Integrated diagnosis and fault tolerance method suitable for any-phase single switch tube fault in five-phase permanent magnet synchronous motor driving system

Abstract

The invention discloses an integrated diagnosis and fault tolerance method suitable for any-phase single switch tube faults in a five-phase permanent magnet synchronous motor driving system, and belongs to the technical field of motor driving and control. The method comprises the following steps: unifying 10 kinds of single switch tube faults at different positions in a five-phase inverter system into two types; the diagnosis of the fault of the upper switch tube and the fault of the lower switch tube is realized by analyzing the third harmonic part of the five-phase current when the single switch tube has the fault; reconstructing a novel basic voltage vector under the fault; determining the action time proportion of the voltage vector combination and synthesizing a virtual voltage vector; constructing a universal virtual voltage vector suitable for the overall upper pipe fault or the overall lower pipe fault; and corresponding vector action time is determined, and obtained switching signals are input into a control motor, so that fault-tolerant control of any-phase single-switch faults of the five-phase permanent magnet synchronous motor driving system is realized. The method is suitable for occasions with high requirements on the reliability of the motor.

Description

Technical Field

[0001] This invention relates to the field of motor drive and control technology, specifically to an integrated diagnostic and fault-tolerant method for single-phase switch failures in a five-phase permanent magnet synchronous motor drive system. Background Technology

[0002] Five-phase permanent magnet synchronous motors are widely used in high-reliability fields such as aerospace, electric vehicles, and precision machining due to their advantages of low torque ripple, high power density, and high reliability. Their drive systems are typically powered by multi-phase inverters, and the power switching transistors in these inverters are among the components with the highest failure rate. If a switching transistor fails and is not addressed promptly, it can lead to drastic fluctuations in motor torque, system shutdown, and even catastrophic consequences. Therefore, rapid and accurate fault diagnosis and fault-tolerant control of the drive system are crucial to ensuring the safe and stable operation of the entire system.

[0003] In fault diagnosis, existing methods, such as analyzing voltage and current residuals or constructing observers based on voltage and current models, often require diagnosing multiple fault states, making the diagnostic methods very complex and prone to misdiagnosis. In fault-tolerant control, mainstream technologies typically employ hardware methods (such as adding extra redundant bridge arms or inverter reconnection) or software algorithms (such as injecting harmonic currents or changing the transformation matrix). While these methods can maintain system operation after a fault, they still have significant limitations. Hardware methods significantly increase system size and additional cost, while most software fault-tolerant algorithms do not fully utilize the characteristics of single-switch faults, and injected harmonic currents can make the system unstable. Furthermore, existing fault-tolerant methods require different algorithms designed for different fault locations (such as the upper A-phase switch and the lower C-phase switch), resulting in complex control systems and a lack of a unified control strategy to handle all possible single-switch faults. In addition, traditional fault-tolerant control systems typically treat "fault diagnosis" and "fault-tolerant operation" as two independent directions, but in practical applications, these two modules need to be closely integrated into a unified solution. Summary of the Invention

[0004] To address the problems of complex diagnosis, non-universal fault tolerance methods, and non-integrated diagnosis and fault tolerance in existing methods for single-switch faults in five-phase permanent magnet synchronous motor drive systems, this invention proposes an integrated diagnosis and fault tolerance method applicable to single-switch faults in any phase of a five-phase permanent magnet synchronous motor drive system.

[0005] To achieve the technical objectives, the present invention adopts the following technical solution:

[0006] An integrated diagnostic and fault-tolerant method for single-phase switch faults in a five-phase permanent magnet synchronous motor drive system includes the following steps:

[0007] Step 1: Since traditional five-phase inverter fault diagnosis methods need to identify multiple fault states, and fault tolerance methods are often for faults of a specific switch in a specific phase and position, if the fault occurs in different phases and positions, fault tolerance methods need to be redesigned. In order to reduce the number of diagnostic states and enhance the versatility of fault tolerance methods, the 10 different single switch faults in the five-phase inverter system are unified into two types: overall upper switch fault and lower switch fault.

