A fault-tolerant control method for open-circuit fault of symmetrical six-phase permanent magnet motor

Abstract

The present invention belongs to the technical field of electromechanical control, and more specifically, relates to a symmetrical six-phase permanent magnet motor open-circuit fault-tolerant control method. The method comprises: obtaining the actual voltage vector in the αβ and xy subspaces under normal conditions; then selecting six actual voltage vectors whose size in the xy subspace is zero and whose size in the αβ subspace is not zero as a finite control set; then using these six actual voltage vectors to divide the αβ subspace into six sectors; then using the deadbeat control principle to calculate the reference voltage vector, and selecting two actual voltage vectors in the sector where the reference voltage vector is located as candidate voltage vectors; selecting the optimal voltage vector through a cost function, and obtaining the corresponding power device drive signal. The present invention solves the problem that the smooth transition from the healthy condition to the fault condition of the symmetrical six-phase motor under the healthy condition and the open-circuit fault condition requires fault diagnosis and control reconstruction, the drive system structure is complex, and the system transient switching process is unstable.

Description

Technical Field

[0001] The present invention belongs to the technical field of electromechanical control, and more specifically, relates to a symmetrical six-phase permanent magnet motor open-circuit fault-tolerant control method. Background Art

[0002] In recent years, six-phase permanent magnet synchronous motors have attracted widespread attention in many high-end fields due to their excellent fault tolerance and flexible control strategies. Applications in these high-end fields usually require six-phase permanent magnet synchronous motors to maintain high reliability under long-term and high-load conditions, which requires advanced fault-tolerant control to ensure that the motor can automatically adjust its operation strategy and continue to maintain its function when a fault occurs. Electrical faults occurring in motor drive systems can generally be divided into short-circuit faults and open-circuit faults. Short-circuit faults can cause overcurrent problems, which can damage the equipment. The general method for such faults is to isolate the faulty components and convert short-circuit faults into open-circuit faults. Therefore, the research on fault-tolerant control of motors for open-circuit faults has attracted more attention. In addition, the increase in the number of phases of six-phase permanent magnet synchronous motors will undoubtedly further increase the incidence of open-circuit faults. Therefore, the open-circuit fault-tolerant control of six-phase permanent magnet motor drive systems has always been a research hotspot.

[0003] Chinese invention patent CN110336511A discloses a fault-tolerant dual-vector predictive control method and device for a six-phase permanent magnet motor, including: obtaining the rotor position, speed and phase current of the six-phase motor; obtaining the fault reverse potential according to the rotor position, speed, permanent magnet flux on the rotor and the position of the fault phase axis, and establishing a motor inverter model considering the fault reverse potential; establishing a reduced-order motor mathematical model according to the position of the fault phase; establishing a switch table of the six-phase motor after the fault, obtaining the phase voltage corresponding to each switch vector, substituting it into the reduced-order motor mathematical model, and determining the phase current prediction value after each switch vector acts for a sampling period; evaluating the error between the phase current prediction value corresponding to each switch vector and the phase current command value, and selecting the switch vector that produces the smallest error as the optimal switch vector; modulating the optimal switch vector and the zero vector within the sampling period to determine the duty cycle of the optimal switch vector.

[0004] The research on fault-tolerant control of six-phase permanent magnet synchronous motors mainly focuses on the fields of fault detection, isolation, reconstruction control and real-time monitoring. Researchers have achieved fault diagnosis and early warning through a variety of methods, including model-based methods and data-driven machine learning techniques. These methods can effectively identify and predict motor faults and improve the reliability of the system. At the same time, the improvement of control strategies enables the motor to be quickly reconstructed in the event of a fault and maintain stable operation. However, the existing fault-tolerant control strategies also have some shortcomings. First, many fault-tolerant control methods rely on simplified mathematical models, which may not accurately reflect the true dynamic characteristics of the motor under complex working conditions, resulting in a decrease in the accuracy of fault identification. Secondly, most fault-tolerant control methods need to reconstruct the transformation matrix, mathematical model, control method, etc. according to the diagnosed fault type, which will not only affect the accuracy of fault diagnosis, but the reconstruction process will also cause the drive system to be unstable. In addition, fault diagnosis and control reconstruction will lead to a complex control system, which will undoubtedly aggravate the impact of control delay and thus affect the performance of fault-tolerant control. Although some fault-tolerant control methods for six-phase permanent magnet motors that do not require reconstruction of transformation matrices and mathematical models have been proposed, which greatly simplifies the control system, these methods still cannot avoid the process of fault diagnosis and control switching after the fault.

