A direct torque control method for dual three-phase motors with single-phase open-circuit fault tolerance

Abstract

The present invention discloses a direct torque control method for fault tolerance of single-phase open-circuit fault of dual three-phase motors, comprising the following steps: using a multi-stage hysteresis loop, considering the positive and negative and amplitude of torque and magnetic flux at the same time, selecting different voltage vector combinations for voltage modulation of α-β plane and x-y harmonic plane in dynamic and steady state, and performing fault-tolerant operation for single-phase open-circuit fault of dual three-phase permanent magnet synchronous motors, correcting the voltage modulation of x-y harmonic plane and α-β plane, and realizing direct torque control of multi-stage hysteresis loop of dual three-phase permanent magnet synchronous motors. The present invention considers the positive and negative and amplitude of torque and magnetic flux, selects different voltage vector combinations for voltage vector synthesis in dynamic and steady state, further reduces torque pulsation, applies it to single-phase open-circuit fault condition of direct torque control, adopts full-dimensional model after fault for fault-tolerant operation, modifies current reference value after motor fault, does not need to change controller, has simple structure, and has small torque pulsation.

Description

Technical Field

[0001] The invention belongs to the field of dual three-phase motor drive control, and in particular relates to a direct torque control method for dual three-phase motors with single-phase open-circuit fault tolerance. Background Art

[0002] Dual three-phase permanent magnet synchronous motors have relatively good fault tolerance capabilities. The current mainstream method of fault handling is to convert all faults into phase loss faults and perform fault-tolerant control to convert them into phase loss operation. Therefore, the current fault-tolerant control strategy of dual three-phase motors is mainly focused on the research of motor phase loss operation.

[0003] At present, the control of dual three-phase PMSM motors generally utilizes space decoupling transformation. According to the vector space decoupling theory, through the stationary coordinate transformation (1), all variables of the dual three-phase motor are mapped to the α-β plane, xy plane and o1-o2 plane. Among them, the α-β sub-plane participates in the electromechanical energy conversion, the xy plane is the harmonic plane, and the 5th and 7th harmonic currents are mapped to this plane. Under normal control, they should be suppressed to 0. The o1-o2 sub-plane is the zero-sequence sub-plane. When the neutral points are isolated from each other, the variables corresponding to the o1-o2 sub-plane are all zero.

[0004] In this way, the motor can be controlled by controlling the physical quantities of the α-β plane, the xy plane, and the o1-o2 plane, and then controlling the physical quantities of each phase. During normal operation, these three sub-planes are decoupled from each other.

[0005] Disadvantages of the existing dual three-phase permanent magnet synchronous motor direct torque fault-tolerant control strategy:

[0006] The fault-tolerant control of the direct torque of the dual three-phase permanent magnet synchronous motor requires the establishment of a dimension reduction mathematical model after the fault, and different motor models need to be established for different faults, which is relatively complicated. In addition, the controller needs to be switched before and after the fault, and the control is not simple enough.

[0007] Currently, most fault-tolerant strategies based on direct torque control are based on virtual voltage vectors. Although the direct torque control strategy based on virtual voltage vectors eliminates the influence of the inverter voltage output on the xy harmonic plane, it cannot suppress the harmonic current caused by factors such as back electromotive force, dead time and external interference. Summary of the invention

[0008] In view of the shortcomings of the prior art, the present invention proposes a direct torque control method with single-phase open-circuit fault tolerance for dual three-phase motors, which takes into account the positive and negative and amplitude of the torque and magnetic flux, selects different voltage vector combinations for voltage vector synthesis in dynamic and steady states, and further reduces the torque pulsation. It is applied to the single-phase open-circuit fault condition of direct torque control, and adopts a full-dimensional model after the fault for fault-tolerant operation. The current reference value is modified after the motor fails, without changing the controller, and has a simple structure and small torque pulsation.

