Permanent magnet synchronous motor model predictive torque control method based on hybrid mode optimization

Abstract

The application discloses a permanent magnet synchronous motor model prediction torque control method based on a hybrid mode optimization. The method first constructs a reference voltage vector through a torque error target function and a discretization prediction model of the permanent magnet synchronous motor, and then designs a hybrid mode optimization algorithm. The algorithm first designs a double-vector mode optimization on the sector boundary of a voltage space vector diagram, and then designs a three-vector optimization mode based on a discrete space vector modulation inside the sector. The optimal synthesized voltage vector obtained by the hybrid mode optimization algorithm is transmitted to a voltage error target function to realize the switching of the two modes, so that the control execution mode and the optimal synthesized voltage vector are determined. The control method disclosed by the application has high freedom to adjust the amplitude and position of the synthesized voltage vector, and thus has high steady-state performance of the control system and fast torque dynamic response in the full speed range.

Description

Technical Field

[0001] This invention relates to the field of model predictive torque control technology for permanent magnet synchronous motors, and in particular to a model predictive torque control method for permanent magnet synchronous motors based on hybrid mode optimization. Background Technology

[0002] Permanent magnet synchronous motors (PMSMs) have been widely used in electric vehicles and robotics. Due to their high power density, high reliability, and compact structure, PMSMs are widely applied in these fields. However, the control requirements for the motor vary significantly across different applications. In the automotive industry, rapid dynamic response and a wide speed range are crucial indicators for evaluating vehicle performance. Conversely, high precision, high reliability, and rapid dynamic response are key standards for robot performance. In PMSM drive systems, the quality of the control method directly determines the overall system performance.

[0003] Previous research has shown that model predictive torque control (MMTC) for permanent magnet synchronous motors faces many challenges, such as large torque ripple, computational complexity for microprocessors, and high switching frequencies. One reason for the large torque ripple is that traditional MMTC uses only one switching vector applied to the converter per sampling cycle. One solution is to use multiple vectors applied to the power electronic converter within a single sampling cycle. For example, the dual-vector MMTC method applies multiple voltage vectors to the power electronic converter in a single sampling cycle. The specific details of dual-vector MMTC are as follows: First, it is assumed that a sector in the voltage space vector diagram has two effective vectors and one zero vector. The voltage vector can be a combination of one effective vector and the zero vector, or a combination of two effective vectors. Therefore, there are three possible backup voltage combinations. In the voltage space vector diagram, the triangular boundary formed by the three basic vectors contains the three backup voltage combinations. In particular, the composite voltage vector has a large amplitude on the outermost line segment. Generally, this combination may only be used when the rotor is running at high speed or under sudden change conditions. Furthermore, due to the three backup voltage combinations, the position of the end of the composite voltage vector is restricted to the triangular boundary. This will result in a relatively large deviation between the composite vector and the reference vector. This will cause the control target to "spurt", that is, the composite vector deviates from the reference vector to a large extent in a certain operating state, which reduces the system control accuracy. Therefore, the dual-vector model predictive torque control method lacks the degree of freedom of the composite voltage amplitude. Summary of the Invention

[0004] The problem this invention aims to solve is the reduced control accuracy caused by the lack of synthetic voltage vector degrees of freedom in dual-vector model predictive torque control (PMSM). This invention proposes a hybrid-mode optimization-based PMSM model predictive torque control method. The method first discretizes the voltage equation, flux linkage equation, and torque equation of the PMSM in continuous time, thereby establishing a discretized predictive model of the PMSM. Based on the discretized predictive model of the PMSM and the deadbeat principle, the reference voltage vector for the next moment is calculated in real time. Then, a hybrid-mode optimization algorithm is designed. This algorithm executes a dual-vector optimization mode when the reference voltage vector is close to the sector boundary of the voltage space vector diagram, and a three-vector optimization mode based on discrete space vector modulation when the reference voltage vector is far from the sector boundary and inside the sector. Finally, the three voltage vectors obtained by the hybrid-mode optimization algorithm are passed to a voltage error-based objective function that enables mode switching, thereby determining the optimal voltage vector. Finally, the duty cycle corresponding to the optimal voltage vector is converted into a switching pulse signal. To solve the above technical problem, the technical solution adopted in this invention is a hybrid-mode optimization-based PMSM model predictive torque control method, which includes the following steps:

[0005] Step S1. Real-time acquisition of motor stator current, rotor speed and motor electrical angle, and construction of reference voltage vector based on the objective function of torque error and the discretized prediction model of permanent magnet synchronous motor;

[0006] Step S2. Generate a dual-vector synthesized voltage based on a dual-vector mode and a virtual vector voltage based on a three-vector mode according to the reference voltage vector and the preset hybrid mode optimization algorithm;

[0007] Step S3. Based on the voltage error objective function, switch between the dual-vector mode and the three-vector mode to output the optimal voltage vector, and convert the duty cycle corresponding to the optimal voltage vector into a switching pulse signal to be applied to the two-level inverter.

