A motor field-weakening control method and system based on a limited set model predictive control

Abstract

The application provides a motor field weakening control method and system based on limited set model predictive control, can solve the problem that the output voltage of limited control set model predictive control cannot be directly used for closed-loop field weakening control, and comprises the following steps: obtaining operation parameters of a current sampling period of a motor, and predicting dq axis currents of a next sampling period; constructing a reference voltage of the next sampling period, the reference voltage is divided into a sampling voltage component and a transient voltage component, performing boundary judgment on a reference current vector corresponding to the transient voltage component, and determining whether to perform dynamic amplitude limiting processing on the transient voltage component according to a boundary judgment result; adding the sampling voltage component and the processed transient voltage component to obtain a corrected reference voltage, then obtaining a field weakening current component through comparison and closed-loop feedback, superimposing the field weakening current component to obtain a corrected d-axis reference current, and updating the reference current; and obtaining an optimal voltage vector through an evaluation function of limited set model predictive control to perform control.

Description

Technical Field

[0001] This invention relates to the field of motor control technology, specifically to a motor field weakening control method and system based on finite set model predictive control. Background Technology

[0002] With the development of advanced microprocessor technologies such as programmable gate arrays (FPGAs), Model Predictive Control (MPC) has been widely used in motor drive systems due to its excellent support for dynamic performance. Based on their working principles, MPC algorithms are mainly divided into two categories: Continuous Control Set Model Predictive Control (CCS-MPC) and Finite Control Set Model Predictive Control (FCS-MPC). CCS-MPC optimizes the desired output voltage into a continuous control signal, requiring the synthesis of a voltage vector through a modulator. While this is beneficial for fixing the switching frequency and reducing steady-state torque ripple, its implementation is relatively complex and typically requires offline optimization, making it difficult to implement online in real-time motor control systems. FCS-MPC, on the other hand, optimizes the desired output voltage into a discrete control signal, directly selecting the inverter's switching state. It has the advantages of simple structure and online optimization, and is widely used in motor control systems.

[0003] Besides the need for high dynamic response, industrial applications such as compressors, new energy vehicles, CNC machine tools, and drones also require motors to operate over a wide speed range. Field weakening control is a crucial technique for extending the high-speed operating range of motors. Its basic idea is to reduce the motor's rotational back electromotive force (EMF) by weakening the motor's magnetic field under conditions where the inverter output voltage is limited, thereby increasing the proportion of the back EMF in the output voltage and achieving speed expansion. Simultaneously, field weakening control must also adhere to the output current limitations of both the inverter and the motor. Therefore, field weakening control is essentially a motor operating path optimization problem under multiple constraints.

[0004] To simultaneously meet the performance requirements of high dynamic range and wide speed range, it is necessary to combine finite set model predictive control and field weakening control. However, due to the discreteness of the output voltage of finite set model predictive control, field weakening control based on closed-loop feedback is difficult to implement directly on finite set model predictive control.

[0005] Existing field weakening control methods are mainly divided into two categories: open-loop flux linkage control and closed-loop voltage control. Open-loop flux linkage control utilizes the relationship between motor flux linkage parameters and speed to calculate the given field weakening current at different speeds offline. This method does not involve controller parameters and can therefore be directly applied to finite set model predictive control. However, due to its open-loop design, this method is prone to over-weakening under light load conditions and under-weakening under heavy load conditions: over-weakening increases motor operating losses, while under-weakening reduces the motor's load-carrying capacity. Closed-loop voltage control, on the other hand, uses the controller's output reference voltage amplitude as the feedback quantity and the inverter's maximum output voltage amplitude as the reference quantity, generating the required field weakening current through the difference between the two. This method plans the reference current path through real-time feedback, effectively avoiding the load-field weakening mismatch problem encountered in open-loop flux linkage control.

[0006] According to the voltage closed-loop control method, the controller output reference voltage is the key variable in the closed-loop control. However, in finite set model predictive control, the output reference voltage is directly selected from the discrete voltage set consisting of 6 effective voltage vectors and 2 zero vectors through the evaluation function, and its amplitude is only 2 / 3u. dc Or 0, u dc If the discrete reference voltage, representing the DC bus voltage, is directly used for closed-loop field weakening control, it will cause significant fluctuations in the generated field weakening current trajectory, severely affecting the motor's high-speed control performance and even leading to motor runaway. Therefore, how to construct a low-fluctuation reference voltage suitable for closed-loop field weakening control within a finite control set model predictive control framework remains an unsolved technical problem in the current technology. Summary of the Invention

[0007] To address the aforementioned problems, this invention provides a motor field weakening control method and system based on finite set model predictive control, which can solve the problem that the output voltage of finite set model predictive control in the prior art cannot be directly used for closed-loop field weakening control.