[0009] To achieve the above object, the present invention provides the following solutions:

[0010] A direct torque control method for a dual three-phase motor with single-phase open-circuit fault tolerance comprises the following steps:

[0011] By utilizing multi-stage hysteresis loops and considering the positive and negative and amplitude of torque and flux linkage at the same time, different voltage vector combinations are selected in dynamic and steady states to perform voltage modulation in the α-β plane and xy harmonic plane. Fault-tolerant operation is performed for single-phase open-circuit faults of dual three-phase permanent magnet synchronous motors, and the xy harmonic plane and α-β plane voltage modulation are corrected to realize direct torque control of multi-stage hysteresis loops of dual three-phase permanent magnet synchronous motors.

[0012] Preferably, the method for performing voltage modulation in the α-β plane and the xy harmonic plane includes:

[0013] The xy harmonic plane voltage and α-β plane voltage are modulated, and the flux observer is used to observe the flux and torque of the motor to perform multi-level hysteresis closed-loop control, calculate the errors of flux and torque, and select the voltage vector according to the error values ​​of flux and torque.

[0014] Preferably, the method of selecting different voltage vector combinations includes:

[0015] Three adjacent voltage vectors are selected as a voltage vector group, including two adjacent D4 vectors and one D2 vector, to synthesize the output voltage.

[0016] Preferably, a model reference adaptive method is used to design a flux observer based on the stator flux as a state variable to identify the flux of the motor and perform closed-loop control of the flux.

[0017] Preferably, the method of using the model reference adaptive method to design a flux observer based on the stator flux as a state variable to identify the flux of the motor and perform closed-loop control of the flux includes:

[0018] Using the basic principle of model reference adaptive control, a reference model and an adjustable model of flux linkage are constructed to observe the flux linkage of the motor;

[0019] Define the error vector and make a difference between the reference model and the adjustable model;

[0020] According to the stability theory, the transfer function of the forward path is obtained;

[0021] Substituting the difference between the reference model and the adjustable model into the transfer function to obtain an estimated flux linkage;

[0022] The estimated stator current is obtained by using the estimated flux, and the difference between the estimated stator current and the actual value of the motor current forms a closed-loop feedback link to achieve closed-loop control of the flux.

[0023] Preferably, the reference model is:

[0024] In the formula, u′ q =u q , R s represents the stator resistance, ω e is the electrical angular frequency, L d and L q They are the d-axis and q-axis stator direct-axis inductances and quadrature-axis inductances ψ d , q is the stator flux of d and q axes, u d 、u q are the d-axis and q-axis stator voltages respectively;

[0025] The adjustable model is:

[0026] are the estimated values ​​of the stator flux of the d and q axes, Estimated value of electrical speed.

[0027] Preferably, the method for performing fault-tolerant operation for a single-phase open-circuit fault of a dual three-phase permanent magnet synchronous motor includes:

[0028] After a single-phase open-circuit fault occurs in a dual three-phase permanent magnet synchronous motor, a full-dimensional model of the dual three-phase permanent magnet synchronous motor after the fault is used to select the corresponding minimum copper loss or maximum torque mode fault-tolerant current according to the position of the broken phase, and add it to the harmonic plane current reference value. The PR controller is used to track the harmonic current reference value to achieve single-phase open-circuit fault-tolerant operation.

[0029] Preferably, the transfer function of the proportional resonant PR controller in the discrete domain can be expressed as:

[0030]

[0031] The parameters are:

[0032]

[0033] k pr is the resonance gain; ω cis the cut-off frequency; ω0 is the resonant frequency; T s is the stator torque.

[0034] Compared with the prior art, the present invention has the following beneficial effects:

[0035] The present invention mainly aims at the problems that the fault-tolerant operation of dual three-phase permanent magnet synchronous motors in phase loss-tolerant operation adopts a dimensionality reduction model for fault-tolerant control, which requires the replacement of the controller before and after a fault and has a limited effect on suppressing torque pulsation. The present invention proposes direct torque control for the phase-loss-tolerant operation of dual three-phase permanent magnet synchronous motors based on a full-dimensional model.