[0008] As a preferred embodiment of the present invention, the step S1 of constructing the reference voltage vector based on the objective function of torque error and the discretized prediction model of the permanent magnet synchronous motor is specifically as follows:

[0009] The formula for calculating the reference voltage vector is constructed based on the objective function of torque error and the discretized prediction model of surface-mounted PMSM. In the two-phase dq rotating coordinate system, the specific structure is as follows:

[0010]

[0011] in,

[0012]

[0013] In the formula, R is the stator resistance and L is the armature winding inductance. For a permanent magnet motor with constant flux linkage, ω e This refers to the rotor speed, expressed in rad / s. For reference electromagnetic torque, p n For extreme logarithms, T s The sampling period of the discrete control system. and The direct-axis and quadrature-axis components of the stator current at time k. and The direct-axis and quadrature-axis components of the stator current are predicted one step ahead at time k+1.

[0014] Construct a reference voltage vector V based on the αβ coordinate system s * It can be expressed by the following formula:

[0015]

[0016] In the formula, θ is the electrical angle of the motor.

[0017] As a preferred embodiment of the present invention, step S2, the step of generating a dual-vector synthesized voltage based on a dual-vector mode according to the reference voltage vector and the preset hybrid mode optimization algorithm, specifically includes generating a dual-vector synthesized voltage based on a dual-vector mode according to the hybrid mode optimization algorithm, calculating the corresponding duty cycle, and establishing a pre-selection rule for the dual-vector synthesized voltage, specifically,

[0018] When the reference voltage vector Located within the first sector, the first sector is defined as including four voltage vectors V1, V2, V0, and V7. V1 and V2 are the basic voltage vectors provided by the inverter, while V0 and V7 are both zero vectors. The first sector is divided into three regions Z1, Z2, and Z3. When the reference voltage vector... When located in different regions, the dual vector voltage is selected according to the preset pre-selection rules.

[0019] As a preferred embodiment of the present invention, the calculation process of the dual-vector duty cycle in the three regions Z1, Z2, and Z3 is as follows:

[0020] When the reference voltage vector is located in region Z1, voltage vectors V1 and V0 are selected for synthesizing a dual-vector voltage. The optimal duty cycle is calculated using the following formula:

[0021]

[0022] Where d 11 Let d be the duty cycle of the voltage vector V1. 12 Let be the duty cycle of the voltage vector V0, and d12 =1-d 11 ,

[0023] The duty cycle d of the voltage vector V1 11 The calculation process is as follows:

[0024]

[0025] When the reference voltage vector is located in the Z2 region, voltage vectors V1 and V2 are selected for synthesizing a dual-vector voltage. The optimal duty cycle is calculated using the following formula:

[0026]

[0027] Where d 21 Let d be the duty cycle of the voltage vector V1. 22 Let be the duty cycle of the voltage vector V2, and d 22 =1-d 21 ;

[0028] The duty cycle d of the voltage vector V1 21 The calculation process is as follows:

[0029]

[0030] When the reference voltage vector is located in region Z3, voltage vectors V2 and V7 are selected for synthesizing a dual-vector voltage. The optimal duty cycle is calculated using the following formula:

[0031]

[0032] Where d 31 Let d be the duty cycle of the voltage vector V2. 32 Let be the duty cycle of the voltage vector V7, and d 32 =1-d 31 ;

[0033] The duty cycle d of the voltage vector V2 31 The calculation process is as follows:

[0034]

[0035] As a preferred embodiment of the present invention, the step of establishing the pre-selection rule for the dual-vector synthesized voltage and selecting the dual-vector voltage according to the preset pre-selection rule specifically includes: setting J1 = |V *s -V1|,J2=|V s * -V2|,J3=|V s * -v0|, the pre-selection rules are as follows:

[0036]

[0037] Where (d) 11 ,d 12 ),(d 21 ,d 22 ), (d 21 ,d 22 The numbers (V1, V0), (V1, V2), and (V2, V7) represent the duty cycles of V1, V0, V1, V2, and V7, respectively.

