Fault-tolerant control method for permanent magnet synchronous motor drive system under current sensor failure

Abstract

The present invention discloses a fault-tolerant control method under a current sensor fault of a permanent magnet synchronous motor drive system, which runs a fault diagnosis algorithm in real time to obtain a fault diagnosis flag, diagnoses and locates a healthy phase and a fault phase; predicts three-phase current in real time, and writes the three-phase predicted current as a healthy phase current prediction value and a fault phase current prediction value in combination with the fault diagnosis flag; calculates a dead zone compensation voltage, a current controller expected current correction value, and a feedback value; sends the obtained expected current correction value and feedback current to the current controller for closed-loop control, and at the same time, sends the obtained reference voltage superimposed on the dead zone compensation voltage to a space vector pulse width modulation algorithm to generate an inverter switch signal, thereby realizing fault-tolerant operation under a current sensor fault. The present invention can quickly and accurately locate the fault phase and the healthy phase, realize continuous and reliable operation of the permanent magnet synchronous motor drive system under a current sensor fault, and the change of motor parameters has little effect on the current steady-state control effect.

Description

Technical Field

[0001] The present invention belongs to the technical field of motor drive and control, and in particular relates to a fault-tolerant control method under a current sensor failure of a permanent magnet synchronous motor drive system. Background Art

[0002] Permanent Magnet Synchronous Motor (PMSM) is widely used in industrial drive, electric vehicles and other fields due to its advantages such as high power density, high efficiency and high power factor. At least two current sensors are required to achieve high-performance closed-loop control of permanent magnet synchronous motors. However, due to harsh working environments such as high temperature, high humidity and high vibration, current sensors may fail, affecting control performance. For electric vehicle applications, in some special scenarios, the motor system is often required to continue to operate in the absence of current measurement signals. Therefore, it is of great significance to study fault-tolerant control strategies under current sensor failures.

[0003] Fault diagnosis is the premise for achieving fault-tolerant control. After diagnosing the current sensor fault, there are currently two main methods to achieve fault-tolerant operation: one is to switch the system operation mode, switching the system from closed-loop to open-loop operation; the other is to use the current reconstruction strategy to achieve closed-loop control using existing current sensors. The current reconstruction strategy can be generally divided into the method of reconstructing the three-phase current based on the inverter switching state and the method of reconstructing the three-phase current based on the residual phase current sensor.

[0004] For a voltage source inverter using a space vector pulse width modulation algorithm, when different effective voltage vectors act, the DC bus current corresponds to different phase currents. Therefore, by installing current sensors on the DC bus, the three-phase current can be reconstructed; however, there is a reconstruction blind spot in the low-profile area and sector switching boundary. At the same time, the method of reconstructing the three-phase current based on the inverter switching state requires the addition of additional current sensors, which increases the cost of the drive system.

[0005] Since the probability of all phase current sensors in an actual motor system failing at the same time is very small, the use of residual phase current sensors to reconstruct three-phase currents and achieve high-performance closed-loop control has better application prospects. The method of reconstructing three-phase currents based on residual phase current sensors mainly uses observers to reconstruct three-phase currents. However, since the motor parameters will change with the operating conditions of the motor, current estimation errors will occur, so error correction is required. Since only one phase current can be obtained through sensor measurement after the current sensor fails, it is difficult to obtain a complete current estimation error correction term, so the cumbersome parameter setting process is a common problem of existing methods. Summary of the invention

[0006] In view of the above, the present invention provides a fault-tolerant control method for a permanent magnet synchronous motor drive system under current sensor failure, which greatly simplifies the parameter setting process. By correcting the expected current of the current controller, the current estimation error caused by the change of motor parameters can be offset, and the current control effect is less affected by the change of motor parameters.