[0036] First, the algorithm with reduced dimension needs to modify the controller before and after the fault, while the proposed algorithm only needs to modify the current reference value after the fault to achieve single-phase open circuit fault-tolerant operation;

[0037] Second, the fault tolerance of the dimensionality reduction model is based on the virtual voltage vector, which has equal amplitude and has limited effect on reducing torque pulsation. On this basis, the voltage vector is corrected for fault tolerance, but the effect on harmonic suppression is limited. The proposed algorithm is based on multi-stage hysteresis control, and selects voltage vector groups based on dynamic and steady states respectively. It can more finely synthesize the voltage vector of the required amplitude, and can better synthesize the voltage more finely according to the positive and negative magnetic flux torque and the amplitude for control, and the harmonic suppression effect is better.

[0038] The third is to use a flux observer based on the model reference adaptive method to achieve accurate identification of the motor flux and perform closed-loop control of the motor flux. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] In order to more clearly illustrate the technical solution of the present invention, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0040] Figure 1 A schematic diagram of the topological structure of the stator winding and the converter of the present invention;

[0041] Figure 2 : is a space voltage vector diagram of a dual three-phase two-level voltage source converter in an embodiment of the present invention; wherein (a) is a space voltage vector diagram of a dual three-phase two-level voltage source converter in an α-β plane; (b) is a space voltage vector diagram of a dual three-phase two-level voltage source converter in an xy plane;

[0042] Figure 3 It is a fault-tolerant operation control block diagram of direct torque control based on multi-level hysteresis loop in an embodiment of the present invention;

[0043] Figure 4Schematic diagram of the selection of voltage vector groups proposed in the present invention; (a) is a schematic diagram of the selection of α-β plane voltage vector groups; (b) is a schematic diagram of the selection of xy plane voltage vector groups;

[0044] Figure 5 Schematic diagram of the process of the method for synthesizing vectors of reference voltage vectors in the xy plane in an embodiment of the present invention;

[0045] Figure 6 Schematic diagram of the synthesized xy plane zero vector in the embodiment of the present invention; wherein (a) is a schematic diagram of the α-β plane zero vector; (b) is a schematic diagram of the xy plane zero vector;

[0046] Figure 7 Schematic diagram of the selection and action time calculation of the v0 vector group in an embodiment of the present invention;

[0047] Figure 8 It is a flow chart of the fault-tolerant operation of ST-DTC based on multi-level hysteresis loop in an embodiment of the present invention;

[0048] Fig. 9 is a forward structure diagram of an observer system based on MRAS in an embodiment of the present invention;

[0049] Fig.10 4 is a structural diagram of a flux observer in an embodiment of the present invention. DETAILED DESCRIPTION

[0050] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0051] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0052] Embodiment 1

[0053] like Figure 1 As shown, the research object of the present invention is a dual three-phase permanent magnet synchronous motor (DT-PMSM) with a double neutral point phase shift of 30°, the stator winding is powered by a two-level voltage source inverter, and the two sets of neutral points are isolated from each other. The present invention mainly performs fault-tolerant control when a single-phase open circuit fault occurs when the dual three-phase PMSM adopts direct torque control.

[0054] The present invention performs direct torque control for single-phase open-circuit fault tolerance of dual three-phase permanent magnet synchronous motors based on multi-stage hysteresis, proposes a phase-loss fault-tolerant control strategy based on normal decoupling transformation, keeps the system decoupling transformation matrix unchanged, and the voltage equation, flux equation and torque equation will not be affected. It only needs to modify the harmonic subspace reference current, analyze the calculation of different plane reference values ​​of voltage vector synthesis after the single-phase open-circuit fault of the dual three-phase motor, consider the influence of harmonic plane voltage modulation on α-β voltage modulation after the fault, and correct the voltage modulation during the fault to adapt to the fault-tolerant working condition. The direct torque control based on multi-stage hysteresis simultaneously considers the positive and negative and amplitude of the torque flux, and can more thoroughly describe the degree to which the system needs to increase or decrease the torque or flux, select a more suitable voltage vector combination for direct torque control, and can further reduce the torque pulsation.

[0055] The motor spatial decoupling matrix can be expressed as:

[0056]

[0057] The voltage square of the dual three-phase permanent magnet synchronous motor can be expressed as:

[0058]

[0059] Among them, R s represents the stator resistance; stator voltage u s =u α +ju β ,u z =u x +ju y ; stator current i s =i α +ji β ,i z =i x +ji y ; stator flux ψ s =ψ α +jψ β , ψ z =ψ x +jψ y .