[0008] Step 2: Sample the five-phase current i of the five-phase motor using a current sensor. A i B i C i D i E By analyzing the third harmonic component of the five-phase current during a single-switch fault, A3 i B3 i C3 i D3 i E3 Diagnostic criteria are obtained from the positive and negative values ​​of the fault cycle, enabling the diagnosis of both overall and downstream switching transistor faults.

[0009] Step 3: When a single switch failure occurs in the five-phase permanent magnet synchronous motor inverter, the phase voltage of the inverter can be obtained according to the switching state at the time of the failure. The five-phase voltage is then transformed into α-β axis voltage and α3-β3 axis voltage, thereby reconstructing a new basic voltage vector under the fault.

[0010] Step 4: Select voltage vector combinations suitable for modulation from the new set of basic voltage vectors. Based on geometric principles, with the goal of minimizing the voltage components of the third harmonic subspace, determine the action time ratio of each voltage vector combination, and then synthesize a virtual voltage vector with low harmonic characteristics.

[0011] Step 5: Unify the virtual voltage vectors of five specific phase switch faults or lower switch faults in the inverter into a unified vector set, and construct a general virtual voltage vector suitable for overall switch faults or lower switch faults by finding the intersection area.

[0012] Step 6: Select a suitable general virtual voltage vector based on the spatial position of the reference voltage vector and establish the corresponding vector action time. Input the obtained switching signal into the inverter to control the motor, thereby realizing fault-tolerant control of any single-phase switch failure in the five-phase permanent magnet synchronous motor drive system.

[0013] Furthermore, the diagnostic method described in step 2 is as follows:

[0014] Step 2.1: Using the Clarke transform matrix and inverse Clarke transform matrix, according to the transformation flowchart, the third harmonic component i of the five-phase current is transformed. A3 i B3 i C3 i D3 i E3 From phase current i A i B i C i D i E Separate from the middle, satisfying the following relationship:

[0015] ;

[0016] Among them, i A i B i C i D i E i represents the five-phase current of the motor under normal operating conditions. A3 i B3 i C3 i D3 i E3 The third harmonic component of the phase current under normal operating conditions is represented by the current coefficient, which is related to the Clarke transform matrix and the inverse Clarke transform matrix. The matrix can be expressed as:

[0017] ,

[0018] ;

[0019] Among them, T Clarke The Clarke transformation matrix is ​​T-1, and the inverse Clarke transformation matrix is ​​TC, where α = 2 / 5π.

[0020] Since the third harmonic space of a five-phase motor does not participate in electromechanical energy conversion, the α3-β3 axis current under normal operating conditions should satisfy:

[0021] ;

[0022] Therefore, the third harmonic component of the phase current obtained by matrix transformation should also remain zero:

[0023] ;

[0024] Step 2.2, assuming the fault occurs at the switching transistor on phase A, the phase A current during the fault can be represented by segments of the upper half-cycle of the fault and the lower half-cycle of normal operation: ;

[0025] Where iFault Atotal is the phase A current when the switching transistor on phase A fails, M is the amplitude of the phase current, and θ e It is an electrical angle;

[0026] During the first half of the fault cycle, phase A current will be lost. At this time, the third harmonic component of the five-phase current will change and will no longer remain at 0.

[0027] ;

[0028] Where iFault A, iFault B, iFault C, iFault D, and iFault E are the five-phase currents when the switching transistor on phase A fails, and iFault A3, iFault B3, iFault C3, iFault D3, and iFault E3 are the third harmonic components of the phase current when the switching transistor on phase A fails.

[0029] At this time, during the fault cycle, the third harmonic component of the five-phase current will inevitably exhibit N values ​​greater than 0 and (5-N) values ​​less than 0. Since the five-phase motor structure is completely symmetrical, the positive and negative values ​​of the third harmonic component will have the same total number of positive and negative values ​​when the upper switch tube of phase A is faulty and the upper switch tubes of phases B, C, D, and E are faulty. The only difference is that the distribution of positive and negative values ​​changes, but the overall number remains the same. The upper switch tube fault and the lower switch tube fault of the inverter are reversible. The lower switch tube will exhibit a completely opposite state, that is, (5-N) values ​​greater than 0 and N values ​​less than 0. Based on the positive and negative value relationship, the diagnosis of the upper switch tube fault and the lower switch tube fault of the entire five-phase inverter can be realized.