[0038] As a preferred embodiment of the present invention, step S2, the step of generating a virtual vector voltage based on a three-vector mode, includes synthesizing a virtual vector based on discrete space vector modulation, wherein the synthesized virtual vector voltage V vir The calculation formula is as follows:

[0039]

[0040] In the formula, d j V represents the true vector duty cycle, N represents the interval divided into each sampling period, and V represents the true vector duty cycle. j real Represents the actual voltage vector;

[0041] When the reference voltage vector V s * When located in the first sector, the sector is divided into six sub-regions, among which, the virtual voltage vector Located at the midpoint inside an equilateral triangle, three virtual voltage vectors The midpoint of the line segment between each vertex of an equilateral triangle sector and the midpoint of the triangle is expressed by the following formula:

[0042]

[0043] The coefficients of V1, V2, and V0 are the duty cycles of the corresponding virtual voltage vectors;

[0044] The duty cycle of the virtual voltage vector at the corresponding position is shown below:

[0045]

[0046] Where d i,j Let represent the duty cycle of the three basic voltage vectors that form the composite virtual voltage vector, where i represents the index of the three basic voltage vectors V1, V2, and V0, and i∈{1,2,3}, and j represents the virtual voltage vector. The lower left index, and j∈{8,9,10,11}.

[0047] The virtual voltage vector is selected according to the aforementioned pre-selection rules, specifically as follows:

[0048]

[0049] 8. As a preferred embodiment of the present invention, the step S3 of switching between the dual-vector mode and the three-vector mode based on the voltage error objective function to output the optimal voltage vector specifically involves generating a dual-vector composite voltage. and virtual vector voltage As a candidate output voltage V out The optimal voltage vector V is obtained by solving the objective function that minimizes the voltage error. opt This allows for switching between the two operating modes. The solution process is as follows:

[0050]

[0051] The global optimal voltage vector V opt The corresponding duty cycle is converted into a switching pulse signal that acts on the two-level inverter.

[0052] The beneficial effects of this invention are as follows: This invention discloses a model predictive torque control method for permanent magnet synchronous motors (PMSMs) based on hybrid mode optimization. It designs a dual-vector mode optimization at the boundary of a triangular sector, providing three duty cycles. Inside the sector, a three-vector mode optimization based on discrete space vector modulation is added to increase the available duty cycles. Therefore, this control method has variable duty cycles and a high degree of freedom to adjust the amplitude and position of the synthesized voltage vector, more accurately tracking the voltage vector, thus achieving more accurate control of torque, flux linkage, and stator current, reducing ripple, and meeting the high control accuracy requirements of PMSMs. Furthermore, the duty cycle calculation of the virtual voltage vector in the three-vector mode optimization based on discrete space vector modulation disclosed in this invention is simple, specifically manifested as the equal division of the action time of each basic voltage vector used to synthesize the virtual voltage vector. It has good scalability; users can adjust the number of vector segment intervals according to the control accuracy requirements to improve the algorithm's control accuracy. Attached Figure Description

[0053] Figure 1 This invention relates to a two-level inverter servo surface-mounted permanent magnet synchronous motor AC drive system.

[0054] Figure 2 This is a block diagram of the predicted torque control model of the permanent magnet synchronous motor based on hybrid mode optimization as described in this invention.

[0055] Figure 3 This is a flowchart of the algorithm for predictive torque control of a permanent magnet synchronous motor model based on hybrid mode optimization, as described in this invention.

[0056] Figure 4Vector combination optimized for dual-vector mode in this invention

[0057] Figure 5 This is a schematic diagram showing the reference voltage vector located in sector I under the dual-vector mode optimization described in this invention.

[0058] Figure 6 This is a schematic diagram of the virtual voltage vector position when the vector line segment described in this invention is divided into two intervals.

[0059] Figure 7 This is a schematic diagram of virtual voltage vector pre-selection when the vector line segment is evenly divided into two intervals, as described in this invention.

[0060] Figure 8 This is a schematic diagram of the virtual voltage vector position when the vector line segment described in this invention is divided into three intervals.

[0061] Figure 9 This is a schematic diagram of virtual voltage vector pre-selection when the vector line segment is divided into three intervals, as described in this invention. Detailed Implementation

[0062] The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and should not be construed as limiting the present invention.