[0007] A fault-tolerant control method for a permanent magnet synchronous motor drive system under a current sensor fault, comprising the following steps:

[0008] (1) Use current sensors to collect the stator current of the permanent magnet synchronous motor phases AB i A ~ i B , while detecting the rotational angular velocity of the permanent magnet synchronous motor oh r , electrical angular velocity oh e and electrical angle i e The d-axis component of the stator current in the dq coordinate system is obtained by rotating the coordinate system i d and the q-axis component i q ;

[0009] (2) Perform fault diagnosis on the current sensor and determine the fault flag Fault and the measured values ​​of the healthy phase current sensors i H ;

[0010] (3) When a phase current sensor fails, the current prediction model is combined with the fault flag to Fault Calculate and determine the predicted value of healthy phase current i Hp And the predicted value of fault phase current i F1p and i F2p ;

[0011] (4) Calculate the dead zone compensation voltage in the stationary coordinate system through the inverter dead zone compensation algorithm u α0 and u β0 ;

[0012] (5) Using the predicted value of healthy phase current i Hp Combined with the fault flag Fault Calculate the desired d-axis current correction value of the current controller i drefp and the q-axis expected current correction value iqrefp ;

[0013] (6) Using the measured value of the healthy phase current sensor i H And the predicted value of fault phase current i F1p and i F2p Combined with the fault flag Fault , calculate the d-axis feedback current of the current controller i drec and q-axis feedback current i qrec ;

[0014] (7) Correct the expected current value of the d-axis i drefp and the q-axis expected current correction value i qrefp and the d-axis feedback current i drec and q-axis feedback current i qrec Input into the current controller for closed-loop control to obtain the stator voltage reference value in the dq coordinate system and , and then and Perform coordinate transformation to obtain the stator voltage reference value in the αβ stationary coordinate system and , and finally , Respectively u α0 , u β0 After superposition, it is used as the input of SVPWM (Space Vector Pulse Width Modulation) algorithm to generate PWM (Pulse Width Modulation) signal for switching control of power devices in permanent magnet synchronous motor inverter.

[0015] Furthermore, the specific implementation of step (2) is as follows:

[0016] 2.1 Calculate the α-axis component of the stator current by the following expression: i α , α1 axis component i α1 , α-axis expected value 、α1 axis expected value :

[0017] , , ;

[0018]

[0019] in: and They are the d-axis expected value and q-axis expected value of the stator current (according to the torque command value and the angular velocity oh r Obtained by looking up the table);

[0020] 2.2 According to i α , i α1 , , Generate a residual signal by taking the difference and taking the absolute value err and err 1. The specific formula is as follows:

[0021] ,

[0022] 2.3 The residual signal err and err 1 and set threshold e Compare to diagnose the fault of the current sensor and determine the fault flag Fault , specifically:

[0023] like err < e and err 1< e ,but Fault =0, indicating that no current sensor fails;

[0024] like err < e and err 1≥ e ,but Fault =1, indicating that only the A phase current sensor is healthy, then i H = i A ;

[0025] like err ≥ e and err 1< e ,but Fault =2, indicating that only the B phase current sensor is healthy, then i H = i B ;

[0026] like err ≥ e and err 1≥ e ,but Fault =3, indicating that all current sensors are faulty.

[0027] Furthermore, the specific implementation of step (3) is as follows:

[0028] 3.1 The d-axis predicted value and q-axis predicted value of the stator current are calculated by the following current prediction model:

[0029]

[0030] in: and They represent the d-axis predicted value and q-axis predicted value of the stator current in the previous control cycle, respectively. and They represent the d-axis predicted value and q-axis predicted value of the stator current in the current control period respectively, and They represent the d-axis reference value and q-axis reference value of the stator voltage in the previous control cycle, respectively. L dn , L qn , R sn , are the nominal values ​​of the permanent magnet synchronous motor d-axis stator inductance, q-axis stator inductance, stator phase resistance, and permanent magnet flux linkage (these parameters are used in the prediction model). T s Indicates the control cycle duration;

[0031] 3.2 The d-axis predicted value and q-axis predicted value of the stator current are transformed by rotating the coordinates to obtain the estimated value of the three-phase stator current i Ap ~ i Cp ;

[0032] 3.3 Based on estimated values i Ap ~ i Cp Combined with the fault flag Fault , determine the predicted value of healthy phase current i Hp And the predicted value of fault phase current i F1p and i F2p , specifically:

[0033] when Fault =1, then i Hp = i Ap , i F1p = i Bp , i F2p = i Cp ;

[0034] when Fault =2, then i Hp = i Bp , i F1p = i Cp , i F2p = i Ap .