[0005] The existing natural fault-tolerant control method that does not require fault diagnosis and switching process is to use two basic voltage vectors to synthesize a voltage vector with a size of zero in the xy subspace, avoiding the control target conflict problem between the dq controller and the xy controller caused by the open circuit fault. However, this method of using a synthetic voltage vector will undoubtedly increase the complexity of fault-tolerant control. In addition, most of the current research objects on fault-tolerant control of six-phase permanent magnet motors are asymmetric six-phase motors, and the characteristics of the basic voltage vectors with a size of zero in the xy subspace and a non-zero size in the αβ subspace in the symmetric six-phase permanent magnet motor control are not taken into account. Summary of the invention

[0006] The present invention aims to overcome at least one defect of the above-mentioned prior art and provides a symmetrical six-phase permanent magnet motor open-circuit fault-tolerant control method. After an open-circuit fault occurs in the motor, a smooth transition from a healthy operating condition to a faulty operating condition can be achieved without the need for fault diagnosis and reconstruction of the control mode, so as to solve the problems of diagnostic delay, complex calculation, and control instability caused by reconstruction of the control mode existing in the current multi-phase motor fault-tolerant control.

[0007] The detailed technical scheme of the present invention is as follows:

[0008] A symmetrical six-phase permanent magnet motor open-circuit fault-tolerant control method, comprising:

[0009] S1. Using a vector space decoupling transformation matrix, the voltage vectors corresponding to 64 switch states of the voltage source inverter driving the symmetrical six-phase permanent magnet motor are transformed into the αβ subspace and the xy subspace, and the actual voltage vectors and their spatial distribution in the αβ subspace and the xy subspace are obtained under normal conditions;

[0010] S2, based on the principle of controlling the current in the xy subspace to be equal to zero, six actual voltage vectors whose size is zero in the xy subspace and non-zero in the αβ subspace are selected as the finite control set of the model predictive voltage control;

[0011] Furthermore, the model used in the model prediction voltage control is the voltage equation of the symmetrical six-phase permanent magnet motor in the dq subspace:

[0012]

[0013] In formula (7), u d and u q is the voltage in the dq subspace, R s is the stator resistance, i d and i q is the current in the dq subspace, L d and L q are the d-axis inductance and the q-axis inductance, ω is the rotor electrical angular velocity, ψ f is the permanent magnet flux amplitude.

[0014] S3, based on the six actual voltage vectors of the finite control set, the αβ subspace is divided into six sectors: sector I, sector II, sector III, sector IV, sector V and sector VI;

[0015] S4, using the deadbeat control principle to calculate the reference voltage vector, determine the specific position of the reference voltage vector in the sector in S3 according to the electrical angle of the reference voltage vector, and select two actual voltage vectors in the sector as candidate voltage vectors for model prediction voltage control;

[0016] S5. Design a cost function, apply the candidate voltage vector to the cost function, select the candidate voltage vector that makes the cost function value smaller as the optimal voltage vector, and obtain the corresponding power device switching signal for controlling the voltage source inverter.

[0017] Furthermore, the voltage vectors corresponding to the 64 switching states of the voltage source inverter driving the symmetrical six-phase permanent magnet motor are transformed into the αβ subspace and the xy subspace using the vector space decoupling transformation matrix T as follows:

[0018]

[0019] In formula (1), u α and u βis the voltage of the αβ subspace; u x and u y is the voltage in the xy subspace; U dc is the DC side voltage of the inverter driving the symmetrical six-phase permanent magnet motor; the inverter driving the symmetrical six-phase permanent magnet motor has 6 bridge arms, corresponding to the six phases A, U, B, V, C, and W respectively. Each bridge arm is divided into an upper bridge arm and a lower bridge arm, and there is a power device in each upper bridge arm and lower bridge arm. In the formula, s represents the switching state of the power devices in the upper bridge arm and the lower bridge arm of each bridge arm, and s i =1 (i = A, B, C, U, V, W) means that the upper bridge arm power device is turned on and the lower bridge arm power device is turned off, s i =0 (i=A, B, C, U, V, W) means that the upper arm power device is turned off and the lower arm power device is turned on.

[0020] Furthermore, the switching state of the voltage source inverter is represented by a six-bit binary number, which corresponds to the switching state of the power device on each bridge arm of the six-phase bridge arm of the voltage source inverter. 1 indicates that the power device on the upper bridge arm of this phase is turned on and the power device on the lower bridge arm is turned off, and 0 indicates that the power device on the upper bridge arm of this phase is turned off and the power device on the lower bridge arm is turned on. Each digit of the six-bit binary number corresponds to a bridge arm, and the actual voltage vector serial number is a decimal number represented by the six-bit binary number.