[0030] Furthermore, the novel voltage vector reconstruction process described in step 3 is as follows:

[0031] A standard five-phase inverter can form 2 5 There are 32 switching states, corresponding to 32 basic voltage vectors. A "0" indicates the upper switch is on, and a "1" indicates the lower switch is on. Assuming the fault occurs at the upper switch of phase A, then the 16 switching states "0xxxx" (x=0&1) are used. The phase voltage U of the inverter at the time of the fault is obtained using the KCL and KVL laws for the affected switching states. AN U BN U CN U DN U EN Then it is transformed onto the α-β axis and the α3-β3 axis:

[0032] ;

[0033] Among them, U AN U BNU CN U DN U EN U is the phase voltage of the inverter. αβ and U αβ3 These are the α-β axis voltage and the α3-β3 axis voltage, respectively. After reconstructing the 16 voltage vectors affected by the fault, they are combined with the original unaffected voltage vectors to form a new type of voltage vector under single-switch fault.

[0034] Furthermore, the virtual voltage vector synthesis process described in step 4 is as follows:

[0035] Step 4.1: Due to the complex distribution of the new voltage vector at the time of the reconstructed fault, preprocessing is required. Taking the fault of the switching transistor on phase A as an example, after excluding the voltage vector corresponding to the small vector under normal operating conditions, the remaining voltage vector U... 16 U 25 U 29 U 24 , U8, U 28, U 30, U 12 , U4, U 14 U 15 , U6, U2, U7, U 23 , U3, U1, U 19 U 27 U 17 They can be combined in pairs to form 10 groups, namely [U 16 U 25 ], [U 29 U 24 ],[U8, U 28 ], [U 30 U 12 ],[U4, U 14 ], [U 15 [U6], [U2, U7], [U 23 [U1, U3], [U3, U3] 19 ], [U 27 U 17 This also corresponds to the large and medium vectors under normal operating conditions;

[0036] Step 4.2: To simplify the spatial distribution of voltage vectors, the voltage vectors from the 10 combinations will be synthesized into a virtual voltage vector according to a certain ratio. Simultaneously, to suppress the influence of the harmonic subspace, the third harmonic voltage after the virtual voltage vector synthesis is minimized; where [U4, U...] 14 ], [U 15In the three combinations [U6], [U2, U7], the voltage vector is not affected by the fault. After synthesis using the action time ratio of 0.618:0.382, the virtual voltage vector with the third harmonic voltage is 0.

[0037] Step 4.3: The remaining 7 combinations cannot be completely synthesized into zero because the two vectors within each combination are not on the same straight line. Therefore, the goal is to minimize the third harmonic voltage after synthesis. Geometrically, this is transformed into a mathematical problem of finding the shortest distance from the zero point to the third side of the triangle formed by the two vectors. Constructing perpendicular lines and solving the triangle determines the respective time ratios, thus synthesizing the virtual voltage vector with the minimum third harmonic voltage. Finally, these 10 combinations are organized into 10 virtual voltage vectors V1-V1 under switching transistor fault conditions. 10 .

[0038] Furthermore, the process for establishing the general virtual voltage vector described in step 5 is as follows:

[0039] Step 5.1, taking the overall switching transistor fault as an example, lists the virtual voltage vectors of the five specific phase switching transistor faults. The virtual voltage vectors of each phase fault are distributed in the fundamental frequency space with a 72-degree rotation relationship. Then, these five voltage vectors are unified into a whole vector set, which includes the 10 virtual voltage vectors V1-V1 of each phase fault. 10 The resulting subset;

[0040] Step 5.2, based on V1-V 10 Within the coverage area, the subsets can initially divide the space into 10 large sectors IX. Since each subset is axially symmetric, each large sector can be further divided into 2 smaller sectors, thus subdividing the spatial plane into 20 smaller sectors. ;

[0041] Step 5.3: Obtain the modulation region formed by the virtual voltage vectors of each phase fault. Find the intersection region of vector modulation in each small sector. The boundary of the intersection region is the prototype of the general virtual voltage vector. Considering the continuity of the control algorithm, the intersection region needs to meet the requirement of covering the entire plane 0-360°. Slightly adjust the phase of the general virtual voltage vector to establish the final general virtual voltage vector.