[0035] Furthermore, the specific expression of the inverter dead zone compensation algorithm in step (4) is as follows:

[0036] ;

[0037]

[0038] in: t d is the dead time of the permanent magnet synchronous motor inverter, u dc is the DC bus voltage of the permanent magnet synchronous motor inverter, sgn( ) represents the sign function, , , are the expected values ​​of the three-phase stator currents, and are the d-axis expected value and q-axis expected value of the stator current respectively, T s Indicates the control cycle duration.

[0039] Furthermore, the specific implementation of step (5) is as follows:

[0040] 5.1 Predicted value of healthy phase current i Hp Input to center frequency is oh e The second-order generalized integrator outputs two orthogonal current signals. i out1 and iout2 , and then calculate the current amplitude by the following formula I m and Phase :

[0041]

[0042] 5.2 According to the current amplitude I m and Phase Obtain the expected current correction value of the fault phase by phase shifting i F1ref and i F2ref , the specific expression is as follows:

[0043]

[0044] 5.3 According to the fault flag Fault The expected current of the healthy phase And the expected current correction value of the fault phase i F1ref and i F2ref The desired d-axis current correction value of the current controller is obtained by coordinate transformation i drefp and the q-axis expected current correction value i qrefp , specifically:

[0045] when Fault =1, the expected current of the healthy phase ,but:

[0046]

[0047] when Fault =2, the expected current of the healthy phase ,but:

[0048]

[0049] in: and are the expected values ​​of stator current for phase A and phase B respectively, T Indicates transpose.

[0050] Furthermore, the specific implementation of step (6) is as follows:

[0051] when Fault =1, d-axis feedback current i drec and q-axis feedback current i qrec Calculated by the following expression:

[0052]

[0053] when Fault =2, d-axis feedback current i drec and q-axis feedback current i qrec Calculated by the following expression:

[0054]

[0055] in: T Indicates transpose.

[0056] A computer device comprises a memory and a processor, wherein the memory stores a computer program, and the processor is used to execute the computer program to implement the fault-tolerant control method under current sensor failure of the permanent magnet synchronous motor drive system.

[0057] A computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the computer program is used to implement the fault-tolerant control method under current sensor failure of the permanent magnet synchronous motor drive system.

[0058] Based on the above technical solution, the present invention has the following beneficial technical effects:

[0059] 1. The present invention does not need to add an additional current sensor on the DC side, but makes full use of the existing phase current sensors, so as to realize the continuous operation of the permanent magnet synchronous motor system under the current sensor failure, thereby saving costs and ensuring the safety of the system.

[0060] 2. The present invention can offset the current estimation error caused by the change of motor parameters by correcting the expected current of the current controller. Therefore, the current control performance of the present invention is less affected by the change of motor parameters.

[0061] 3. Compared with the existing fault-tolerant control method, the present invention has fewer design parameters, which greatly simplifies the parameter setting process. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] Figure 1 It is an overall control block diagram of the fault-tolerant control method under current sensor failure of the permanent magnet synchronous motor drive system of the present invention.

[0063] Figure 2 It is a flow chart of the fault diagnosis algorithm of the present invention.

[0064] Figure 3 It is a flow chart of the expected current correction module of the present invention.

[0065] Figure 4This is the structural block diagram of the second-order generalized integrator.