[0060] The flux formula of the dual three-phase permanent magnet synchronous motor can be expressed as:

[0061]

[0062] Where: L s =3L m +L z ; L m and L z are the stator winding self-inductance and leakage inductance respectively; ψ f is the amplitude of the rotor flux; θe is the electrical angle of the rotor position;

[0063] The electromagnetic torque of the motor can be expressed as:

[0064] T e =3n p (i s ×ψ s ) (4)

[0065] n p is the number of motor pole pairs.

[0066]

[0067] Apply equation (5) to map the voltage vector to two orthogonal sub-planes α-β and xy. Suppose the switching function of the six-phase inverter is S = [S A S B S C S D S E S F ], taking phase A as an example, if S A =1, it means the switch on the inverter phase A bridge arm is turned on; S A =0 means the switch under the A-phase bridge arm is turned on. The six-bit switch corresponds to 64 voltage vectors, including 60 effective vectors and 4 zero vectors. According to the voltage vector amplitude, it can be divided into four groups, namely large vector D4, medium vector D3, medium and small vector D2, and small vector D1, with amplitudes of and Its mapping relationship between the α-β plane and the xy plane is as follows: Figure 2 (a) and (b), and the corresponding relationship of their amplitudes is shown in Table 1. The amplitude of the voltage vector of the dual three-phase permanent magnet synchronous motor.

[0068] Table 1

[0069]

[0070]

[0071] The present invention selects a voltage vector combination to perform multi-level hysteresis control, considers the positive and negative and amplitude of magnetic flux and torque at the same time, selects the best voltage vector group for voltage modulation, performs multi-level hysteresis direct torque control, and performs fault-tolerant operation for a single-phase open-circuit fault of a dual three-phase permanent magnet synchronous motor, and corrects the xy harmonic plane and α-β plane voltage modulation.

[0072] The control strategy block diagram proposed by the present invention is as follows: Figure 3 As shown, the subscript "ref" represents a reference value.

[0073] we is the motor electrical speed, T e is the electromagnetic torque, e represents the error, T A -T F is the vector action time duty cycle, i A -i F is the stator current, i s is the stator current, i z is the harmonic current. c and v o They are active vector group and zero vector group respectively.

[0074]

[0075] In the formula, ε ψ =ψ s，ref -ψ s , ψ T =T e，ref -T e +ω e ψ s , ΔT is the hysteresis width of torque control. The hysteresis width ΔT is reasonably selected according to the speed and torque and other working conditions for hysteresis control.

[0076] The harmonic plane uses a proportional resonant (PR) controller to control the harmonic plane current to achieve zero-error regulation of the harmonic current iz. During normal operation, the harmonic plane current should be suppressed to 0. In the event of a fault, it is used to suppress the 5th and 7th harmonic currents and track the given fault-tolerant operation reference current.

[0077] The transfer function of the PR controller in the discrete domain can be expressed as:

[0078]

[0079] The parameters are:

[0080]

[0081] k pr is the resonance gain; ω c is the cut-off frequency; ω0 is the resonant frequency. pr It can improve the controller amplitude-frequency gain and increase ω c The controller bandwidth can be increased.

[0082] The resonant frequency is ω e (ω e is the electrical angular frequency), 5ω e and 7ω e The PR controller is connected in parallel as an xy plane current controller, and closed-loop control is used to suppress the 5th and 7th current harmonics and track the fundamental frequency current reference value.

[0083] To realize planar linear voltage modulation, three adjacent voltage vectors are selected as a voltage vector group, including two adjacent D4 vectors and one D2 vector, to synthesize the output voltage, as shown in Figure 4 The black (bold) vectors in (a) and (b) are shown. The vector combination is represented by v c There are 12 such vector combinations, and the vector combinations and their positions in the xy plane are shown in Table 2v c1 Vector groups and positions are shown.

[0084] Table 2

[0085]

[0086]

[0087] When a single-phase open-circuit fault occurs in the dual three-phase PMSM, the current reference value of the harmonic plane is modified according to the principle of maximum torque or minimum copper loss for fault tolerance, as shown in Table 3, the open-phase fault position and the corresponding current constraint.