[0063] The present invention discloses a model predictive torque control method for permanent magnet synchronous motors based on hybrid mode optimization. The application objects of this invention are as follows: Figure 1 As shown, the method disclosed in this invention is a control algorithm for the AC drive system of a surface-mounted permanent magnet synchronous motor serving as a two-level voltage source inverter. The purpose is to improve the control accuracy of torque, flux linkage, stator current and other parameters based on the existing dual-vector model predictive torque control algorithm.

[0064] Figure 2 This is the block diagram of the predicted torque control model of the permanent magnet synchronous motor (PMSM) based on hybrid mode optimization as described in this invention. Real-time acquisition of the PMSM's electrical angle, rotor speed, and stator current is used as the reference voltage vector V. s * Calculation of torque reference value. It is obtained from the speed regulation loop, that is, by transmitting the error between the given reference speed and the measured speed to the proportional-integral element. Magnetic flux linkage reference value Calculated based on the maximum torque-to-current ratio (MPTA) formula. Figure 2 In this context, N represents the number of evenly spaced vector line segments. (At the torque reference value...) Magnetic flux reference value Rotor speed ω e The electric angle θ of the motor and the stator current I in the dq coordinate system dqUnder the influence of torque error, a formula for calculating the reference voltage vector is constructed using an objective function based on torque error and a discretized prediction model of the permanent magnet synchronous motor. Assume the inverter output voltage V... s It can accurately track the reference voltage vector V s * This enables tracking of torque reference values, stator current, and flux linkage reference values. The hybrid mode optimization algorithm and the objective function based on voltage error aim to synthesize the optimal inverter output voltage V. opt To achieve the reference voltage vector V s * Tracking, V opt The duty cycle is converted into switching pulses that act on the two-level inverter, thereby controlling the torque, stator current, and flux linkage.

[0065] Figure 3 The flowchart illustrates the model predictive torque control method for permanent magnet synchronous motors based on hybrid mode optimization according to the present invention, combined with... Figure 2 and Figure 3 The specific implementation plan is explained in detail below. First, the rotor speed and electrical angle of the surface-mounted permanent magnet synchronous motor are acquired in real time, and the stator current of the permanent magnet synchronous motor is acquired in real time using a current sensor. Then, the real-time acquired data is transmitted to a reference voltage vector calculation formula based on a discretized torque prediction model, a discretized stator current prediction model, and a discretized torque error objective function, thereby calculating the reference voltage vector in real time. Subsequently, a hybrid mode optimization algorithm is designed, specifically in two steps: the first step is to design a dual-vector mode optimization algorithm on the sector boundaries of the voltage space vector diagram; the second step is to design a three-vector mode optimization algorithm based on discrete space vector modulation within the sector. Figure 2 Taking a vector line segment with two intervals as an example, the hybrid mode optimization algorithm can obtain three voltage vectors and pass them to the voltage error-based objective function to realize mode switching, thereby determining the optimal voltage vector; finally, the duty cycle corresponding to the optimal voltage vector is converted into a switching pulse signal. The specific implementation process of this invention is as follows:

[0066] Step S1. Real-time acquisition of motor stator current, rotor speed and motor electrical angle, and construction of reference voltage vector based on the objective function of torque error and the discretized prediction model of permanent magnet synchronous motor.

[0067] In the two-phase dq rotating coordinate system, the mathematical expression (1) of the voltage equation of the surface-mounted PMSM is as follows:

[0068]

[0069] The mathematical expression (2) of the flux linkage equation is as follows:

[0070]

[0071] The mathematical expression (3) of the electromagnetic torque equation is as follows:

[0072]

[0073] Among them, V d With V q Stator voltage V s The components on the dq axis, i d with i q stator current i abc After the Clarke transformation and Parker transformation, the three-phase abc stationary coordinate system is transformed into the quadrature and direct axis components in the two-phase dq rotating coordinate system. R is the stator resistance, and L is the armature winding inductance. For a permanent magnet motor with constant flux linkage, ω e This refers to the rotor speed, expressed in rad / s. For reference electromagnetic torque, p n It is an extreme logarithm.

[0074] In digital control systems, by discretizing equation (1) using the forward Euler theorem, the mathematical expression (4) of the stator current prediction equation in the two-phase dq rotating coordinate system can be obtained as follows:

[0075]

[0076] The mathematical expression (5) for the stator flux linkage prediction equation is as follows:

[0077]

[0078] The mathematical expression (6) for the electromagnetic torque prediction equation is as follows:

[0079]

[0080] T s The sampling period of the discrete control system. and The direct-axis and quadrature-axis components of the stator current at time k. and The direct-axis and quadrature-axis components of the stator current are predicted one step ahead at time k+1. and This is the estimated value of the stator voltage at time k. and The direct and quadrature components of the magnetic flux are predicted for the next step forward at time k+1.