[0021] Furthermore, the electrical angles corresponding to the six actual voltage vectors of the finite control set in the αβ subspace are 0°, 60°, 120°, 180°, 240°, and 300°, respectively.

[0022] Furthermore, the calculation of the reference voltage vector using the deadbeat control principle specifically includes:

[0023] S41, obtaining the dq subspace current sampling value at time k by performing vector space decoupling transformation and Park transformation on the six-phase current collected at time k;

[0024] S42, the dq subspace voltage reference value calculated at time (k-1) is delayed by one sampling period as the dq subspace voltage estimation value at time k, and combined with the dq subspace current sampling value at time k, according to the discretization results of the motor dq subspace voltage and current differential equations, the dq subspace current prediction value at time (k+1) is obtained as follows:

[0025]

[0026] In formula (2), and is the predicted value of the dq subspace current at time (k+1), and is the dq subspace current sampling value at time k, and is the estimated voltage in the dq subspace at time k, ψ f is the permanent magnet flux amplitude, T s is the sampling period, R s is the stator resistance, L d is the d-axis inductance, L q is the q-axis inductance, ω is the rotor electrical angular velocity;

[0027] S43, according to the influence of a sampling cycle delay, the q-axis current reference value obtained by the speed loop PI controller and the given d-axis current reference value Combined with the predicted value of the dq subspace current at time (k+1), the dq subspace voltage reference value is calculated as:

[0028]

[0029] In formula (3), and is the dq subspace voltage reference value;

[0030] S44, the obtained dq subspace voltage reference value is subjected to Park inverse transformation to obtain the αβ subspace voltage reference value, and the αβ subspace voltage reference value is used to calculate the electrical angle of the reference voltage vector:

[0031]

[0032] In formula (4), θ ref is the electrical angle of the reference voltage vector, and is the voltage reference value of αβ subspace.

[0033] Furthermore, the S5 specifically includes:

[0034] S51. The designed cost function g is:

[0035]

[0036] In formula (5), u id and u iq is the dq subspace voltage of the candidate voltage vector.

[0037] S52, the αβ subspace components corresponding to the candidate voltage vector are subjected to Park transformation to obtain the dq subspace components corresponding to the candidate voltage vector, and the calculation formula is as follows:

[0038]

[0039] In formula (6), u iα and uiβ are the αβ subspace components of the candidate voltage vector.

[0040] S53, then substitute the dq subspace components corresponding to the two candidate voltage vectors into the cost function respectively, select the candidate voltage vector that makes the cost function smaller as the optimal voltage vector, and obtain the power device switching signal according to the switching state of the voltage source inverter corresponding to the optimal voltage vector, which is used to control the voltage source inverter.

[0041] Compared with the prior art, the present invention has the following beneficial effects:

[0042] (1) The present invention provides a symmetrical six-phase permanent magnet motor open-circuit fault-tolerant control method, which does not require fault diagnosis after an open-circuit fault occurs in the motor and can achieve a smooth transition from before the fault to after the fault occurs, thereby fundamentally avoiding the impact of misdiagnosis and diagnosis delay on the symmetrical six-phase permanent magnet motor drive.

[0043] (2) The present invention provides a symmetrical six-phase permanent magnet motor open-circuit fault-tolerant control method, which applies the same control mode before and after the motor open-circuit fault, and there is no switching of control modes, which fundamentally solves the problems of program complexity and control instability caused by control mode reconstruction; at the same time, the actual voltage vector is directly selected as the finite control set of the model predictive voltage control, avoiding the complex program caused by the synthetic voltage vector.

[0044] (3) The present invention provides a symmetrical six-phase permanent magnet motor open-circuit fault-tolerant control method, which simplifies the selection process of the optimal voltage vector by using the zero-beat control principle, reduces the calculation burden of the fault-tolerant control method, and can achieve satisfactory fault-tolerant control when the motor has a single-phase open-circuit fault or a two-phase open-circuit fault. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 It is a schematic diagram of the architecture of the open-circuit fault-tolerant control method for a symmetrical six-phase permanent magnet motor provided by the present invention;

[0046] Figure 2 It is a voltage space vector diagram of a symmetrical six-phase two-level inverter according to a specific embodiment of the present invention;

[0047] Figure 3 is a voltage space vector and sector distribution diagram of a finite control set in a specific embodiment of the present invention;

[0048] Figure 4 is a diagram of phase current simulation results of a single-phase open circuit fault in a specific embodiment of the present invention;

[0049] Figure 5 is a diagram of the simulation results of the speed and torque of a single-phase open circuit fault in a specific embodiment of the present invention;

[0050] Figure 6 is a diagram of phase current simulation results of a two-phase open circuit fault in a specific embodiment of the present invention;

[0051] Figure 7 It is a diagram of the speed and torque simulation results of a single-phase open circuit fault in a specific embodiment of the present invention. DETAILED DESCRIPTION

[0052] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0053] It should be noted that the following detailed descriptions are exemplary and are intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meanings as those commonly understood by those skilled in the art to which the present invention belongs.