[0066] Figure 5 When the second-order generalized integrator gain k =0.5 center frequency oh e =628rad / s when the Bode diagram of the transfer function, where (a) corresponds to the transfer function D ( s ), (b) corresponds to the transfer function Q ( s ).

[0067] Figure 6 Schematic diagram of experimental results of the transition process of the fault-tolerant control method of the present invention when a B-phase current sensor fails, wherein (a) is the experimental result when the controller parameters are accurate, and (b) is the experimental result when the controller parameters are mismatched.

[0068] Figure 7 It is a schematic diagram of the steady-state current tracking experiment results of the motor running under the condition of B-phase current sensor failure without adding dead zone compensation algorithm and the controller parameters are accurate using the fault-tolerant control of the present invention.

[0069] Figure 8 It is a schematic diagram of the steady-state current tracking experiment results of the motor running under the condition of a B-phase current sensor failure, adding a dead zone compensation algorithm and using the fault-tolerant control of the present invention when the controller parameters are accurate.

[0070] Figure 9 It is a schematic diagram of the steady-state current tracking experiment results of the motor running with the B-phase current sensor fault, adding the dead zone compensation algorithm and the controller parameters mismatching using the fault-tolerant control of the present invention. DETAILED DESCRIPTION

[0071] In order to describe the present invention more specifically, the technical solution of the present invention is described in detail below in conjunction with the accompanying drawings and specific implementation methods.

[0072] like Figure 1 As shown, the fault-tolerant control method of the permanent magnet synchronous motor drive system under current sensor failure of the present invention comprises the following steps:

[0073] (1) Use the permanent magnet synchronous motor drive system measurement circuit to obtain the permanent magnet synchronous motor A phase current i A and B phase measuring current i B ,Will i A and i B The motor d and q axis currents are obtained by coordinate transformation i d and iq , the motor rotation angular velocity is obtained by the rotary transformer and signal processing circuit oh r , electrical angular velocity oh e and electrical angle i e , by the torque command and the current motor rotation angular velocity oh r The expected current of the d and q axes is obtained by looking up the table (which can be obtained through offline calibration) and And the expected current of three phases A, B and C is obtained by coordinate transformation , , .

[0074] Among them, the motor d and q axis currents i d and i q The calculation formula is as follows:

[0075]

[0076] A, B, C three-phase expected current , , The calculation formula is as follows:

[0077]

[0078] (2) Run the fault diagnosis algorithm in real time to obtain the fault flag Fault , and locate the phase where the healthy sensor is located, define it as the healthy phase, and the rest as the fault phase, and get the measured value of the healthy phase current sensor i H .

[0079] The fault diagnosis algorithm should also be executed once in each control cycle when the motor is operating normally, such as Figure 2 As shown, the specific implementation is as follows:

[0080] 2.1 Calculate by the following formula i α , i α1 , , :

[0081] , , ;

[0082]

[0083] in: i α is the α-axis current obtained by Clarke transformation when the α-axis coincides with the A-axis; i α1 is the α1-axis current obtained by Clarke transformation when the α1-axis coincides with the B-axis; , Represents the expected current of the α and α1 axes.

[0084] 2.2 Measurement value i α , i α1 and expected value , Subtract and take the absolute value to generate the residual err and err 1;

[0085] ,

[0086] 2.3 Residual err and err 1 respectively with the threshold e Compare to determine whether the corresponding phase current sensor is faulty and get the fault flag Fault and healthy phase current sensor measurements i H , as follows:

[0087] like err < e and err 1< e ,but Fault =0, indicating that no current sensor fails; if err < e And err1≥ e ,but Fault =1, indicating that only the A phase current sensor is healthy. i H = i A ;like err ≥ e and err 1< e ,but Fault =2, indicating that only the B phase current sensor is healthy. i H = i B ;like err ≥ e and err 1≥ e ,but Fault =3, indicating that all current sensors are faulty.