[0088] Table 3

[0089]

[0090] According to the position of the phase failure, select the corresponding minimum copper loss or maximum torque mode fault-tolerant current, add it to the harmonic plane current reference value, and use the PR controller to track the harmonic current reference value. Taking phase F as an example, the xy harmonic plane current reference value when using the minimum copper loss mode is:

[0091] i x，ref =0,i y，ref -i β (9)

[0092] When using the maximum torque mode, the current reference value is:

[0093] i x，ref =-i α ,i y，ref =-i β (10)

[0094] For the modulation of harmonic plane voltage and α-β plane voltage, a flux observer is used to observe the flux and torque of the motor to perform multi-level hysteresis closed-loop control. The flux observer uses a current observer to calculate the errors of flux and torque, and selects the voltage vector based on these two error values.

[0095] According to formulas (6) and (7), the current hysteresis width ΔT is set, and the most appropriate voltage vector v is selected according to the flux error and torque error. cand v0 are the active vector group and the zero vector group respectively, λ∈[0,1] is the adjustment coefficient, and the vector selection switch table is shown in Table 4ST-DTC switch table.

[0096] Table 4

[0097]

[0098] When e T ≠0, select v c Vector group, select 2 adjacent D4 vectors and 1 D2 vector to synthesize the output voltage, such as Figure 4 As shown, the corresponding voltage vector group is selected according to the flux linkage and torque error, as shown in Table 3, and then the required voltage vector is synthesized using the selected three voltage vectors.

[0099] The harmonic voltage reference value before the fault is 0, that is, u z，ref =0, when a single-phase open circuit fault occurs, the harmonic current reference value after the fault is modified, and the corresponding harmonic reference voltage u is generated through the PR controller z，ref , using two adjacent vectors of the harmonic voltages in the three selected voltages to synthesize u z，ref ,like Figure 5 shown.

[0100] When the motor is running normally z，ref = 0, only the α-β plane reference voltage vector needs to be synthesized. When a single-phase open circuit fault occurs in the motor, the xy plane harmonic voltage reference value u z，ref , as shown in formula (11).

[0101] v c1 T c1,1 +v c2 T c2,1 +v c3 T c3,1 =u z,ref (11)

[0102] according to Figure 4 The corresponding relationship of the voltage vectors shown in the figure, the voltage vector generated by the synthesized xy plane harmonic voltage vector in the α-β plane is:

[0103] v c1 T c1,1 +v c2 T c2,1 +v c3 T c3,1 =u αβ，1 (12)

[0104] The harmonic voltage after the fault is relatively large. When synthesizing the α-β plane voltage reference value, it is necessary to consider the impact of the synthesized xy plane harmonic voltage on the α-β plane and correct the voltage vector to be synthesized:

[0105] u′ αβ，ref =u αβ，ref -u αβ，1 (13)

[0106] Then, according to the mapping relationship of the voltage vector, the remaining time is used to synthesize the corrected α-β plane reference voltage u′ αβ，ref , and make the voltage vector of its composite voltage mapped in the xy plane zero, such as Figure 6 (a) and (b) in can be expressed as:

[0107] v c1 T c1，2 +v c2 T c2,2 +v c3 T c3,2 =u′ αβ，ref (14)

[0108] In this way, the total action time of the three vectors in one cycle can be obtained, which can be expressed as:

[0109]

[0110] Selecting a smaller λ can reduce the torque ripple without affecting the voltage modulation on the harmonic plane, but λ cannot be reduced infinitely.

[0111] When e T = 0, since the traditional zero vector v0 cannot realize the modulation of the xy harmonic plane voltage, it cannot well realize the modulation of the harmonic plane voltage after the fault, and cannot effectively suppress the harmonic current, the present invention provides a new vector selection scheme to replace the traditional zero vector such as Figure 7 shown.

[0112] The amplitude of the D1 vector is the smallest in the α-β plane and the largest in the xy plane. The two adjacent D1 vectors in the xy plane differ by 150°. The two vectors have similar effects on the harmonic plane, but have a smaller effect on the α-β plane. These two vectors are selected as a vector group and recorded as v0. There are also 12 groups of v0, such as Figure 7 The selection and synthesis of the voltage vector are the same as above. The xy harmonic plane reference voltage is also synthesized to realize the motor control in steady state, as shown in equation (16).