[0081] In digital control systems, there is a command delay, so delay compensation is performed. Therefore, the stator current prediction equation (7) after delay compensation is as follows:

[0082]

[0083] The stator flux prediction equation (8) after time delay compensation is as follows:

[0084]

[0085] The electromagnetic torque prediction equation (9) after time delay compensation is as follows:

[0086]

[0087] in,

[0088]

[0089] The mathematical expression (10) for the electromagnetic torque error objective function of model predictive torque control is as follows:

[0090]

[0091] in, This is the torque reference value.

[0092] Substituting the electromagnetic torque formula (3) and prediction model (9) into (10), the torque error objective function (11) is as follows:

[0093]

[0094] in This is the reference current.

[0095] make It is zero, achieved based on the principle of zero beat rate. This can be achieved Based on the torque error objective function (10), the stator current prediction model (7), and the torque prediction model (3), the reference voltage vector calculation equation (12) is as follows:

[0096]

[0097] Among them, X1, X2, and X3 are consistent with those in (7).

[0098] Transform the reference voltage vector into a two-phase stationary αβ coordinate system:

[0099]

[0100] Where θ is the electrical angle of the motor, and V s * This is the reference voltage vector in the αβ coordinate system.

[0101] The above formula is the calculation formula for the reference voltage vector. The reference voltage vector V for the corresponding operating state can be calculated in real time by acquiring data in real time. s * Assume the inverter output voltage V s It can accurately track the reference voltage vector V s * This enables both torque reference value tracking and flux linkage reference value tracking. Based on the maximum torque-to-current ratio (MPTA) formula (14), we can derive:

[0102]

[0103] At this time, the inverter output voltage V s Under the influence of the prediction models (7) and (8), the results calculated by these models are as follows: Able to achieve magnetic flux reference value track.

[0104] Step S2. Generate a dual-vector synthesized voltage based on the dual-vector mode and a virtual vector voltage based on the three-vector mode according to the reference voltage vector and the preset hybrid mode optimization algorithm.

[0105] In this embodiment, the implementation of this scheme using a hybrid mode optimization algorithm includes two main steps:

[0106] Step 1: Dual-vector mode optimization design. Figure 4 The vector combinations optimized for the dual-vector mode are as follows: each of the three boundaries within a sector corresponds to a dual-vector voltage vector combination, resulting in a total of twelve dual-vector combinations. Figure 5 With reference voltage vector V s * Taking the first sector as an example, it contains four basic voltage vectors: V1, V2, V0, and V7. V1 and V2 are the basic voltage vectors provided by the inverter, while V0 and V7 are both zero vectors, used alternately to reduce switching frequency. The sector is divided into three regions: Z1, Z2, and Z3. Then, V1 and V0 are selected to synthesize a dual-vector voltage when the reference voltage vector is in region Z1; V1 and V2 are selected when the reference voltage vector is in region Z2; and V2 and V7 are selected when the reference voltage vector is in region Z3. Each vector lies on one of the three boundaries of a triangle. The pre-selection of the optimal dual-vector voltage vector is equivalent to the reference voltage vector V... s * In determining the location of the three regions Z1, Z2, and Z3.

[0107] When the reference voltage vector is located in the Z1 region, V1 is selected; V0 is used to synthesize a dual-vector voltage. The optimal duty cycle is calculated using the following formula:

[0108]

[0109] Where d 11 Let d be the duty cycle of V1. 12 Let be the duty cycle of V0, and d 12 =1-d 11 ,

[0110] Calculate the duty cycle d of voltage vector V1. 11 as follows:

[0111] d 11 =|V s * ·V1| / V1 2 (16)

[0112] When the reference voltage vector is located in the Z2 region, V1 and V2 are selected to synthesize the dual-vector voltage. The optimal duty cycle is calculated using the following formula:

[0113]

[0114] Where d 21 Let d be the duty cycle of V1. 22 Let be the duty cycle of V2, and d 22 =1-d 21 ;

[0115] From this, the duty cycle d of the voltage vector V1 can be calculated. 21 As follows:

[0116] d 21 =1 / 2+|V s * ·(V1+V2)| / V1 2 (18)