[0088] (3) Run the current prediction model and obtain the three-phase current estimation value through rotating coordinate transformation i Ap , i Bp and i Cp , combined with the fault flag Fault , write the three-phase current prediction value as the healthy phase current prediction value i Hp and the predicted value of the fault phase current i F1p and i F2p .

[0089] First, the d and q axis current prediction values ​​are obtained through the current prediction model i dp and i qp , as follows:

[0090]

[0091] in: and They represent the predicted values ​​of the d-axis and q-axis currents of the previous control cycle respectively; and They represent the d-axis and q-axis reference voltages of the previous control cycle, respectively, which are obtained by the output of the current controller; L dn , L qn , R sn , are the nominal values ​​of the d-axis stator inductance, q-axis stator inductance, stator phase resistance, and permanent magnet flux linkage of the permanent magnet synchronous motor (these parameters are used by the prediction model); T s To control the cycle.

[0092] Then i dp and i qp Obtain three-phase current estimation through rotating coordinate transformation i Ap , i Bp and i Cp , as follows:

[0093]

[0094] Finally, combined with the fault flag Fault , write the three-phase current prediction value as the healthy phase current prediction value i Hp and the predicted value of the fault phase current i F1p and i F2p , as follows:

[0095] when Fault =1, we have:

[0096] i Hp = i Ap , i F1p = i Bp , i F2p = i Cp

[0097] when Fault =2, we have:

[0098] i Hp = i Bp , i F1p = i Cp , i F2p = i Ap

[0099] (4) Execute the inverter dead zone compensation algorithm to obtain the dead zone compensation voltage in the stationary coordinate system u α0 and u β0 .

[0100] The inverter dead zone compensation algorithm is as follows:

[0101]

[0102] Where: sgn represents the sign function; t d is the inverter dead time; u dc is the DC bus voltage.

[0103] (5) Calculate the expected current correction values ​​of the d and q axes of the current controller i drefp and iqrefp , this correction value replaces and As the desired current of the current controller (using PI control, i.e. proportional integral control), the specific process is as follows Figure 3 As shown:

[0104] 5.1 Predicted value of healthy phase current i Hp The input center frequency is oh e The second-order generalized integrator outputs two orthogonal signals i out1 and i out2 . Figure 4 is the structure of a second-order generalized integrator. s Domain i out1 and i out2 It can be written as:

[0105] ;

[0106]

[0107] in: k is the gain of the second-order generalized integrator; D ( s ) is the input i Hp To Output i out1 The transfer function of Q ( s ) is the input i Hp To Output i out2 The transfer function of oh g is the center frequency of the second-order generalized integrator.

[0108] like Figure 5 The second-order generalized integrator gain is shown as k =0.5, center frequency oh g = oh e =628rad / s D ( s )and Q ( s ) from the Bode diagram, we can see that when oh g = oh e hour, iout1 Tracking input signal, i out2 The input signal lags behind by 90 degrees. Therefore, it can be calculated by the following formula i Hp The amplitude I m and Phase :

[0109]

[0110] 5.2 Obtaining the expected current correction value of the two fault phases by phase shifting i F1ref , i F2ref :

[0111]

[0112] 5.3 Combined Fault Flag Fault , the expected current of the healthy phase and the expected current correction value of the two-phase fault i F1ref , i F2ref The desired current correction values ​​of the d and q axes of the current controller are obtained by coordinate transformation. i drefp and i qrefp :

[0113] when Fault =1, the expected current of the healthy phase ,but i drefp and i qrefp It can be calculated as:

[0114]

[0115] when Fault =2, the expected current of the healthy phase ,but i drefp and i qrefp It can be calculated as:

[0116]

[0117] in: T Indicates transpose.

[0118] (6) Calculate the d-axis and q-axis feedback currents of the current controller i drec and iqrec , the feedback current replaces i d and i q Serves as feedback current for the current controller.