[0042] The present invention has the following beneficial effects:

[0043] 1. This invention proposes a unified diagnostic and fault-tolerant control strategy, integrating the "fault diagnosis" and "fault-tolerant operation" modules into a cohesive whole. Regardless of whether the fault occurs on the upper or lower phase, the system can switch to the same preset overall upper or lower phase switch fault-tolerant control algorithm after diagnosing the fault. This changes the complex situation in traditional methods that require pre-designing multiple different algorithms for different fault locations, greatly simplifying system design and software implementation, reducing algorithm complexity, and improving the versatility and maintainability of the control system.

[0044] 2. This invention extracts the third harmonic characteristics of the phase current during a fault. By combining the positive and negative values ​​with the symmetry of the five-phase motor and the reciprocity of the inverter, it can achieve overall diagnosis of the upper and lower switching transistors. This method greatly simplifies the complexity of the diagnostic algorithm, effectively avoids misdiagnosis, and enables rapid fault location, saving valuable time for subsequent fault-tolerant control.

[0045] 3. This invention adopts a pure software solution, which does not require adding hardware redundancy or changing the main circuit topology. By reconstructing a new voltage vector during a fault and conducting in-depth analysis, the virtual voltage vector synthesized using geometric meaning can greatly suppress the influence of the third harmonic space, achieving a high level of fault-tolerant operation. It is particularly in line with the strict requirements for power density and cost control in aerospace, electric vehicle and other fields. Attached Figure Description

[0046] Figure 1 A general, integrated diagnostic and fault-tolerant approach;

[0047] Figure 2 The process of extracting the third harmonic component of the phase current;

[0048] Figure 3 Simplified diagnostic flowchart;

[0049] Figure 4 : Reconstructed voltage vector distribution under single-switch fault (taking a switch fault on phase A as an example); (a) fundamental frequency space; (b) third harmonic frequency space;

[0050] Figure 5 Solution of virtual voltage vector action time ratio (using U8 and U8 as examples) 28 (For example)

[0051] Figure 6 : Virtual voltage vector distribution of the fault switch on phase A; (a) fundamental frequency space; (b) third harmonic frequency space;

[0052] Figure 7Virtual voltage vector distribution of fault switches on any phase; (a) Phase A; (b) Phase B; (c) Phase C; (d) Phase D; (e) Phase E; (f) Five-phase set;

[0053] Figure 8 : Division of large and small sectors;

[0054] Figure 9 The final modulation region of the universal virtual voltage vector;

[0055] Figure 10 : Block diagram of integrated diagnosis and fault-tolerant control for single-phase switching transistor faults in a five-phase permanent magnet synchronous motor drive system;

[0056] Figure 11 Overall diagnostic results of single-switch tube faults in a five-phase permanent magnet synchronous motor; (a) from the fault of the upper switch tube in phase A to the fault of the lower switch tube in phase B (experimental load is 2.6 Nm, speed is 200 r / min); (b) from the fault of the upper switch tube in phase A to the fault of the lower switch tube in phase C (experimental load is 2.6 Nm, speed is 400 r / min).

[0057] Figure 12 The current and torque waveforms of a five-phase permanent magnet synchronous motor system from normal to fault to fault tolerance (experimental load is 2.9 Nm, speed is 380 r / min); (a) fault of the switch on phase A; (b) fault of the switch on phase B; (c) fault of the switch on phase C;

[0058] Figure 13 The switching of the switching transistors on different phases of the five-phase permanent magnet synchronous motor system was tested (phase A-phase-B-phase-C phase, experimental load of 2.8 Nm, speed of 360 r / min). Detailed Implementation

[0059] The specific implementation methods and effects of this embodiment will be described in detail below with reference to the accompanying drawings.