[0054] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit exemplary embodiments according to the present invention. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "comprising" and / or "including" are used in this specification, it indicates the presence of features, steps, operations, devices, components and / or combinations thereof.

[0055] In the absence of conflict, the embodiments of the present invention and the features of the embodiments may be combined with each other.

[0056] Example 1

[0057] This embodiment takes a symmetrical six-phase permanent magnet motor with two sets of three-phase windings that are 60° apart and whose neutral points are isolated from each other as the object, and proposes a symmetrical six-phase permanent magnet motor open-circuit fault-tolerant control method. The structural block diagram of the method is as follows Figure 1 As shown, the phase sequence of the symmetrical six-phase permanent magnet motor is distributed counterclockwise as A, U, B, V, C, and W, and the difference between two adjacent phases is 60° electrical angle, as follows:

[0058] S1. Use the vector space decoupling transformation matrix to transform the voltage vectors corresponding to the 64 switching states of the voltage source inverter driving the symmetrical six-phase permanent magnet motor into the αβ subspace and the xy subspace, and obtain the actual voltage vectors and their spatial distribution in the αβ subspace and the xy subspace under normal circumstances.

[0059] Specifically, the vector space decoupling transformation matrix T used in the symmetrical six-phase symmetrical permanent magnet motor is:

[0060]

[0061] Preferably, the voltage vectors corresponding to the 64 switch states of the voltage source inverter driving the symmetrical six-phase permanent magnet motor are transformed into the αβ subspace and the xy subspace by using the vector space decoupling transformation matrix T:

[0062]

[0063] In formula (1), u α and u β is the voltage of the αβ subspace; u x and u y is the voltage in the xy subspace; U dc is the DC side voltage of the inverter driving the symmetrical six-phase permanent magnet motor; the inverter driving the symmetrical six-phase permanent magnet motor has 6 bridge arms, corresponding to the six phases of AUBVCW, each bridge arm is divided into an upper bridge arm and a lower bridge arm, and there is a power device in each upper bridge arm and lower bridge arm. In the above formula, s represents the switching state of the power devices of the upper bridge arm and the lower bridge arm of each bridge arm, s i =1 (i = A, B, C, U, V, W) means that the upper bridge arm power device is turned on and the lower bridge arm power device is turned off, s i = 0 (i = A, B, C, U, V, W) means that the upper bridge arm power device is turned off and the lower bridge arm power device is turned on. In this way, the spatial voltage vector diagram of the αβ subspace and the xy subspace can be obtained under normal conditions, as shown in Figure 2 shown.

[0064] Among them, the corresponding relationship between the sequence number of the actual voltage vector and the 64 switch states is shown in Table 1:

[0065] Table 1 Correspondence between the serial number of the actual voltage vector and the 64 switch states

[0066]

[0067] Among them, the switching state of the voltage source inverter is represented by a six-bit binary number, and the six-bit binary number corresponds to the switching state of the power device on each bridge arm of the six-phase bridge arm of the voltage source inverter. The switching state of the voltage source inverter is represented by a six-bit binary number, and the six-bit binary number corresponds to the switching state of the power device on each bridge arm of the six-phase bridge arm of the voltage source inverter. 1 indicates that the power device of the upper bridge arm of the phase is turned on and the power device of the lower bridge arm is turned off, and 0 indicates that the power device of the upper bridge arm of the phase is turned off and the power device of the lower bridge arm is turned on. Each digit of the six-bit binary number corresponds to a bridge arm: 1 indicates that the power device of the upper bridge arm of the phase is turned on and the power device of the lower bridge arm is turned off, and 0 indicates that the power device of the upper bridge arm of the phase is turned off and the power device of the lower bridge arm is turned on; each digit of the six-bit binary number corresponds to a bridge arm, and 1 or 0 indicates the switching state of the power device on the bridge arm. For example, 000000 indicates that all six bridge arms have the power devices of the lower bridge arm turned on and the power devices of the upper bridge arm turned off. The actual voltage vector serial number is the decimal number represented by the six-bit binary number.