[0119] when Fault =1, then i drec and i qrec It can be calculated as:

[0120]

[0121] when Fault =2, then i drec and i qrec It can be calculated as:

[0122]

[0123] (7) Correct the expected current value i drefp , i qrefp With feedback current i drec , i qrec The current controller is sent to perform closed-loop control to obtain the d and q axis reference voltages. and ,Will and The reference voltage in the stationary coordinate system is obtained by coordinate transformation and , on this basis superimpose the dead zone compensation voltage u α0 and u β0 The superimposed voltage is used as the input of the space vector pulse width modulation algorithm to modulate the reference voltage, and the output switching signal is applied to the inverter to achieve closed-loop control of the motor.

[0124] in and The reference voltage in the stationary coordinate system is obtained by coordinate transformation and , the specific formula is as follows:

[0125]

[0126] In order to verify the effectiveness and superiority of the fault-tolerant control method of the permanent magnet synchronous motor current sensor failure of the present invention, we conducted an experimental verification. During the experiment, the control cycle of the algorithm was 50us. In the experiment, the built-in permanent magnet synchronous motor was taken as an example. The parameters are shown in Table 1:

[0127] Table 1

[0128]

[0129] When the motor runs normally at 750r / min and 300N·m, the measured value of the phase B current sensor is artificially changed to 0 in the controller to simulate the signal loss fault of the phase B current sensor. The fault enable variable is defined as Fault_ enable , when the B phase current sensor has a signal loss fault, its value is set to 1. Figure 6 The figure shows the transient process experimental results of the fault-tolerant control method of the present invention at 750 r / min and 300 N·m, wherein the first line is the torque waveform and the second line is the fault enabling variable. Fault_enable and fault flag Fault Waveform diagram and partial enlargement diagram. The third line is the expected current of the motor d axis. and the actual current i d Waveform, the fourth line is the expected current of the motor q axis and the actual current i q Waveform, the fifth line is the actual three-phase current of the motor i A , i B , i C Waveform. Figure 6 As shown in (a), when the controller parameters are accurate, the fault diagnosis algorithm can quickly locate the fault phase in one control cycle. After locating the fault phase, it can smoothly switch to the corresponding fault-tolerant control algorithm, and the motor system can achieve continuous and reliable operation when the current measurement signal is lost. Figure 6 As shown in (b), continuous operation can be achieved when the current measurement signal is lost even in the case of controller parameter mismatch (the actual motor resistance is 1.4 times the nominal value, the d-axis inductance is 0.7 times the nominal value, the q-axis inductance is 0.8 times the nominal value, and the permanent magnet flux is 0.87 times the nominal value).

[0130] In order to verify the current steady-state control performance of the fault-tolerant control method of the present invention, the motor is operated in the fault-tolerant control condition after the B-phase current sensor fails, the motor speed is 500r / min, and the torque is 200N·m. Figure 7 This is the experimental result when the dead zone compensation algorithm is not added and the controller parameters are accurate. Figure 8These are the experimental results when the dead zone compensation algorithm is added and the controller parameters are accurate. Figure 9 The experimental results are as follows when the dead zone compensation algorithm is added and the controller parameters are mismatched (the actual motor resistance is 1.4 times the nominal value, the d-axis inductance is 0.7 times the nominal value, the q-axis inductance is 0.8 times the nominal value, and the permanent magnet flux is 0.87 times the nominal value). Figure 7~Figure 9 The experimental results include the torque waveform, the expected current of the motor d and q axes , and the actual current i d , i q Waveform, actual three-phase current of the motor i A , i B , i C Waveform, motor d-axis current i d The result of FFT transformation (Fast Fourier Transform) is shown in Figure 1. st , 2 nd , 6 th Respectively represent the 1st, 2nd, and 6th harmonics. Figure 7 and Figure 8 It can be seen that after adding the dead zone compensation algorithm, the motor output torque fluctuation is reduced, the three-phase current symmetry is increased, and the FFT of the motor d-axis current shows that the second harmonic and sixth harmonic in the d-axis current are significantly reduced, proving the effectiveness of the dead zone compensation algorithm in improving the current tracking performance of the present invention. Figure 8 and Figure 9 It can be seen that after adding the dead zone compensation algorithm, in the case of controller parameter mismatch, the motor torque fluctuation increases slightly compared to when the controller parameters are accurate. FFT of the motor d-axis current shows that the second harmonic component in the motor d-axis current increases slightly. Therefore, it can be concluded that the controller parameter mismatch has little effect on the current tracking performance of the present invention.