[0008] The technical solution is as follows: a field weakening control method for motors based on finite set model predictive control, comprising the following steps: Step S1: Obtain the operating parameters of the motor in the current sampling period, including the dq axis current sampling value, motor speed, rotor position and bus voltage; Step S2: Based on the obtained operating parameters, enumerate each voltage vector in the inverter voltage vector set, and predict the dq axis current of the next sampling period based on the motor discretization model; Step S3: Based on the operating parameters and reference current collected in step S1, construct the reference voltage for the next sampling period. The reference voltage includes a sampled voltage component and a transient voltage component. The sampled voltage component is obtained based on the dq-axis sampled current and motor speed in the current sampling period, and the transient voltage component is obtained based on the difference between the reference current and the dq-axis sampled current. Step S4: Based on the predicted current obtained in step S2, perform boundary judgment on the reference current vector corresponding to the transient voltage component, and decide whether to perform dynamic limiting processing on the transient voltage component based on the boundary judgment result. Step S5: Add the sampled voltage component to the transient voltage component processed in step S4 to obtain the corrected reference voltage. Compare the amplitude of the corrected reference voltage with the amplitude of the inverter's maximum output voltage. The difference is used by the closed-loop feedback controller to obtain the field weakening current component. The field weakening current component is used as a correction amount and superimposed on the d-axis reference current to obtain the corrected d-axis reference current. Update the reference current. Step S6: Based on the predicted current and the updated reference current, the optimal voltage vector that minimizes the evaluation function value is selected from the voltage vector set using the evaluation function of the finite set model prediction control, and the inverter drives the motor to run through the optimal voltage vector.

[0009] Furthermore, in step S2, based on the finite control set model predictive control, the voltage vectors corresponding to all switching states of the inverter are enumerated, and the predicted current vector for the next sampling period is calculated, expressed as: , Among them, i, u, ω e R, L, ψ f These are represented by current, voltage, rotational speed, resistance, inductance, and permanent magnet flux linkage, respectively, T. s Where n is the sampling time, k is the sampling period, the superscript n represents the voltage index within the voltage set, and the subscripts d and q represent the d and q axes of the dq-axis coordinate system. For the dq-axis voltage vector values ​​enumerated in the current sampling period, This is the predicted current vector for the next sampling period.

[0010] Furthermore, in step S3, the required initial reference voltage along the dq axis is calculated using the reference current in the dq-axis coordinate system, and the calculation formula is as follows: ; The superscript ref represents a reference value. R is the initial reference voltage for the dq axis. s L is the stator resistance of the motor. d L q For stator dq axis inductance, Let k be the motor speed during the sampling period. Let ω be the dq-axis current sample value with sampling period k. e T is the motor speed. s Sampling time, For permanent magnet flux linkage; and These are the reference current vectors in the dq-axis coordinate system with a sampling period of k+1; The initial reference voltage of the dq axis is decomposed into a sampled voltage component and a transient voltage component, as shown in the following formula: ; In the formula, These represent the sampled voltage components along the d and q axes, respectively, with the superscript 's' indicating that it represents the sampled voltage. These represent the transient voltage components along the d and q axes, respectively.

[0011] Furthermore, the d-axis reference current vector The output torque command of the k-sampling-period speed loop is obtained using the maximum torque-to-current ratio method; the q-axis reference current vector is... The output torque command of the speed loop is obtained by proportional-integral regulation and current amplitude limitation based on the k sampling period.

[0012] Furthermore, step S4 is executed as follows: Determine whether the reference current vector is within the predicted current boundary range of the finite control set model predictive control; If the reference current vector exceeds the predicted current boundary range, the transient voltage component is dynamically limited to obtain the limited transient voltage component; if the reference current vector is within the predicted current boundary range, the original value of the transient voltage component is retained.

[0013] Furthermore, determining whether the reference current vector is within the predicted current boundary range of the finite control set model predictive control in step S4 specifically includes the following steps: In the predicted current vector of the next sampling period, select the two predicted current points that are closest to the reference current vector to determine the sector where the reference current vector is located; A linear equation is constructed based on the two selected predicted current points. The predicted current vector and the reference current vector corresponding to the zero vector are substituted into the linear equation to obtain two sign judgment calculation values. The reference current vector is then judged to be within the predicted current boundary range based on the obtained sign judgment calculation values. If the signs of the two sign judgment calculation values ​​are consistent, the reference current vector is judged to be within the predicted current boundary range. If the signs are inconsistent, the reference current vector is judged to be outside the predicted current boundary range.