[0060] From the appendix Figure 1 The general integrated diagnosis and fault tolerance concept shows that the diagnosis status has been reduced from the original 10 states of five-phase upper and lower switch tube failure to 2 states of overall upper switch tube failure and overall lower switch tube failure. Correspondingly, only 2 types of fault tolerance methods need to be prepared: one to deal with all five-phase upper switch tube failures and one to deal with all five-phase lower switch tube failures.

[0061] From the appendix Figure 2 The process of extracting the third harmonic of the phase current allows us to understand the specific functions of the Clarke transform matrix and the inverse Clarke transform matrix. First, the Clarke transform matrix is ​​used to extract the third harmonic of the five-phase current i. A i B i C iD i E The current is transformed into α-β axis currents and α3-β3 axis currents. The α-β axis currents are set to 0, and then transformed into the third harmonic component of the five-phase current through the inverse Clarke transformation matrix. A3 i B3 i C3 i D3 i E3 .

[0062] From the appendix Figure 3 The entire diagnostic process can be visualized. First, the third harmonic component of the five-phase current is extracted. Then, the number of positive and negative values ​​within the fault cycle is recorded (actual measurements show that upper switch faults contain 2 positive and 3 negative values, while lower switch faults show the opposite). The diagnostic result is derived based on the inverse relationship between positive and negative values. Averaging the variables in the flowchart is to eliminate random errors in the actual measurements.

[0063] Appendix Figure 4 This represents the reconfigured voltage vector distribution under a fault in the switching transistor on phase A. A standard five-phase inverter can form 2... 5 There are 32 switching states, corresponding to 32 basic voltage vectors. A "0" indicates the upper switch is on, and a "1" indicates the lower switch is on. When the upper switch of phase A fails, the 16 voltage vectors with the value "0xxxx" (x=0&1) will be affected, corresponding to voltage vector U. 16 -U 31 With voltage vector U 17 For example, according to the KCL and KVL laws, the phase voltage U of the inverter under normal operating conditions can be obtained. AN U BN U CN U DN U EN 3 / 5U in sequence dc -2 / 5U dc -2 / 5U dc -2 / 5U dc 3 / 5U dc When a fault occurs in the switching transistor on phase A, due to the absence of phase A, the phase voltage U of the inverter will decrease. AN U BN U CN U DN U EN It will become 3 / 5U dc -2 / 5U dc -2 / 5U dc -2 / 5U dc 3 / 5U dc Then it is transformed onto the α-β axis and the α3-β3 axis:

[0064] ;

[0065] Among them, U dc U is the bus voltage. AN U BN U CN U DN U EN U is the phase voltage of the inverter. αβ and U αβ3 These are the α-β axis voltage and the α3-β3 axis voltage, respectively, with reference voltage vector U. 17 The voltage vector U affected by the fault can be 16 -U 31 All reconstructions are complete, and the original unaffected voltage vector U0-U is used in conjunction with them. 15 This constitutes the following: Figure 4 The diagram shows the reconfiguration voltage vector distribution under a fault in the switching transistor on phase A.

[0066] From the appendix Figure 5 The process of calculating the time ratio of the virtual voltage vector can be understood, mainly solving the problem that the two vectors in the combination under fault conditions are not on the same straight line and cannot be completely synthesized into 0. Taking voltage vectors U8 and U... 28 For example, the amplitude length of these two voltage vectors in cubic space is l1 = 0.4U. dc and l2=0.1453U dc The phase angle θ 8-28 The angle is 128°. To minimize the synthesized third harmonic voltage, this objective can be transformed into a mathematical problem of finding the shortest distance from the zero point to the third side of the triangle formed by the two vectors, which can be obtained by drawing a perpendicular line. The amplitude length l3 of the third side and the amplitude length h of the perpendicular line can be obtained by solving the triangle: ;