[0131] The above description of the embodiments is to facilitate the understanding and application of the present invention by those skilled in the art. It is obvious that those skilled in the art can easily make various modifications to the above embodiments and apply the general principles described herein to other embodiments without creative work. Therefore, the present invention is not limited to the above embodiments. Improvements and modifications made by those skilled in the art to the present invention based on the disclosure of the present invention should be within the protection scope of the present invention.

Claims

1. A fault-tolerant control method for a permanent magnet synchronous motor drive system under a current sensor fault, comprising the following steps: (1) Use current sensors to collect the stator current of the permanent magnet synchronous motor phases AB i A ~ i B , while detecting the rotational angular velocity of the permanent magnet synchronous motor ω r , electrical angular velocity ω e and electrical angle θ e The d-axis component of the stator current in the dq coordinate system is obtained by rotating the coordinate system i d and the q-axis component i q ; (2) Perform fault diagnosis on the current sensor and determine the fault flag Fault and the measured values ​​of the healthy phase current sensors i H ; (3) When a phase current sensor fails, the current prediction model is combined with the fault flag to Fault Calculate and determine the predicted value of healthy phase current i Hp And the predicted value of fault phase current i F1p and i F2p ; (4) Calculate the dead zone compensation voltage in the stationary coordinate system through the inverter dead zone compensation algorithm u α0 and u β0 ; (5) Using the predicted value of healthy phase current i Hp Combined with the fault flag Fault Calculate the desired d-axis current correction value of the current controller i drefp and the q-axis expected current correction value i qrefp ; (6) Using the measured value of the healthy phase current sensor i H And the predicted value of fault phase current i F1p and i F2p Combined with the fault flag Fault , calculate the d-axis feedback current of the current controller i drec and q-axis feedback current i qrec ; (7) Correct the expected current value of the d-axis i drefp and the q-axis expected current correction value i qrefp and the d-axis feedback current i drec and q-axis feedback current i qrec Input into the current controller for closed-loop control to obtain the stator voltage reference value in the dq coordinate system and , and then and Perform coordinate transformation to obtain the stator voltage reference value in the αβ stationary coordinate system and , and finally , Respectively u α0 , u β0 After superposition, they are used as the input of the SVPWM algorithm to generate a PWM signal for switching control of the power devices in the permanent magnet synchronous motor inverter.

2. The fault-tolerant control method for a permanent magnet synchronous motor drive system under current sensor failure according to claim 1, characterized in that: The specific implementation of step (2) is as follows: 2.1 Calculate the α-axis component of the stator current by the following expression: i α , α1 axis component i α1 , α-axis expected value 、α1 axis expected value : ， ， ; ; in: and are the d-axis expected value and q-axis expected value of the stator current respectively; 2.2 According to i α , i α1 , , Generate a residual signal by taking the difference and taking the absolute value err and err 1. The specific formula is as follows: ， ; 2.3 The residual signal err and err 1 and set threshold ε Compare to diagnose the fault of the current sensor and determine the fault flag Fault , specifically: like err < ε and err 1< ε ,but Fault =0, indicating that no current sensor fails; like err < ε and err 1≥ ε ,but Fault =1, indicating that only the A phase current sensor is healthy, then i H = i A ; like err ≥ ε and err 1< ε ,but Fault =2, indicating that only the B phase current sensor is healthy, then i H = i B ; like err ≥ ε and err 1≥ ε ,but Fault =3, indicating that all current sensors are faulty.