[0113] u o1 T o1 +v o2 T o2 =uz，ref (16)

[0114] Denote the six-dimensional switch column vector as v k =[S A S B S C S D S E S F ] T , so the vector action time is:

[0115] T 6P =T c1 v c1 +T c2 v c2 +T c3 v c3 (17)

[0116] or:

[0117] T 6P =T o1 v o1 +T o2 v o2 (18)

[0118] Where T 6P =[T A T B T C T D T E T F ] T .

[0119] Through the pulse width modulation (PWM) module, T 6P By comparing with the triangular carrier, the switching signal of the six-phase bridge arm of the VSC can be obtained as follows: Figure 8 shown.

[0120] The flux observer based on the model reference adaptive method (MRAS) is used to observe the flux of the motor and perform closed-loop control of the flux.

[0121] According to the voltage equation (2), it is converted into an equation in the dq coordinate system and written in matrix form as follows:

[0122]

[0123] The subscript letters in the formula represent the parameters of the variables in their corresponding planes.

[0124] Similarly, according to the magnetic flux equation (4), it can be rewritten as:

[0125]

[0126] Where, L d and L q are the d-axis and q-axis stator direct-axis inductance and quadrature-axis inductance, respectively, L aal is the leakage self-inductance of each phase stator winding, and is the flux amplitude of the permanent magnet.

[0127] Using the basic principle of model reference adaptive control, a reference model and an adjustable model of flux are constructed to observe the flux of the motor. Therefore, the stator flux in the dq coordinate system is selected as the state variable, and the reference model can be expressed as:

[0128]

[0129] In the formula, u′ q =u q , ψ d , q is the stator flux of d and q axes, u d 、u q are the d-axis and q-axis stator voltages respectively.

[0130] Replacing the flux and speed in equation (22) with estimated values, we get the adjustable model of MRAS, which can be expressed as:

[0131]

[0132] In the formula, are the estimated values ​​of the stator flux of the d and q axes, Estimated value of electrical speed.

[0133] Defining the error vector Subtracting the reference model from the adjustable model yields:

[0134]

[0135] The above formula can be simplified as:

[0136]

[0137] In the formula,

[0138] According to stability theory, the necessary and sufficient condition for system stability is that the forward path is a strictly positive real transfer function matrix. Fig. 9 As shown, the transfer function of its forward path can be expressed as:

[0139] H(s)=(sI-A) -1 I (25)

[0140] The calculation result is:

[0141]

[0142] Its characteristic equation is:

[0143]

[0144] Solving for its extreme point yields:

[0145]

[0146] From formula (28), we can see that the poles of H(s) are all located in the left half plane of s, the system is stable, and it can be proved that H(jω)+H * (jω)>0. From the stability theory, we know that the system H(s) is strictly positive definite, that is, the system is stable and convergent.

[0147] The estimated stator current is obtained by using the estimated flux linkage, and the difference between the estimated stator current and the actual value of the motor current forms a closed-loop feedback link. The structure diagram of the flux linkage observer is shown in Fig.10 shown.

[0148] G is the feedback matrix. When G = 0, it is equivalent to an open loop observer. A reasonable value of G is selected to ensure its stability. Generally, the value of the feedback gain can be:

[0149] G=R s (29)

[0150] According to stability theory, when the current estimate converges to the actual current value, it means that the adjustable model tends to the reference model, and the estimated flux can be considered equal to the actual motor flux. In this way, the observed flux can be obtained for fault-tolerant operation of direct torque control.

[0151] The embodiments described above are only descriptions of the preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Without departing from the design spirit of the present invention, various modifications and improvements made to the technical solutions of the present invention by ordinary technicians in this field should all fall within the protection scope determined by the claims of the present invention.