[0014] Furthermore, in step S4, the two predicted current points closest to the reference current vector are selected as follows: and ; A straight line is constructed based on the two selected predicted current points. The equation of the constructed straight line is: ; Substituting the predicted current vector and the reference current vector into the linear equation, we obtain the sign determination values ​​d1 and d2, which are expressed as: ; Where d1 and d2 are the calculated values ​​for sign determination. and It refers to the d-axis and q-axis components of the predicted current vector at one of the selected predicted current points n1 during the sampling period k+1. and These are the d-axis and q-axis components of the predicted current vector at another selected predicted current point n2 during sampling period k+1. and It is the predicted current vector at sampling period k+1, which is the zero vector. and It is the reference current vector in the sampling period k+1.

[0015] Furthermore, in step S4, the transient voltage component is dynamically limited, specifically as follows: The reference current vector is pulled back within the limiting value r by proportionally shortening it. The limiting value r is calculated based on the predicted current point corresponding to the two predicted current points and the zero vector. The value of the limiting value r is determined by the following formula: ; in, and It is the predicted current vector of one of the selected predicted current points n1 during the sampling period k+1. and It is the predicted current vector at another selected predicted current point n2 during sampling period k+1. and It is the predicted current vector at sampling period k+1, which is a zero vector. For reference current points that require dynamic limiting processing Reference current vector , The formula for proportional reduction is as follows: ; Where r is the amplitude limit value, and It is the reference current point after limiting. The reference current vector is used to adjust the transient voltage component based on the reference current vector after the amplitude is limited.

[0016] Furthermore, in step S5, the closed-loop feedback controller is a proportional-integral controller, and the maximum output voltage amplitude of the inverter is based on the voltage modulation coefficient m and the DC bus voltage. Confirmed, the calculation formula is: U max =m·(2u dc / π), The value of the voltage modulation coefficient m corresponds to the ratio of the inscribed radius to the circumscribed radius of the voltage vector hexagon.

[0017] Furthermore, in step S6, the formula for the evaluation function of the finite set model predictive control is as follows: ; In the formula, g is the evaluation function value. Based on the predicted current vector obtained in step S2 and the updated reference current in step S5, the evaluation function value g is calculated according to the formula of the evaluation function. The voltage vector that minimizes the value of g is selected from the voltage vector set as the optimal voltage vector. The inverter drives the motor to operate through the optimal voltage vector. , The predicted current vector for the next sampling period. and These are the reference current vectors in the dq-axis coordinate system with a sampling period of k+1.

[0018] A motor drive system includes an inverter, a motor, and a controller, wherein the controller is configured to execute the above-described motor field weakening control method based on finite set model predictive control.

[0019] A computer-readable storage medium having a program stored thereon, which, when executed by a processor, implements the field weakening control method for a motor based on finite set model predictive control as described above.

[0020] A computer program product includes a computer program / instructions that, when executed by a processor, implement the steps of the above-described method.

[0021] The present invention has the following advantages: This invention, within the framework of finite control set model predictive control, draws upon the design concept of deadbeat control. It constructs a continuous reference voltage suitable for closed-loop field weakening control through a discretized mathematical model of the motor. This overcomes the problem in existing technologies where the output voltage of finite control set model predictive control is discrete and cannot be directly used for voltage closed-loop field weakening control. This invention splits the reference voltage into two parts: a sampled voltage component and a transient voltage component. Only the transient voltage component is amplitude-limited, effectively suppressing its instantaneous fluctuations without affecting the dynamic response of the reference voltage.

[0022] The method of the present invention is based on the prediction of the current vector to determine the reference current boundary and dynamically limit the amplitude. It can adaptively adjust the amplitude limit value according to the motor operating state, avoiding the problem of reduced response performance or amplitude limit failure caused by fixed amplitude limit value under different operating conditions.