[0067] And voltage vectors U8 and U 28 The time ratio corresponds to the length ratio of l5 and l4 in the triangle, which can be calculated using the Pythagorean theorem: ;

[0068] Thus, the voltage vectors U8 and U are calculated. 28 The effective time ratio is 0.22:0.78. Similarly, the effective time ratios of other voltage vectors that are not on the same straight line can be obtained and synthesized into a virtual voltage vector. Finally, combining the virtual voltage vector synthesized from voltage vectors on the same straight line according to 0.618:0.382, the virtual voltage vector V1-V under switching transistor fault can be derived. 10 .

[0069] Appendix Figure 6This refers to the virtual voltage vector distribution of the switching transistor fault on phase A in the fundamental and harmonic spaces. Compared to before the virtual voltage vector was synthesized, the voltage component in the third harmonic space was significantly reduced.

[0070] Appendix Figure 7 Taking overall switching transistor failure as an example, the virtual voltage vectors of the five specific phase switching transistor failures are listed. The virtual voltage vectors of each phase failure are distributed in the fundamental frequency space with a 72-degree rotation relationship. Then, these five voltage vectors are unified into a whole vector set, which includes 10 virtual voltage vectors V1-V for each phase failure. 10 The resulting subset.

[0071] Appendix Figure 8 This involves dividing the market into large and small sectors, based on V1-V. 10 Within the coverage area, the subset initially divides the space into 10 large sectors IX. Since each subset is axially symmetric, each large sector is further divided into two smaller sectors, thus subdividing the spatial plane into 20 smaller sectors. .

[0072] Appendix Figure 9 It is the final modulation region of the general virtual voltage vector after adjusting the intersection region of the five-phase virtual voltage vectors, so that the intersection region meets the requirement of covering the entire plane 0-360°.

[0073] Appendix Figure 10 This is a block diagram for integrated diagnosis and fault-tolerant control of single-phase switch faults in a five-phase permanent magnet synchronous motor drive system.

[0074] Appendix Figure 11 This is the overall diagnostic result of a single switch tube fault in a five-phase permanent magnet synchronous motor. "1" represents a fault in the upper switch tube, and "0" represents a fault in the lower switch tube. Under different operating conditions, the diagnostic method proposed in this invention can simply and effectively diagnose both the overall upper switch tube fault and the overall lower switch tube fault.

[0075] Appendix Figure 12 This invention describes the current and torque waveforms of a five-phase permanent magnet synchronous motor system from normal to fault to fault-tolerant. When the switching transistors of different phases fail, the universal fault-tolerant method proposed in this invention can reduce the torque ripple caused by the fault.

[0076] Appendix Figure 13 This demonstrates the switching process when different phase switch tubes of a five-phase permanent magnet synchronous motor system fail. There is no need to redesign the algorithm based on the phase position. The same universal fault-tolerant method proposed in this invention can switch smoothly without causing system disturbance.

[0077] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "illustrative embodiment," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0078] Although embodiments of the invention have been shown and described, those skilled in the art will understand that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the claims and their equivalents.