Claims

1. A direct torque control method for a dual three-phase motor with single-phase open-circuit fault tolerance, characterized in that: The following steps are involved: By using multi-stage hysteresis loops and considering the positive and negative values ​​and amplitudes of torque and flux linkage, different voltage vector combinations are selected in dynamic and steady states to perform voltage modulation in the α-β plane and xy harmonic plane, and fault-tolerant operation is performed for single-phase open-circuit faults of dual three-phase permanent magnet synchronous motors. The voltage modulation in the xy harmonic plane and α-β plane is corrected to realize direct torque control of multi-stage hysteresis loops of dual three-phase permanent magnet synchronous motors. Methods for performing voltage modulation in the α-β plane and the xy harmonic plane include: For the modulation of harmonic plane voltage and α-β plane voltage, the flux observer is used to observe the flux of the motor, and multi-level hysteresis closed-loop control is performed on the flux and torque. The flux observer uses a current observer to calculate the errors of the flux and torque, and the voltage vector is selected based on these two error values. The method of using the model reference adaptive method to design a flux observer based on the stator flux as a state variable to identify the flux of the motor and perform closed-loop control of the flux includes: Using the basic principle of model reference adaptive control, a reference model and an adjustable model of flux linkage are constructed to observe the flux linkage of the motor; Define the error vector and make a difference between the reference model and the adjustable model; According to the stability theory, the transfer function of the forward path is obtained; Substituting the difference between the reference model and the adjustable model into the transfer function to obtain an estimated flux linkage; The estimated stator current is obtained by using the estimated flux, and the difference between the estimated stator current and the actual value of the motor current forms a closed-loop feedback link to achieve closed-loop control of the flux. According to the formula e ψ 、e T is the hysteresis output, ε ψ is the flux error, ε T For torque error, set the hysteresis width Δ T , and select the most appropriate voltage vector according to the flux error and torque error, v c and v0 are the active vector group and the zero vector group respectively, λ∈[0,1] is the adjustment coefficient, when e T ≠0, select v c Vector group, select 2 adjacent D4 vectors and 1 D2 vector to synthesize the output voltage, then select the corresponding voltage vector group according to the flux linkage and torque error, and then use the selected 3 voltage vectors to synthesize the required voltage vector; among them, the D4 vector has an amplitude of The vector of D2 is the magnitude of vector of; The harmonic voltage reference value before the fault is 0, that is, u z，ref =0, when a single-phase open circuit fault occurs, the harmonic current reference value after the fault is modified, and the corresponding harmonic reference voltage u is generated through the PR controller z，ref , using two adjacent vectors of the harmonic voltages in the three selected voltages to synthesize u z，ref ; When the motor is running normally z，ref = 0, only the α-β plane reference voltage vector needs to be synthesized, and when a single-phase open circuit fault occurs in the motor, the xy plane harmonic voltage reference value u is synthesized z，ref , as shown in formula (11): v c1 T c1，1 +v c2 T c2，1 +v c3 T c3，1 =u z，ref (11) According to the corresponding relationship of the voltage vector, the voltage vector generated by the synthesized xy plane harmonic voltage vector in the α-β plane is: v c1 T c1，1 +v c2 T c2，1 +v c3 T c3，1 =u αβ，1 (12) The harmonic voltage after the fault is relatively large. When synthesizing the α-β plane voltage reference value, it is necessary to consider the impact of the synthesized xy plane harmonic voltage on the α-β plane and correct the voltage vector to be synthesized: in' αβ，ref =in αβ，ref -in αβ，1 (13) Then, according to the mapping relationship of the voltage vector, the remaining time is used to synthesize the corrected α-β plane reference voltage u′ αβ，ref , and make the voltage vector of its synthetic voltage mapped in the xy plane zero, expressed as: v c1 T c1,2 +v c2 T c2,2 +v c3 T c3,2 =u′ αβ，ref (14) The total action time of the three vectors in one cycle is obtained, which is expressed as: Selecting a smaller λ to reduce torque ripple will not affect the voltage modulation on the harmonic plane, but λ cannot be reduced infinitely; When e T = 0, the amplitude of the D1 vector is the smallest in the α-β plane and the largest in the xy plane, and the two adjacent D1 vectors in the xy plane differ by 150°. The two vectors have similar effects on the harmonic plane, but have a smaller effect on the α-β plane. These