[0068] S2. Based on the principle of controlling the current in the xy subspace to be equal to zero, six actual voltage vectors whose size is zero in the xy subspace and not zero in the αβ subspace are selected as the finite control set for model predictive voltage control.

[0069] Furthermore, the model used in the model prediction voltage control is the voltage equation of the symmetrical six-phase permanent magnet motor in the dq subspace:

[0070]

[0071] In formula (7), u d and u q is the voltage in the dq subspace, R s is the stator resistance, i d and i q is the current in the dq subspace, L d and L q are the d-axis inductance and the q-axis inductance, ω is the rotor electrical angular velocity, ψ f is the permanent magnet flux amplitude.

[0072] The serial numbers of the six actual voltage vectors selected by S2 are: 37 # , 52 # ,twenty two # , 26 # , 11 # , 41 # The corresponding voltage source inverter switching states are: 100101, 110100, 010110, 011010, 001011, 101001.

[0073] The main reason why the traditional control method fails due to an open-circuit fault in a symmetrical six-phase motor is that the open-circuit fault will cause a constraint relationship between the αβ subspace current and the xy subspace current, causing the two to no longer be independent of each other. Therefore, the traditional independent control method cannot operate normally after an open-circuit fault. Therefore, the key to achieving fault-tolerant faults is to achieve independent control of the αβ subspace current and the xy subspace current after the fault. In addition, when the six-phase motor is running healthily, the xy subspace current needs to be controlled to zero to reduce the harmonic current;

[0074] In summary, the symmetrical six-phase permanent magnet motor fault-tolerant control method proposed in the present invention takes the principle of controlling the xy subspace current to be equal to zero when selecting the voltage vector, and selects six actual voltage vectors whose size is zero in the xy subspace and not zero in the αβ subspace as the finite control set of the model predictive voltage control; this selection method can not only reduce the harmonic current in the healthy operation state of the six-phase motor, but also realize independent control of the αβ subspace current and the xy subspace current in both the healthy operation and open circuit fault operation states of the six-phase motor, and thus realize fault-tolerant control without changing the control mode before and after the open circuit fault.

[0075] Specifically, from the spatial voltage vector diagrams of the αβ subspace and the xy subspace, it can be found that the 64 actual voltage vectors in the αβ subspace and the xy subspace can be divided into large voltage vectors (LVV), medium voltage vectors (MVV), small voltage vectors (SVV), and zero voltage vectors (ZVV) according to the magnitude of the amplitude. # , 7 # , 11 # ,twenty two # , 26 # , 37 # , 41 # , 52 # , 56 # and 63 # The size of the xy subspace is zero, where the vector 0 # , 7 # , 56 # and 63 # The size of the αβ subspace is also zero, which does not meet the selection criteria. The remaining 11 # ,twenty two # , 26 # , 37 # , 41 # and 52 # The size in the αβ subspace is not zero, which meets the selection criteria. Therefore, the serial numbers of the six actual voltage vectors finally selected are: 37 # , 52 # ,twenty two # , 26 # , 11 #, 41 # , the corresponding voltage source inverter switch states are: 100101, 110100, 010110, 011010, 001011, 101001. The six actual voltage vectors finally selected are used as the finite control set of model predictive voltage control. It is not difficult to find from the selection process that a finite vector control set that meets the standard can be selected from only 64 actual basic voltage vectors. This is exactly the characteristic of the symmetrical six-phase motor, which avoids the disadvantage that the fault-tolerant control method applied to the asymmetrical six-phase motor needs to synthesize the voltage vector to meet the selection criteria.

[0076] S3. Based on the six actual voltage vectors of the finite control set, the αβ subspace is divided into six sectors: sector I, sector II, sector III, sector IV, sector V and sector VI.

[0077] Specifically, the spatial distribution of the six actual voltage vectors selected in S2 in the αβ subspace is as follows: Figure 3 As shown, vector 37 # , 52 # ,twenty two # , 26 # , 11 # , 41 # The corresponding electrical angles in the αβ subspace are 0°, 60°, 120°, 180°, 240°, and 300° respectively.

[0078] from Figure 3 It can be seen from the figure that these six actual voltage vectors divide the αβ subspace into six sectors. # and 52 # The space between is sector I, vector 52 # and 22 # In between is sector II, vector 22 # and 26 # In between is sector III, vector 26 # and 11 # The area between is sector IV, vector 11 # and 41 # In between is sector V, vector 41 # and 37 # The area in between is sector VI.