[0023] This invention applies the modified low-fluctuation reference voltage to voltage closed-loop field weakening control, enabling the field weakening current command to adaptively adjust according to load changes. This effectively avoids the problems of excessive field weakening under light load and insufficient field weakening under heavy load in the open-loop flux field weakening control method. Using the method of this invention, after the motor enters the field weakening zone, the current path can be stabilized at the voltage limit boundary, achieving an ideal field weakening state. Attached Figure Description

[0024] Figure 1 This is a schematic diagram illustrating the steps of a motor field weakening control method based on finite set model predictive control in an embodiment of the present invention. Figure 2 A predicted current vector diagram provided for an embodiment of the present invention; Figure 3 This is a schematic diagram of the dynamic limiting value provided in an embodiment of the present invention; Figure 4 A control block diagram of a motor field weakening control method based on finite set model predictive control provided in an embodiment of the present invention; Figure 5 The rotational speed and dq-axis current waveforms are obtained using the existing open-loop magnetic flux weakening method. Figure 6 The rotational speed and dq-axis current waveforms using the method of the present invention; Figure 7 This is a comparison diagram of the dq-axis current paths of the method of the present invention and the magnetic flux open-loop weakening method. Detailed Implementation

[0025] See Figure 1 The present invention provides a motor field weakening control method based on finite set model predictive control, comprising at least the following steps: Step S1: Obtain the operating parameters of the motor in the current sampling period. The operating parameters include the dq axis current sampling value, motor speed, rotor position, and bus voltage. Step S2: Based on the obtained operating parameters, enumerate each voltage vector in the inverter voltage vector set, and predict the dq axis current of the next sampling period based on the motor discretization model; Step S3: Based on the operating parameters and reference current collected in step S1, construct the reference voltage for the next sampling period. The reference voltage includes a sampled voltage component and a transient voltage component. The sampled voltage component is obtained based on the dq-axis sampled current and motor speed in the current sampling period, and the transient voltage component is obtained based on the difference between the reference current and the dq-axis sampled current. Step S4: Based on the predicted current obtained in step S2, perform boundary judgment on the reference current vector corresponding to the transient voltage component, and decide whether to perform dynamic limiting processing on the transient voltage component based on the boundary judgment result. Step S5: Add the sampled voltage component to the transient voltage component processed in step S4 to obtain the corrected reference voltage. Compare the amplitude of the corrected reference voltage with the amplitude of the inverter's maximum output voltage. The difference is used by the closed-loop feedback controller to obtain the field weakening current component. The field weakening current component is used as a correction and superimposed on the d-axis reference current to obtain the corrected d-axis reference current. Update the reference current. Step S6: Based on the predicted current and the updated reference current, the evaluation function of the predictive control is used with a finite set model. The optimal voltage vector that minimizes the evaluation function value is selected from the voltage vector set. The inverter drives the motor to run by controlling the optimal voltage vector.

[0026] In one specific embodiment of the present invention, in step 1, the three-phase current of the motor with a sampling period k is collected and the dq-axis sampling current i is obtained through coordinate transformation. d (k) and i q (k); Acquire DC bus voltage u dc Collect or observe the motor speed and rotor position .

[0027] Based on the state variables acquired in step S1, predictive control is performed using a finite control set model. The voltage vectors corresponding to all switching states of the inverter are enumerated. For each voltage vector in the inverter voltage vector set U={u0,u1,u2,u3,u4,u5,u6}, in this embodiment, two zero vectors are merged into u0. Based on the discretized mathematical model of the motor, the dq-axis current for the k+1 sampling period is predicted. The formula for predicting the current is shown in the following equation: (1), Among them, i, u, ω e R, L, ψ fThese are represented by current, voltage, rotational speed, resistance, inductance, and permanent magnet flux linkage, respectively, T. s Where n is the sampling time, k is the sampling period, the superscript n represents the voltage index within the voltage set, and the subscripts d and q represent the d and q axes of the dq-axis coordinate system. This refers to the d-q axis voltage vector values ​​enumerated for the current sampling period, including d- and q-axis components. The predicted current vector for the next sampling period includes d-axis and q-axis components.

[0028] The finite control set model predictive control uses an evaluation function to select the optimal voltage vector that minimizes the evaluation function value among all predicted current vectors. The evaluation function is shown in equation (2): (2); In the formula, g is the evaluation function value. The optimal voltage vector is selected, and the inverter drives the motor to operate by controlling the optimal voltage vector. and These are the reference current vectors in the dq-axis coordinate system with a sampling period of k+1.