[0117] When the reference voltage vector is located in region Z3, V2 and V7 are selected for synthesizing the dual-vector voltage. The optimal duty cycle is calculated using the following formula:

[0118]

[0119] Where d 31 Let d be the duty cycle of V2. 32 Let d be the duty cycle of V7, and d 32 =1-d 31 ;

[0120] The duty cycle d of voltage vector V2 can be calculated from this. 31 As follows:

[0121]

[0122] Subsequently, J1 = |V s * -V1|,J2=|V s * -V2|,J3=|V s * -V0|, when When located in region Z1, J1 = max{J1, J2, J3}, and the optimal combination is V2 V7. When located in region Z2, J3 = max{J1, J2, J3}, and the optimal combination is V1 V2. When located in region Z3, J2 = max{J1, J2, J3}, and the optimal combination is V1 V0.

[0123] The corresponding value can be obtained through three calculations. The present invention provides a pre-selection rule for a dual-vector voltage vector as follows:

[0124]

[0125] Where (d) 11 ,d 12 ),(d 21 ,d 22 ), (d 21 ,d 22 The numbers (V1, V0), (V1, V2), and (V2, V7) represent the duty cycles of V1, V0, V1, V2, and V7, respectively.

[0126] Step 2: Three-vector mode optimization design based on Discrete Space Vector Modulation (DSVM). A DSVM technique is introduced to synthesize a virtual voltage vector. The formula for calculating the synthesized virtual vector is as follows:

[0127]

[0128] Where t j This represents the duty cycle of the true vector, and N represents the interval divided into each sampling period. V represents the true voltage vector; vir This is a virtual voltage vector. Extensive literature confirms that dividing the sampling period into three equal parts, while increasing the computational load reasonably, can effectively improve system control performance.

[0129] In the method disclosed in this invention, the concept of equal division of sampling period in discrete space voltage vector modulation technology is used to synthesize virtual voltage vector. The equal division of vector line segments in the voltage space vector diagram is equivalent to the equal division of the corresponding action time of each basic voltage vector used to synthesize the virtual voltage vector.

[0130] In this invention, N represents the number of evenly divided intervals of vector line segments. Virtual voltage vectors are synthesized at the center of each equilateral triangle sector in the voltage space vector diagram and on the line segments between each vertex and the center of the triangle. Figure 6 This represents the location of all virtual voltage vectors when the vector segment is evenly divided into two intervals. In the voltage space vector diagram, there are 24 virtual voltage vectors. Assuming a reference voltage vector... Located in the first sector, such as Figure 7 As shown, the sector is divided into six sub-regions. Located at the midpoint inside an equilateral triangle, three virtual voltage vectors The location is at the midpoint of the line segment between each vertex and the midpoint of the triangle. The four virtual voltage vectors are as follows:

[0131]

[0132] The coefficients of V1, V2, and V0 are the duty cycles of the corresponding virtual voltage vectors.

[0133] Therefore, the duty cycle used to synthesize the virtual voltage vector is as follows:

[0134]

[0135] Where d i,j Let represent the duty cycle of the three basic voltage vectors that form the composite virtual voltage vector, where i represents the index of the three basic voltage vectors V1, V2, and V0, and i∈{1,2,3}, and j represents the virtual voltage vector. The lower left index, and j∈{8,9,10,11}

[0136] This invention takes the vector line segment being divided into two equal intervals, i.e., N=2, as an example. Figure 6 Each black dot represents the position of the virtual voltage vector. Located at the midpoint inside an equilateral triangle, three virtual voltage vectors The location is the midpoint of the line segment between each vertex and the midpoint of the triangle. The reference voltage vector has already been determined during the optimal dual-vector voltage vector pre-selection process. Located at positions Z1, Z2, and Z3. Each region Z1, Z2, and Z3 is divided into two sub-regions by line segments from the midpoint of the triangle to the midpoint of each side. Each sub-region contains two virtual voltage vectors. This invention re-divides each sector in the voltage space vector diagram into six sub-regions. Therefore, the pre-selection of the optimal virtual voltage vector is equivalent to determining the position of the reference voltage vector in regions R1, R2, ..., R6 within each sector.

[0137] The first step involves three calculations to obtain the values ​​of J1, J2, and J3. When When located in the region between R1 and R6, either J2>J1>J3 or J1>J2>J3, and the optimal virtual voltage vector is: when When located in the region between R2 and R3, either J1>J3>J2 or J3>J1>J2, the optimal virtual voltage vector is... when When located in the region between R4 and R5, either J3>J2>J1 or J2>J3>J1, and the optimal virtual voltage vector is...