[0079] S4. Calculate the reference voltage vector using the deadbeat control principle, determine the specific position of the reference voltage vector in the sector in S3 according to the electrical angle of the reference voltage vector, and select two actual voltage vectors in the sector as candidate voltage vectors for model prediction voltage control.

[0080] Traditional control methods often use a limited control set directly as a candidate voltage vector, which will result in the need to sequentially bring the six actual voltage vectors selected by S2 into the cost function when selecting the optimal voltage vector, and compare the calculation results corresponding to the six actual voltage vectors, in order to select the actual voltage vector that minimizes the cost function as the optimal voltage vector. This method requires six calculations, which greatly increases the calculation burden. The symmetrical six-phase permanent magnet motor fault-tolerant control method proposed in the present invention adopts the deadbeat control principle to reduce the number of candidate voltage vectors from 6 to 2, so only two calculations are required when selecting the optimal voltage vector, which reduces the calculation burden of the fault-tolerant control method.

[0081] Specifically, the calculation of the reference voltage vector using the deadbeat control principle specifically includes:

[0082] S41. The collected six-phase current at time k is transformed by vector space decoupling to obtain the calculation of the αβ subspace and xy subspace currents at time k as follows:

[0083]

[0084] In formula (9), is the current sampling value of phase i (i = A, B, C, U, V, W) at time k; and is the current sampling value of the αβ subspace at time k; and is the current sampling value of the xy subspace at time k.

[0085] Then, the αβ subspace current at time k can be transformed by Park to obtain the dq subspace current sampling value at time k:

[0086]

[0087] In formula (12), and is the current sampling value of the dq subspace at time k; θ is the electrical angle of the rotor position, which is obtained through the decoder.

[0088] S42, the dq subspace voltage reference value calculated at time (k-1) is delayed by one sampling period as the dq subspace voltage estimation value at time k, and combined with the dq subspace current sampling value at time k in S41, according to the discretization results of the motor dq subspace voltage and current differential equations, the dq subspace current prediction value at time (k+1) is obtained by the prediction model:

[0089]

[0090] In formula (2), and is the predicted value of the dq subspace current at time (k+1), and is the dq subspace current sampling value at time k, and is the estimated voltage in the dq subspace at time k, ψ f is the permanent magnet flux amplitude, T s is the sampling period, R s is the stator resistance, L d is the d-axis inductance, L q is the q-axis inductance, ω is the rotor electrical angular velocity, which is obtained through the speed calculation module.

[0091] S43, according to the influence of a sampling cycle delay, the q-axis current reference value obtained by the speed loop PI controller and the given d-axis current reference value Combined with the dq subspace current prediction value at time (k+1) in S42, the dq subspace voltage reference value can be calculated as:

[0092]

[0093] In formula (3), and is the dq subspace voltage reference value.

[0094] S44, the dq subspace voltage reference value obtained in S43 is subjected to Park inverse transformation to obtain the αβ subspace voltage reference value:

[0095]

[0096] In formula (11), and is the voltage reference value of the αβ subspace;

[0097] The electrical angle of the reference voltage vector is calculated using the αβ subspace voltage reference value:

[0098]

[0099] In formula (4), θ ref is the electrical angle of the reference voltage vector.

[0100] Specifically, from Figure 3 It is not difficult to see from the voltage space vector and sector distribution diagram of the finite control set that the corresponding relationship between the electrical angle of the reference voltage vector and the two candidate voltage vectors is shown in Table 2:

[0101] Table 2 Correspondence between two candidate voltage vectors

[0102]

[0103]

[0104] S5. Design a cost function, apply the candidate voltage vectors in S4 to the cost function, select the candidate voltage vector that makes the cost function value smaller as the optimal voltage vector, and obtain the corresponding power device switching signal according to the optimal voltage vector to control the voltage source inverter.

[0105] Specifically, the S5 specifically includes:

[0106] S51. The designed cost function g is:

[0107]

[0108] In formula (5), u id and u iq is the dq subspace voltage of the candidate voltage vector.

[0109] S52, the αβ subspace components corresponding to the candidate voltage vector are subjected to Park transformation to obtain the dq subspace components corresponding to the candidate voltage vector, and the calculation formula is as follows:

[0110]

[0111] In formula (6), u iα and u iβ are the αβ subspace components of the candidate voltage vector.

[0112] S53, then substitute the dq subspace components corresponding to the two candidate voltage vectors into the cost function respectively, select the candidate voltage vector that makes the cost function smaller as the optimal voltage vector, and obtain the power device switching signal according to the switching state of the voltage source inverter corresponding to the optimal voltage vector, which is used to control the voltage source inverter.