[0029] In step S3, due to the strong fluctuations in the output voltage of the finite control set model predictive control, its output voltage cannot be directly used as a reference value for closed-loop field weakening control. Therefore, this embodiment of the invention draws on the concept of deadbeat control: finite control set model predictive control predicts the current at time k+1 using an existing voltage vector, while deadbeat control infers the required reference voltage from the reference current. The two are mathematically dual. The required initial reference voltage on the dq axis is calculated using the reference current in the dq axis coordinate system, and the calculation formula is as follows: (3); The superscript ref represents a reference value. R is the initial reference voltage for the dq axis. s L is the stator resistance of the motor. d L q For stator dq axis inductance, Let k be the motor speed during the sampling period. Let ω be the dq-axis current sample value with sampling period k. e T is the motor speed. s Sampling time, For permanent magnet flux linkage; and These are the reference current vectors in the dq-axis coordinate system with a sampling period of k+1; in this embodiment, the d-axis reference current vector... The q-axis reference current vector is obtained from the output torque command of the speed loop during the k-sampling period using the maximum torque-to-current ratio method. The output torque command of the speed loop is obtained by proportional-integral regulation and current amplitude limitation based on the k sampling period.

[0030] Observing equation (3), it can be seen that the reference voltage at time k+1 without beat prediction consists of two parts: one part is directly related to the sampling current and sampling motor speed at time k, representing the steady-state voltage value of the motor at time k; the other part is related to the difference between the reference current and the sampling current, representing the transient voltage required for the motor to transition from time k to time k+1. Accordingly, in this embodiment, based on equation (3), the initial reference voltage of the dq axis is decomposed into a sampling voltage component and a transient voltage component, as shown in the following formula: (4); In the formula, These represent the sampled voltage components along the d and q axes, respectively, with the superscript 's' indicating that it represents the sampled voltage. These represent the transient voltage components along the d and q axes, respectively.

[0031] From the decomposition form of equation (4), it can be seen that: rotational speed and dq axis current The value does not change drastically within adjacent sampling periods, and the corresponding sampling voltage... The values ​​do not change drastically within adjacent sampling periods; for transient voltage, its amplitude variation mainly depends on the difference between the dq-axis reference current and the sampled current. When the motor speed enters the field weakening region from the base speed region, the inverter output voltage is limited. Without field weakening control, the dq-axis current will not be able to track the reference current, resulting in a difference. At the same time, due to the speed limitation, the current reference value output by the speed loop will quickly saturate, further widening the difference between the reference current and the sampled current. Similarly, under conditions of sudden load increases or decreases within the motor's base speed range, the current reference value will undergo a step change, which will also cause significant fluctuations in the transient voltage. Therefore, by appropriately limiting the transient voltage, a low-fluctuation reference voltage can be obtained for closed-loop field weakening control.

[0032] In this embodiment, step S4 is specifically performed as follows: Determine whether the reference current vector is within the predicted current boundary range of the finite control set model predictive control; If the reference current vector exceeds the predicted current boundary range, the transient voltage component is dynamically limited to obtain the limited transient voltage component; if the reference current vector is within the predicted current boundary range, the original value of the transient voltage component is retained.

[0033] In this embodiment, determining whether the reference current vector is within the predicted current boundary range of the finite control set model predictive control specifically includes the following steps: In the predicted current vector of the next sampling period, select the two predicted current points that are closest to the reference current vector to determine the sector where the reference current vector is located; A linear equation is constructed based on the two selected predicted current points. The predicted current vector and the reference current vector corresponding to the zero vector are substituted into the linear equation to obtain two sign judgment calculation values. The reference current vector is then judged to be within the predicted current boundary range based on the obtained sign judgment calculation values. If the signs of the two sign judgment calculation values ​​are consistent, the reference current vector is judged to be within the predicted current boundary range. If the signs are inconsistent, the reference current vector is judged to be outside the predicted current boundary range.

[0034] See Figure 2 The predicted current vector diagram shown shows i dq (k) is the sampled current vector at time k; (n=0,1,2,...,6) represents the predicted current vectors corresponding to different voltage vectors u0~u6 in the voltage vector set under sampling period k+1. They form a hexagonal distribution in the dq coordinate system, with the center of the hexagon being the predicted current vector corresponding to the zero vector. , and Examples of reference current vectors at two different locations, where Located within the hexagonal boundary, Located outside the hexagonal boundary, the subscript dq here represents the dq axis, which also represents the components of the d and q axes.

[0035] If the reference current vector is within the boundary, it means that the inverter's maximum output voltage in the current state can support the current transition from the sampled value to the reference value; if the reference current vector is outside the boundary, it means that the inverter's maximum output voltage cannot meet the current variation requirements. Therefore, when the reference current is within the boundary, the transient voltage does not need to be limited; however, when the reference current is outside the boundary, the transient voltage needs to be limited to prevent large fluctuations under specific operating conditions. For example, when the motor experiences sudden load increases or decreases within its base speed range, the current reference value will undergo a step change, thus affecting the transient voltage.