Claims

1. An integrated diagnostic and fault-tolerant method for single-phase switch faults in a five-phase permanent magnet synchronous motor drive system, characterized in that, Includes the following steps: Step 1: Since traditional five-phase inverter fault diagnosis methods need to identify multiple fault states, and fault tolerance methods are often for faults of a specific switch in a specific phase and position, if the fault occurs in different phases and positions, fault tolerance methods need to be redesigned. In order to reduce the number of diagnostic states and enhance the versatility of fault tolerance methods, the 10 different single switch faults in the five-phase inverter system are unified into two types: overall upper switch fault and lower switch fault. Step 2: Sample the five-phase current i of the five-phase motor using a current sensor. A i B i C i D i E By analyzing the third harmonic component of the five-phase current during a single-switch fault, A3 i B3 i C3 i D3 i E3 Diagnostic criteria are obtained from the positive and negative values ​​of the fault cycle, enabling the diagnosis of both overall and downstream switching transistor faults. Step 3: When a single switch failure occurs in the five-phase permanent magnet synchronous motor inverter, the phase voltage of the inverter can be obtained according to the switching state at the time of the failure. The five-phase voltage is then transformed into α-β axis voltage and α3-β3 axis voltage, thereby reconstructing a new basic voltage vector under the fault. Step 4: Select voltage vector combinations suitable for modulation from the new set of basic voltage vectors. Based on geometric principles, with the goal of minimizing the voltage components of the third harmonic subspace, determine the action time ratio of each voltage vector combination, and then synthesize a virtual voltage vector with low harmonic characteristics. Step 5: Unify the virtual voltage vectors of five specific phase switch faults or lower switch faults in the inverter into a unified vector set, and construct a general virtual voltage vector suitable for overall switch faults or lower switch faults by finding the intersection area. Step 6: Select a suitable general virtual voltage vector based on the spatial position of the reference voltage vector and establish the corresponding vector action time. Input the obtained switching signal into the inverter to control the motor, thereby realizing fault-tolerant control of any single-phase switch failure in the five-phase permanent magnet synchronous motor drive system.

2. The integrated diagnosis and fault-tolerant method for single-phase switch faults in a five-phase permanent magnet synchronous motor drive system according to claim 1, characterized in that, The specific process of step 2 is as follows: Step 2.1: Using the Clarke transform matrix and inverse Clarke transform matrix, according to the transformation flowchart, the third harmonic component i of the five-phase current is transformed. A3 i B3 i C3 i D3 i E3 From phase current i A i B i C i D i E Separated from the three phases, since the third harmonic space of the five-phase motor does not participate in electromechanical energy conversion, the α3-β3 axis current under normal operating conditions should satisfy: ； Therefore, the third harmonic component of the phase current obtained by matrix transformation should also remain zero: ； Step 2.2, assuming the fault occurs at the switching transistor on phase A, the phase A current during the fault can be represented by segments of the upper half-cycle of the fault and the lower half-cycle of normal operation: ; Where iFault Atotal is the phase A current when the switching transistor on phase A fails, M is the amplitude of the phase current, and θ e It is an electrical angle; During the first half of the fault cycle, phase A current will be lost. At this time, the third harmonic component of the five-phase current will change and will no longer remain at 0. ; Where iFault A, iFault B, iFault C, iFault D, and iFault E are the five-phase currents when the switching transistor on phase A fails, and iFault A3, iFault B3, iFault C3, iFault D3, and iFault E3 are the third harmonic components of the phase current when the switching transistor on phase A fails. At this time, during the fault cycle, the third harmonic component of the five-phase current will inevitably exhibit N values ​​greater than 0 and (5-N) values ​​less than 0. Since the five-phase motor structure is completely symmetrical, the positive and negative values ​​of the third harmonic component will have the same total number of positive and negative values ​​when the upper switch tube of phase A is faulty and the upper switch tubes of phases B, C, D, and E are faulty. The only difference is that the distribution of positive and negative values ​​changes, but the overall number remains the same. The upper switch tube fault and the lower switch tube fault of the inverter are reversible. The lower switch tube will exhibit a completely opposite state, that is, (5-N) values ​​greater than 0 and N values ​​less than 0. Based on the positive and negative value relationship, the diagnosis of the upper switch tube fault and the lower switch tube fault of the entire five-phase inverter can be realized.

3. The integrated diagnosis and fault-tolerant method for single-phase switch faults in a five-phase permanent magnet synchronous motor drive system according to claim 1, characterized in that, The novel basic voltage vector reconstruction process described in step 3 is as follows: A standard five-phase inverter can form 2 5 =32 switching states, corresponding to 32 basic voltage vectors. The switching state is represented by "0" to indicate that the upper switch is on and "1" to indicate that the lower switch is on. Assuming that the fault occurs in the upper switch of phase A, then there are 16 switching states of "0xxxx" (x=0&1). The phase voltage U of the inverter at the time of the fault is obtained by using the KCL and KVL laws for the switching states affected by the fault. AN U BN U CN U DN U EN Then it is transformed onto the α-β axis and the α3-β3 axis: ; Among them, e jα U is a rotation operator, e is the natural constant, j is the imaginary number, α = 2 / 5π, and U AN U BN U CN U DN U EN U is the phase voltage of the inverter. αβ and U αβ3 These are the α-β axis voltage and the α3-β3 axis voltage, respectively. After reconstructing the 16 voltage vectors affected by the fault, they are combined with the original unaffected voltage vectors to form a new type of voltage vector under single-switch fault.