[0113] Furthermore, this embodiment takes two faults, A phase fault and AU fault, as examples for analysis as follows:

[0114] The analysis of the motor open circuit fault is as follows: the collected six-phase current of the motor is converted into two mutually orthogonal two-dimensional decoupled subspaces through the space vector transformation matrix T. The conversion process can be expressed as:

[0115]

[0116] The single-phase open circuit fault is analyzed by taking the A phase open circuit as an example. When the A phase is open circuit, i A =0 and i B +i C =0, by substituting into the above formula, we can get the electrical constraints of the system when the A phase open circuit fault occurs:

[0117] i x =-i α (13);

[0118] According to formula (13), the open circuit fault of phase A will cause the currents in the αβ subspace and the xy subspace to couple with each other and cannot be controlled independently. This will lead to a conflict between the closed-loop controllers of the αβ subspace and the xy subspace, resulting in increased torque pulsation and unstable speed.

[0119] Take the A-phase and U-phase open circuit fault as an example. When the A-phase and U-phase open circuits, i A =0, i B +i C =0, i U =0 and i V +i W =0, and the current constraint of the system when phase A and phase U have open circuit faults can be obtained by substituting it into the decoupling transformation equation of the six-phase current vector space of the motor:

[0120]

[0121] According to formula (14), the current coupling in the αβ subspace and the xy subspace caused by the two-phase open circuit fault is more serious, and when the open circuit fault occurs, it will seriously affect the control effect.

[0122] The symmetrical six-phase permanent magnet motor fault-tolerant control method described in this embodiment avoids the conflict between the current closed-loop controllers in the αβ subspace and the xy subspace caused by the open-circuit fault by selecting a special actual voltage vector. Therefore, this method does not require fault diagnosis, and does not require changing the control framework before and after the fault. The symmetrical six-phase permanent magnet motor drive system can naturally realize the fault-tolerant control of the motor single-phase open-circuit fault and the two-phase open-circuit fault.

[0123] The simulated phase current, speed and torque waveforms of the open-circuit fault-tolerant control method for a symmetrical six-phase permanent magnet motor before and after the A phase fault are as follows: Figure 4 , Figure 5 The simulated phase current, speed and torque waveforms before and after the A-phase and U-phase faults are shown in Figure 6 , Figure 7 As shown. Obviously, no matter it is a phase A fault or a phase A and U fault, the speed and torque waveforms before and after the fault remain stable, and the speed and torque ripples do not increase significantly. It can be seen that the control method proposed in the present invention can achieve a smooth transition from a healthy condition to a fault condition without the need for fault diagnosis and control reconstruction, and is a simple and easy-to-implement fault-tolerant control method for a symmetrical six-phase permanent magnet motor.

[0124] The control method of the present invention is applicable to healthy operating conditions and open-circuit fault conditions of a symmetrical six-phase permanent magnet motor, and can realize fault-tolerant control of the symmetrical six-phase permanent magnet motor after an open-circuit fault without the need for fault diagnosis and control reconstruction.

[0125] Obviously, the above embodiments of the present invention are merely examples for clearly illustrating the technical solution of the present invention, and are not intended to limit the specific implementation methods of the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the claims of the present invention shall be included in the protection scope of the claims of the present invention.