[0036] In this embodiment, the determination of whether the reference current is inside or outside the boundary includes the following steps: Step S401: Predict the six non-zero current vectors in a hexagonal distribution. ~ In the middle, the two predicted current points closest to the reference current vector are selected as follows: and ; and The expression separates the d-axis and q-axis of the predicted current point.

[0037] A straight line is constructed by selecting the two predicted current points. The equation of the constructed straight line is: (5); Substituting the predicted current vector and the reference current vector into the linear equation, we obtain the sign determination values ​​d1 and d2, which are expressed as: (6); Where d1 and d2 are the calculated values ​​for sign determination. and It refers to the d-axis and q-axis components of the predicted current vector at one of the selected predicted current points n1 during the sampling period k+1. and These are the d-axis and q-axis components of the predicted current vector at another selected predicted current point n2 during sampling period k+1. and It is the predicted current vector at sampling period k+1, which is the zero vector. and The reference current vector is at sampling period k+1. If the signs of d1 and d2 are the same, i.e. d1·d2>0, then the reference current is determined to be inside the boundary; if the signs of d1 and d2 are different, i.e. d1·d2<0, then the reference current is determined to be outside the boundary.

[0038] Step S402: The voltage margin reserved for current adjustment is inconsistent under different operating conditions of the motor. Therefore, when the reference current exceeds the boundary range, the limit value needs to be dynamically adjusted according to the different voltage margins. Since the voltage margin cannot be directly measured, this embodiment indirectly characterizes the size of the voltage margin by predicting the current.

[0039] See Figure 3 The diagram showing the dynamic limiting value illustrates what happens when the reference current exceeds the limiting value. Figure 3 Reference current point The reference current point shown needs to be pulled back to within the limiting boundary by proportionally shortening it, such as... Figure 3 middle As shown; the limiting value r is calculated based on the predicted current point corresponding to the two predicted current points and the zero vector. The value of the limiting value r is determined by the following formula: (7); in, and It is the predicted current vector of one of the selected predicted current points n1 during the sampling period k+1. and It is the predicted current vector at another selected predicted current point n2 during sampling period k+1. and It is the predicted current vector at sampling period k+1, which is a zero vector. For reference current points that require dynamic limiting processing Reference current vector , The formula for proportional reduction is as follows: (8); in, and It is the reference current point after limiting. The d-axis and q-axis reference current vectors are used to substitute the reduced reference current values ​​into the transient voltage component formula in equation (4) to obtain the limited transient voltage component. The transient voltage component is then adjusted according to the limited reference current vector. In step S5, the sampled voltage component obtained in step S3 is added to the limited transient voltage component obtained in step S4 to obtain the corrected reference voltage. The amplitude of the corrected reference voltage is compared with the amplitude of the inverter's maximum output voltage, which is based on the voltage modulation coefficient m and the DC bus voltage. Confirmed, the calculation formula is: U max =m·(2u dc / π)(9) The voltage modulation coefficient m corresponds to the ratio of the inscribed radius to the circumscribed radius of the voltage vector hexagon, typically set to 0.907. The difference is compared and processed by a PI controller to obtain the d-axis field-weakening current component. This component is then superimposed as a correction factor onto the d-axis reference current generated by the MTPA method, resulting in an updated d-axis reference current. .

[0040] In step S6 of the embodiment, based on the predicted current vector obtained in step S2 and the reference current updated in step S5, the evaluation function value g is calculated according to the formula of the evaluation function. The voltage vector that minimizes the value of g is selected from the voltage vector set U as the optimal voltage vector, and the switching state corresponding to the optimal voltage vector is directly applied to the inverter to drive the motor to run according to the expected field weakening path.

[0041] This embodiment uses a two-level voltage source inverter driving a permanent magnet synchronous motor (PMSM) as a specific implementation object for illustration, but the method of the present invention is also applicable to other AC motor drive systems with dq axis mathematical models.

[0042] See Figure 4The overall control architecture of the motor field weakening control method based on finite set model predictive control in this embodiment includes: an outer speed loop, a middle current reference generation loop, an inner closed-loop field weakening control loop based on model prediction, and an innermost evaluation function selection loop.