3. The fault-tolerant control method for a permanent magnet synchronous motor drive system under current sensor failure according to claim 1, characterized in that: The specific implementation of step (3) is as follows: 3.1 The d-axis predicted value and q-axis predicted value of the stator current are calculated by the following current prediction model: ; in: and They represent the d-axis predicted value and q-axis predicted value of the stator current in the previous control cycle, respectively. and They represent the d-axis predicted value and q-axis predicted value of the stator current in the current control period respectively, and They represent the d-axis reference value and q-axis reference value of the stator voltage in the previous control cycle, respectively. L dn , L qn , R sn , are the nominal values ​​of the d-axis stator inductance, q-axis stator inductance, stator phase resistance, and permanent magnet flux linkage of the permanent magnet synchronous motor, respectively. T s Indicates the control cycle duration; 3.2 The d-axis predicted value and q-axis predicted value of the stator current are transformed by rotating the coordinates to obtain the estimated value of the three-phase stator current i Ap ~ i Cp ; 3.3 Based on estimated values i Ap ~ i Cp Combined with the fault flag Fault , determine the predicted value of healthy phase current i Hp And the predicted value of fault phase current i F1p and i F2p , specifically: when Fault =1, then i Hp = i Ap , i F1p = i Bp , i F2p = i Cp ; when Fault =2, then i Hp = i Bp , i F1p = i Cp , i F2p = i Ap .

4. The fault-tolerant control method for a permanent magnet synchronous motor drive system under current sensor failure according to claim 1, characterized in that: The specific expression of the inverter dead zone compensation algorithm in step (4) is as follows: ; ; in: t d is the dead time of the permanent magnet synchronous motor inverter, u dc is the DC bus voltage of the permanent magnet synchronous motor inverter, sgn( ) represents the sign function, , , are the expected values ​​of the three-phase stator currents, and are the d-axis expected value and q-axis expected value of the stator current respectively, T s Indicates the control cycle duration.

5. The fault-tolerant control method for a permanent magnet synchronous motor drive system under current sensor failure according to claim 1, characterized in that: The specific implementation of step (5) is as follows: 5.1 Predicted value of healthy phase current i Hp Input to center frequency is ω e The second-order generalized integrator outputs two orthogonal current signals. i out1 and i out2 , and then calculate the current amplitude by the following formula I m and Phase : ; 5.2 According to the current amplitude I m and Phase Obtain the expected current correction value of the fault phase by phase shifting i F1ref and i F2ref , the specific expression is as follows: ; 5.3 According to the fault flag Fault The expected current of the healthy phase And the expected current correction value of the fault phase i F1ref and i F2ref The desired d-axis current correction value of the current controller is obtained by coordinate transformation i drefp and the q-axis expected current correction value i qrefp , specifically: when Fault =1, the expected current of the healthy phase ,but: ; when Fault =2, the expected current of the healthy phase ,but: ; in: and are the expected values ​​of stator current for phase A and phase B respectively, T Indicates transpose.

6. The fault-tolerant control method for a permanent magnet synchronous motor drive system under current sensor failure according to claim 1, characterized in that: The specific implementation of step (6) is as follows: when Fault =1, d-axis feedback current i drec and q-axis feedback current i qrec Calculated by the following expression: ; when Fault =2, d-axis feedback current i drec and q-axis feedback current i qrec Calculated by the following expression: ; in: T Indicates transpose.

7. A computer device comprising a memory and a processor, characterized in that: The memory stores a computer program, and the processor is used to execute the computer program to implement a fault-tolerant control method for a permanent magnet synchronous motor drive system under current sensor failure as described in any one of claims 1 to 6.

8. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by the processor, it implements the fault-tolerant control method for a permanent magnet synchronous motor drive system under current sensor failure as described in any one of claims 1 to 6.

Images