[0138] Based on this, the present invention provides a virtual voltage vector pre-selection scheme to reduce the amount of computation. The pre-selection rules are as follows:

[0139]

[0140] The three-vector optimization mode design based on discrete space vector modulation has good scalability. Users can increase the number of vector segment intervals to improve control accuracy according to the required control precision. If the number of vector segment intervals is three, i.e., N=3, then there are two virtual voltage vectors between each vertex and the midpoint of the triangular sector, such as... Figure 8 The points shown represent virtual voltage vectors divided into three equal intervals by a vector line segment. There are 42 virtual voltage vectors in the voltage space vector diagram. Again, taking the reference voltage vector located in the first sector as an example... Figure 9 The vector line segments between each vertex and the midpoint of the equilateral triangular sector shown are divided into three equal parts by two virtual voltages. At this point, there are seven virtual voltage vectors within each sector. Based on the principle of planar vectors, the synthesis of these virtual voltage vectors is as follows:

[0141]

[0142] The coefficients of V1, V2, and V0 represent the duty cycles of the corresponding virtual voltage vectors.

[0143] Therefore, the action times of each basic vector used to synthesize the virtual voltage vector can be easily calculated as follows:

[0144]

[0145] Where d i,j Let represent the duty cycle of the three basic voltage vectors that form the composite virtual voltage vector, where i represents the index of the three basic voltage vectors V1, V2, and V0, and i∈{1,2,3}, and j represents the virtual voltage vector. The lower left index of the index, and j∈{8,9,10,11,12,13,14}.

[0146] when When located in the region between R1 and R6, either J2>J1>J3 or J1>J2>J3, and the optimal virtual voltage vector is: when When located in the region between R2 and R3, either J1>J3>J2 or J3>J1>J2, the optimal virtual voltage vector is... when When located in the region between R4 and R5, either J3>J2>J1 or J2>J3>J1, and the optimal virtual voltage vector is...

[0147] The pre-selection rules for its virtual three-vector voltages are consistent with the pre-selection rules in the second step of step 3.

[0148]

[0149] If users can design a three-vector mode optimization based on discrete space vector modulation when N=4 and N=5, the control accuracy of the control method in this invention can be effectively improved.

[0150] Step S3. Based on the voltage error objective function, switch between the dual-vector mode and the three-vector mode, finally find the optimal output voltage vector, and convert the duty cycle corresponding to the optimal voltage vector into a switching pulse signal to be applied to the two-level inverter.

[0151] Taking the vector line segment as an example where it is evenly divided into two intervals, i.e., N=2, step 3 uses the hybrid mode optimization algorithm to obtain the dual-vector synthesized voltage. and virtual vector voltage As a candidate output voltage V ou The voltage vector V is obtained by passing t to a voltage error-based objective function and then solving for the minimum value of the objective function. opt This allows for switching between the two operating modes. The solution process is as follows:

[0152]

[0153] The global optimal voltage vector V opt The corresponding duty cycle is converted into a switching pulse signal that acts on the two-level inverter.

[0154] Compared with existing dual-vector model predictive torque control technology, which only uses two adjacent basic vectors to synthesize a voltage vector, there are at most three combinations of dual-vector voltage vectors within a sector, i.e., three duty cycles. Furthermore, the outermost voltage vector combination and its duty cycle are generally only used when the rotor is running at high speed or under sudden changes in state. In addition, this technology restricts the position of the synthesized voltage vector to the boundary of the triangle, which leads to a relatively large deviation between the synthesized vector and the reference vector. Therefore, it lacks the degree of freedom of the synthesized voltage vector and reduces the control accuracy.

[0155] The above description, based on the embodiments shown in the figures, details the structure, features, and effects of the present invention. The above description is only a preferred embodiment of the present invention, but the present invention is not limited to the scope of implementation shown in the figures. Any changes made in accordance with the concept of the present invention, or equivalent embodiments modified to have equivalent changes, that do not exceed the spirit covered by the specification and figures, should be within the protection scope of the present invention.