[0043] The outer speed ring references the motor speed. With actual sampled motor speed The difference is taken as input, and the torque command is output through the proportional-integral regulator; In the middle layer current reference generation loop, the d-axis reference current is generated by torque command through the MTPA method, and the q-axis reference current is generated by torque command after PI regulation and current amplitude limitation. The inner closed-loop field weakening control loop constructs a corrected reference voltage based on the sampled current, sampled motor speed, and reference current, combined with the sampled voltage calculation unit and transient voltage calculation unit. The amplitude of this reference voltage is then compared with the inverter's maximum output voltage amplitude m·(2u). dc The difference is compared with the π / 2, and the magnetic field weakening current component is output by the PI regulator. The magnetic field weakening current component is superimposed on the d-axis reference current. Here, m is the voltage modulation coefficient, which is usually taken as 0.907, that is, the ratio of the inscribed circle radius to the circumscribed circle radius of the voltage vector hexagon. The innermost evaluation function selection loop uses an evaluation function to select the optimal voltage vector that minimizes the evaluation function value from a set of voltage vectors consisting of 6 effective voltage vectors u1~u6 and 2 zero vectors u0, and sets the corresponding switching state [S]. a ,S b ,S c The three-phase motor current is directly applied to the two-level inverter. The Clarke transform yields the two-phase stationary α-β axis currents. Then, through the Park transformation involving the rotor position θ(k), the dq-axis sampled current is obtained, which serves as the feedback quantity for the control system.

[0044] There is no data dependency between the sampling voltage calculation and the transient voltage calculation in step S3. Therefore, in a specific hardware implementation of one embodiment, the sampling voltage calculation, boundary judgment, dynamic limiting and transient voltage calculation and other links can be executed in parallel using a programmable gate array parallel processing architecture to shorten the calculation time in a single control cycle and meet the real-time requirements of the motor when running at high speed.

[0045] Specifically, the sampling voltage calculation unit, boundary judgment unit, dynamic limiting unit, and transient voltage calculation unit can be mapped to independent hardware logic units. After completing the signal acquisition and prediction current enumeration at time k, the sampling voltage calculation and transient voltage calculation start simultaneously and run in parallel. The final output results are synthesized into a reference voltage within the same clock cycle for use by the closed-loop field weakening control loop.

[0046] See Figure 5 , Figure 6 , Figure 7 To verify the effectiveness of the method of the present invention, this embodiment conducted a comparative experiment on the method of the present invention and the magnetic flux open-loop field weakening method on a permanent magnet synchronous motor experimental platform. The experimental setting was the dividing speed between the base speed region and the field weakening region as approximately 1500 rpm, and the target speed as 2500 rpm.

[0047] Figure 5 The rotational speed and dq-axis current waveforms of the open-loop magnetic flux weakening method in the prior art are shown. After the rotational speed reaches the target value, the d-axis current continues to decrease, and a stable magnetic weakening operating point is not formed. Figure 6 The rotational speed and dq-axis current waveforms of the method of the present invention are shown. After the rotational speed reaches the target value, the d-axis current exhibits obvious foldback characteristics and stabilizes at a certain value, indicating that the current path is in a steady state of closed-loop field weakening control. Figure 7 The paper further compares the current paths of the two methods in the dq plane. The current path of the method of the present invention is stably attached to the voltage limit curve and is in an ideal magnetic weakening state, while the current path of the magnetic flux open-loop magnetic weakening method deviates significantly from the voltage limit curve and exhibits excessive magnetic weakening.

[0048] The above experimental results demonstrate that, compared with the open-loop magnetic flux weakening method, the method of the present invention can maintain close tracking of the current path to the voltage limit boundary in the magnetic weakening region, avoid the problems of increased losses and reduced load capacity caused by excessive magnetic weakening, and improve the control performance of the motor in a wide speed range.

[0049] In an embodiment of the present invention, a motor drive system is also provided, including an inverter, a motor, and a controller, wherein the controller is configured to execute the above-described motor field weakening control method based on finite set model predictive control.

[0050] In an embodiment of the present invention, a computer-readable storage medium is also provided, on which a program is stored, which, when executed by a processor, implements the motor field weakening control method based on finite set model predictive control as described above.

[0051] The methods described in the above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any combination thereof. If implemented in software, the functionality can be stored as one or more instructions or code on or transmitted on a computer-readable storage medium. The computer-readable storage medium can include computer storage media and communication media, and can also include any medium capable of transferring a computer program from one place to another. The storage medium can be any target medium accessible by a computer.