4. The integrated diagnosis and fault-tolerant method for single-phase switch faults in a five-phase permanent magnet synchronous motor drive system according to claim 1, characterized in that, The virtual voltage vector synthesis process described in step 4 is as follows: Step 4.1: Due to the complex distribution of the new voltage vector at the time of the reconstructed fault, preprocessing is required. Taking the fault of the switching transistor on phase A as an example, after excluding the voltage vector corresponding to the small vector under normal operating conditions, the remaining voltage vector U... 16 U 25 U 29 U 24 , U8, U 28, U 30, U 12 , U4, U 14 U 15 , U6, U2, U7, U 23 , U3, U1, U 19 U 27 U 17 They can be combined in pairs to form 10 groups, namely [U 16 U 25 ], [U 29 U 24 ],[U8, U 28 ], [U 30 U 12 ],[U4, U 14 ], [U 15 [U6], [U2, U7], [U 23 [U1, U3], [U3, U3] 19 ], [U 27 U 17 This also corresponds to the large and medium vectors under normal operating conditions; Step 4.2: To simplify the spatial distribution of voltage vectors, the voltage vectors from the 10 combinations will be synthesized into a virtual voltage vector according to a certain ratio. Simultaneously, to suppress the influence of the harmonic subspace, the third harmonic voltage after the virtual voltage vector synthesis is minimized; where [U4, U...] 14 ], [U 15 In the three combinations [U6], [U2, U7], the voltage vector is not affected by the fault. After synthesis using the action time ratio of 0.618:0.382, the virtual voltage vector with the third harmonic voltage is 0. Step 4.3: The remaining 7 combinations cannot be completely synthesized into zero because the two vectors within each combination are not on the same straight line. Therefore, the goal is to minimize the third harmonic voltage after synthesis. Geometrically, this is transformed into a mathematical problem of finding the shortest distance from the zero point to the third side of the triangle formed by the two vectors. Constructing perpendicular lines and solving the triangle determines the respective time ratios, thus synthesizing the virtual voltage vector with the minimum third harmonic voltage. Finally, these 10 combinations are organized into 10 virtual voltage vectors V1-V1 under switching transistor fault conditions. 10 .

5. The integrated diagnosis and fault-tolerant method for single-phase switch faults in a five-phase permanent magnet synchronous motor drive system according to claim 1, characterized in that, The process for establishing the general virtual voltage vector described in step 5 is as follows: Step 5.1: Assuming an overall switching transistor fault, list the virtual voltage vectors for the five specific phase switching transistor faults. The virtual voltage vectors for each phase fault are distributed in the fundamental frequency space with a 72-degree rotation relationship. Then, unify these five voltage vectors into a unified vector set, containing the 10 virtual voltage vectors V1-V1 for each phase fault. 10 The resulting subset; Step 5.2, based on V1-V 10 Within the coverage area, the subsets can initially divide the space into 10 large sectors IX. Since each subset is axially symmetric, each large sector can be further divided into 2 smaller sectors, thus subdividing the spatial plane into 20 smaller sectors. ; Step 5.3: Obtain the modulation region formed by the virtual voltage vectors of each phase fault. Find the intersection region of vector modulation in each small sector. The boundary of the intersection region is the prototype of the general virtual voltage vector. Considering the continuity of the control algorithm, the intersection region of vector modulation needs to meet the requirement of covering the entire plane 0-360°. Slightly adjust the phase of the general virtual voltage vector to establish the final general virtual voltage vector.

Images