function includes: S41: The first parallel branch, successively series-connected with the dq-axis current tracking error cost function and the capacitor midpoint potential balance cost function; S42: The second parallel branch, successively series-connected with the dq-axis current tracking error cost function and the switching frequency cost function; S43: Calculate the vector intersection of the two branches and determine the final optimal vector based on the integral sorting algorithm.

6. The fault-tolerant predictive control method for a dual three-phase permanent magnet motor based on virtual vectors and common-mode suppression according to claim 5, wherein: The specific steps of the integral sorting algorithm include: (a) Arrange the candidate vectors screened out by each branch in ascending order of the cost function value and assign decreasing integral values; (b) Accumulate the integral values of the vectors within the intersection and select the vector with the highest integral; (c) If the intersection is empty, select the optimal vector of the dq-axis current tracking error cost function.

7. The fault-tolerant predictive control method for a dual-three-phase permanent magnet motor based on virtual vectors and common-mode suppression according to claim 1, characterized in that: The method further includes S6: compensating for the control delay through a two-step prediction method, specifically including: S61: predicting the current state in the next cycle using the measured value at the current moment; S62: combining the optimal virtual vector to correct the delay error in the prediction model.

8. The fault-tolerant predictive control method for a dual-three-phase permanent magnet motor based on virtual vectors and common-mode suppression according to claim 3, wherein: In the S22, the equivalent harmonic voltage threshold is 10% of the bus voltage.

9. The fault-tolerant predictive control method for a dual-three-phase permanent magnet motor based on virtual vectors and common-mode suppression according to claim 1, characterized in that: In the S3, the suppression target of the common-mode voltage is: (a) The common-mode voltage of the abc phase is suppressed to zero in the low-speed range and the amplitude drops to 1 / 6 of the bus voltage in the high-speed range; (b) The common-mode voltage of the def phase is lower than 1 / 8 of the bus voltage within the full speed range.

10. The fault-tolerant predictive control method for a dual three-phase permanent magnet motor based on virtual vectors and common-mode suppression according to any one of claims 1 to 9, characterized in that: The method is applied to a T-type three-level dual three-phase permanent magnet electric drive system, and the single-phase open-circuit fault is the open-circuit fault of the F phase.

Images