Claims

1. A symmetrical six-phase permanent magnet motor open circuit fault tolerance control method, characterized in that: The method comprises: S1. Using a vector space decoupling transformation matrix, the voltage vectors corresponding to 64 switch states of the voltage source inverter driving the symmetrical six-phase permanent magnet motor are transformed into the αβ subspace and the xy subspace, and the actual voltage vectors and their spatial distribution in the αβ subspace and the xy subspace under normal conditions are obtained, and the actual voltage vectors are numbered with decimal numbers corresponding to the switch states; S2, based on the principle of controlling the current in the xy subspace to be equal to zero and eliminating the control conflict between the αβ and xy subspaces after the open-phase fault, six actual voltage vectors with zero size in the xy subspace and non-zero size in the αβ subspace are selected as the finite control set of the model predictive voltage control; S3. Based on the six actual voltage vectors of the finite control set, the αβ subspace is divided into six sectors: sector I, sector II, sector III, sector IV, sector V and sector VI; S4, using the deadbeat control principle to calculate the reference voltage vector, determine the specific position of the reference voltage vector in the sector in S3 according to the electrical angle of the reference voltage vector, and select two actual voltage vectors in the sector as candidate voltage vectors for model prediction voltage control; S5, designing a cost function, applying the candidate voltage vector to the cost function, selecting the candidate voltage vector that makes the cost function value smaller as the optimal voltage vector, and obtaining the corresponding power device switching signal based on the cost function to control the voltage source inverter; The specific steps of calculating the reference voltage vector by using the deadbeat control principle include: S41, obtaining the dq subspace current sampling value at time k by performing vector space decoupling transformation and Park transformation on the six-phase current collected at time k; S42, the dq subspace voltage reference value calculated at time (k-1) is delayed by one sampling period as the dq subspace voltage estimation value at time k, and combined with the dq subspace current sampling value at time k, according to the discretization results of the motor dq subspace voltage and current differential equations, the dq subspace current prediction value at time (k+1) is obtained as follows: （2）； In formula (2), and is the predicted value of the dq subspace current at time (k+1), and is the dq subspace current sampling value at time k, and is the estimated value of the dq subspace voltage at time k, is the permanent magnet flux amplitude, is the sampling period, is the stator resistance value, is the d-axis inductance, is the q-axis inductance, ω is the rotor electrical angular velocity; S43, according to the influence of a sampling cycle delay, the q-axis current reference value obtained by the speed loop PI controller and the given d-axis current reference value , combined with the predicted value of the dq subspace current at time (k+1), the dq subspace voltage reference value is calculated as: （3）； In formula (3), and is the dq subspace voltage reference value; S44, the obtained dq subspace voltage reference value is subjected to Park inverse transformation to obtain the αβ subspace voltage reference value, and the αβ subspace voltage reference value is used to calculate the electrical angle of the reference voltage vector: （4）； In formula (4), is the electrical angle of the reference voltage vector, and is the voltage reference value of αβ subspace.

2. A symmetrical six-phase permanent magnet motor open circuit fault tolerance control method according to claim 1, characterized in that: The vector space decoupling transformation matrix T is used to transform the voltage vectors corresponding to the 64 switching states of the voltage source inverter driving the symmetrical six-phase permanent magnet motor into the αβ subspace and the xy subspace as follows: （1）； In formula (1), and is the voltage of the αβ subspace; and is the voltage in the xy subspace; is the DC side voltage of the inverter driving the symmetrical six-phase permanent magnet motor; the inverter driving the symmetrical six-phase permanent magnet motor has 6 bridge arms, corresponding to the six phases A, U, B, V, C, and W respectively. Each bridge arm is divided into an upper bridge arm and a lower bridge arm, and there is a power device in each of the upper bridge arm and the lower bridge arm. Indicates the switching state of the power devices in the upper and lower arms of each bridge arm, s i =1 Indicates that the upper bridge arm power device is turned on and the lower bridge arm power device is turned off. s i =0 Indicates that the upper bridge arm power device is turned off and the lower bridge arm power device is turned on. =A,B,C,U,V,W.

3. The method for controlling an open-circuit fault of a symmetrical six-phase permanent magnet motor according to claim 1, characterized in that: The switching state of the voltage source inverter is represented by a six-bit binary number, which corresponds to the switching state of the power device on each bridge arm of the six-phase bridge arm of the voltage source inverter. 1 means that the power device on the upper bridge arm of the phase is turned on and the power device on the lower bridge arm is turned off, and 0 means that the power device on the upper bridge arm of the phase is turned off and the power device on the lower bridge arm is turned on. Each digit of the six-bit binary number corresponds to a bridge arm, and the actual voltage vector serial number is a decimal number represented by the six-bit binary number.

4. The method for controlling an open-circuit fault of a symmetrical six-phase permanent magnet motor according to claim 1, characterized in that: The electrical angles corresponding to the six actual voltage vectors of the finite control set in the αβ subspace are 0°, 60°, 120°, 180°, 240°, and 300° respectively.

5. A symmetrical six-phase permanent magnet motor open circuit fault tolerance control method according to claim 4, characterized in that: The S5 specifically includes: S51. Design cost function g for: （5）； In formula (5), and is the dq subspace voltage of the candidate voltage vector; S52, the αβ subspace components corresponding to the candidate voltage vector are subjected to Park transformation to obtain the dq subspace components corresponding to the candidate voltage vector, and the calculation formula is as follows: （6）； In formula (6), and is the αβ subspace component of the candidate voltage vector; S53, then substitute the dq subspace components corresponding to the two candidate voltage vectors into the cost function respectively, select the candidate voltage vector that makes the cost function smaller as the optimal voltage vector, and obtain the power device switching signal according to the switching state of the voltage source inverter corresponding to the optimal voltage vector, which is used to control the voltage source inverter.

Images