two vectors are selected as a vector group and recorded as v0. There are also 12 groups of v0. Among them, the amplitude of the D1 vector is The vector of the synthesized xy harmonic plane reference voltage realizes the motor control in steady state, as shown in formula (16): v o1 T o1 +v o2 T o2 =u z，ref (16) Denote the six-dimensional switch column vector as v k =[S A S B S C S D S E S F ] T , where S A To S F are the on-off states of the switches on the inverter phases A to F, respectively. The vector action time is: T 6P =T c1 v c1 +T c2 v c2 +T c3 v c3 (17) or: T 6P =T o1 v o1 +T o2 v o2 (18) Where T 6P =[T A T B T C T D T E T F ] T ; Among them, T A -T F is the duty cycle of the vector action time; Through the pulse width modulation PWM module, T 6P The switching signals of the six-phase bridge arms of the converter are obtained by comparing with the triangular carrier; The flux observer of the model reference adaptive method MRAS is used to observe the flux of the motor and perform closed-loop control of the flux. According to the voltage equation Among them, u s is the stator voltage, R s is the stator resistance, i s is the stator current, ψ s is the stator flux, i z is the harmonic current, which is converted into an equation in the dq coordinate system and written in matrix form as follows: The subscript letters in the formula represent the parameters of the variables in their corresponding planes; Similarly, according to the magnetic flux equation Among them, L z is the stator winding leakage inductance, θ0 is the electrical angle of the rotor position, which can be rewritten as: Where, L d and L q are the d-axis and q-axis stator direct-axis inductance and quadrature-axis inductance, respectively, L aal is the leakage self-inductance of each phase stator winding, ψ f is the flux amplitude of the permanent magnet; Using the basic principle of model reference adaptive control, a reference model and an adjustable model of flux are constructed to observe the flux of the motor; therefore, the stator flux in the dq coordinate system is selected as the state variable, and the reference model is expressed as: In the formula, u′ q =u q , ψ d , q is the stator flux of d and q axes, u d 、u q are the d-axis and q-axis stator voltages respectively; Replacing the flux and speed in equation (21) with estimated values, we get the adjustable model of MRAS, which is expressed as: In the formula, are the estimated values ​​of the stator flux of the d and q axes, is the estimated value of electrical speed; Defining the error vector in, is the estimated value of the stator flux, and the difference between the reference model and the adjustable model is: The above formula can be simplified as: In the formula, The transfer function of the forward path is expressed as: H(s)=(sI-A) -1 I (25) The calculation result is: Its characteristic equation is: Solving for its extreme point gives: From formula (28), we can see that the poles of H(s) are all located in the left half plane of s, the system is stable, and it is proved that H(jω)+H * (jω)>0; From the stability theory, we know that the system H(s) is strictly positive definite, that is, the system is stable and convergent; The estimated stator current is obtained by using the estimated flux linkage, and the difference between the estimated stator current and the actual value of the motor current is used to form a closed-loop feedback link; G is the feedback matrix. When G = 0, it is equivalent to an open loop observer. A reasonable value of G is selected to ensure its stability. The value of the feedback gain is: G=R s (29) According to the stability theory, when the current estimate converges to the actual current value, it means that the adjustable model tends to the reference model, and the estimated flux is considered to be equal to the actual motor flux, and the observed flux is obtained for fault-tolerant operation of direct torque control; The method for fault-tolerant operation of a dual three-phase permanent magnet synchronous motor with a single-phase open-circuit fault includes: After a single-phase open-circuit fault occurs in a dual three-phase permanent magnet synchronous motor, a full-dimensional model of the dual three-phase permanent magnet synchronous motor after the fault is used to select the corresponding minimum copper loss or maximum torque mode fault-tolerant current according to the position of the broken phase, and add it to the harmonic plane current reference value. The PR controller is used to track the harmonic current reference value to achieve single-phase open-circuit fault-tolerant operation.

2. The direct torque control method for single-phase open-circuit fault tolerance of dual three-phase motors according to claim 1, characterized in that: The transfer function of the proportional resonant PR controller in the discrete domain can be expressed as: The parameters are: k pr is the resonance gain; ω c is the cut-off frequency; ω0 is the resonant frequency; T s is the stator torque.

Images