Claims

1. A model predictive torque control method for permanent magnet synchronous motors based on hybrid mode optimization, characterized in that: The method includes the following steps: Step S1. Real-time acquisition of motor stator current, rotor speed, and motor electrical angle, and construction of a reference voltage vector based on the objective function of torque error and the discretized prediction model of the permanent magnet synchronous motor, specifically: The formula for calculating the reference voltage vector is constructed based on the objective function of torque error and the discretized prediction model of surface-mounted PMSM. In the two-phase dq rotating coordinate system, the specific structure is as follows: in, In the formula, For stator resistance, For armature winding inductance, For permanent magnet motors with constant flux linkage, This refers to the rotor speed, expressed in rad / s. For reference electromagnetic torque, For extreme logarithms, The sampling period of the discrete control system. and Let be the direct-axis and quadrature-axis components of the stator current at time k. and The direct-axis and quadrature-axis components of the stator current are predicted one step ahead at time k+1. Build based on Reference voltage vector in coordinate system It can be expressed by the following formula: In the formula, The electrical angle of the motor; Step S2. Generate a dual-vector synthesized voltage based on a dual-vector mode and a virtual vector voltage based on a three-vector mode according to the reference voltage vector and a preset hybrid mode optimization algorithm. This includes generating a dual-vector synthesized voltage based on a dual-vector mode and calculating the corresponding duty cycle according to the hybrid mode optimization algorithm, and establishing a pre-selection rule for the dual-vector synthesized voltage. Specifically: When the reference voltage vector Located within the first sector, the first sector is defined to include four voltage vectors. ,in, and The basic voltage vector provided for the inverter, and All are zero vectors. The first sector is divided into three regions Z1, Z2, and Z3. When the reference voltage vector... When located in different regions, dual-vector voltages are selected according to preset pre-selection rules; The calculation process for the dual-vector duty cycle in the three regions Z1, Z2, and Z3 is as follows: When the reference voltage vector is located in region Z1, voltage vectors V1 and V0 are selected for synthesizing a dual-vector voltage. The optimal duty cycle is calculated using the following formula: in Voltage vector duty cycle, Voltage vector The duty cycle, and has , The duty cycle of the voltage vector V1 The calculation process is as follows: Select the voltage vector when the reference voltage vector is located in the Z2 region. and voltage vector To synthesize a dual-vector voltage, the optimal duty cycle is calculated using the following formula: in Voltage vector duty cycle, Voltage vector The duty cycle, and has ; The voltage vector duty cycle The calculation process is as follows: When the reference voltage vector is located in the Z3 region, select the voltage vector. and voltage vector To synthesize a dual-vector voltage, the optimal duty cycle is calculated using the following formula: in Voltage vector duty cycle, Voltage vector The duty cycle, and has ; The voltage vector duty cycle The calculation process is as follows: The step of establishing the pre-selection rules for the dual-vector synthesized voltage and selecting the dual-vector voltage according to the preset pre-selection rules specifically involves: setting... , , The pre-selection rules are as follows: in , They represent respectively duty cycle, Indicates a two-vector voltage; The steps for generating a virtual vector voltage based on a three-vector mode include synthesizing a virtual vector based on discrete space vector modulation, and the synthesized virtual vector voltage. The calculation formula is as follows: In the formula, This represents the duty cycle of the true vector, and N represents the interval divided into each sampling period. Represents the actual voltage vector; When the reference voltage vector When located in the first sector, the sector is divided into six sub-regions, among which, the virtual voltage vector Located at the midpoint inside an equilateral triangle, three virtual voltage vectors , , The midpoint of the line segment between each vertex of the sector and the midpoint of the triangle is expressed by the following formula: in , , The coefficient is the duty cycle of the corresponding virtual voltage vector; The duty cycle of the virtual voltage vector at the corresponding position is shown below: in This represents the duty cycle of the three basic voltage vectors that form the composite virtual voltage vector, where i represents the duty cycle of each of the three basic voltage vectors. The serial number, and j represents the virtual voltage vector , , , The bottom left number, and ; The virtual voltage vector is selected according to the aforementioned pre-selection rules, specifically as follows: Step S3. Based on the voltage error objective function, switch between the dual-vector mode and the three-vector mode to output the optimal voltage vector, and convert the duty cycle corresponding to the optimal voltage vector into a switching pulse signal to be applied to the two-level inverter.

2. The method according to claim 1, characterized in that: In step S3, the step of switching between the dual-vector mode and the three-vector mode based on the voltage error objective function to output the optimal voltage vector specifically involves generating the dual-vector composite voltage. and the virtual vector voltage as a candidate output voltage The optimal voltage vector is obtained by solving the objective function that minimizes the voltage error. This allows for switching between the two operating modes. The solution process is as follows: The global optimal voltage vector The corresponding duty cycle is converted into a switching pulse signal that acts on the two-level inverter.

Images