[0052] In one possible implementation, a computer-readable storage medium may include RAM, ROM, compact disc read-only memory (CD-ROM) or other optical disc storage, magnetic disk storage or other magnetic storage devices, or any other medium targeted to carry or to store the required program code in the form of instructions or data structures, and accessible by a computer. Furthermore, any connection is appropriately referred to as a computer-readable storage medium. For example, if software is transmitted from a website, server, or other remote source using coaxial cable, fiber optic cable, twisted pair, Digital Subscriber Line (DSL), or wireless technologies such as infrared, radio, and microwave, then coaxial cable, fiber optic cable, twisted pair, DSL, or wireless technologies such as infrared, radio, and microwave are included in the definition of medium. As used herein, disks and optical discs include optical discs, laser discs, optical discs, Digital Versatile Discs (DVDs), floppy disks, and Blu-ray discs, where disks typically reproduce data magnetically, while optical discs optically reproduce data using lasers. Combinations of the above should also be included within the scope of computer-readable storage media.

[0053] In an embodiment of the present invention, a computer program product is also provided, including a computer program / instructions that, when executed by a processor, implement the steps of the above-described method.

[0054] This application describes embodiments of methods, apparatus (systems), and computer program products according to embodiments of this application with reference to flowchart illustrations and / or block diagrams. It should be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processing unit of a general-purpose computer, special-purpose computer, embedded processor, or other programmable device to produce a machine, such that the instructions, which execute via the processing unit of the computer or other programmable data processing device, generate instructions for implementing the flowchart illustrations. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0055] The above specific embodiments further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above are merely specific embodiments of the present invention and are not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made on the basis of the technical solution of the present invention should be included within the scope of protection of the present invention.

Claims

1. A field weakening control method for a motor based on finite set model predictive control, characterized in that, Includes the following steps: Step S1: Obtain the operating parameters of the motor in the current sampling period, including the dq axis current sampling value, motor speed, rotor position and bus voltage; Step S2: Based on the obtained operating parameters, enumerate each voltage vector in the inverter voltage vector set, and predict the dq axis current of the next sampling period based on the motor discretization model; Step S3: Based on the operating parameters and reference current collected in step S1, construct the reference voltage for the next sampling period. The reference voltage includes a sampled voltage component and a transient voltage component. The sampled voltage component is obtained based on the dq-axis sampled current and motor speed in the current sampling period, and the transient voltage component is obtained based on the difference between the reference current and the dq-axis sampled current. Step S4: Based on the predicted current obtained in step S2, perform boundary judgment on the reference current vector corresponding to the transient voltage component, and decide whether to perform dynamic limiting processing on the transient voltage component based on the boundary judgment result. Step S5: Add the sampled voltage component to the transient voltage component processed in step S4 to obtain the corrected reference voltage. Compare the amplitude of the corrected reference voltage with the amplitude of the inverter's maximum output voltage. The difference is used by the closed-loop feedback controller to obtain the field weakening current component. The field weakening current component is used as a correction amount and superimposed on the d-axis reference current to obtain the corrected d-axis reference current. Update the reference current. Step S6: Based on the predicted current and the updated reference current, the optimal voltage vector that minimizes the evaluation function value is selected from the voltage vector set using the evaluation function of the finite set model prediction control, and the inverter drives the motor to run through the optimal voltage vector.

2. The motor field weakening control method based on finite set model predictive control according to claim 1, characterized in that, In step S2, based on the finite control set model predictive control, the voltage vectors corresponding to all switching states of the inverter are enumerated, and the predicted current vector for the next sampling period is calculated, expressed as: ， Among them, i, u, ω e R, L, ψ f These are represented by current, voltage, rotational speed, resistance, inductance, and permanent magnet flux linkage, respectively, T. s Where n is the sampling time, k is the sampling period, the superscript n represents the voltage index within the voltage set, and the subscripts d and q represent the d and q axes of the dq-axis coordinate system. For the dq-axis voltage vector values ​​enumerated in the current sampling period, This is the predicted current vector for the next sampling period.

3. The motor field weakening control method based on finite set model predictive control according to claim 1, characterized in that: In step S3, the required initial reference voltage along the dq axis is calculated using the reference current in the dq-axis coordinate system. The calculation formula is as follows: ； The superscript ref represents a reference value. R is the initial reference voltage for the dq axis. s L is the stator resistance of the motor. d L q For stator dq axis inductance, Let k be the motor speed during the sampling period. Let ω be the dq-axis current sample value with sampling period k. e T is the motor speed. s Sampling time, For permanent magnet flux linkage; and These are the reference current vectors in the dq-axis coordinate system with a sampling period of k+1; The initial reference voltage of the dq axis is decomposed into a sampled voltage component and a transient voltage component, as shown in the following formula: ； In the formula, These represent the sampled voltage components along the d and q axes, respectively, with the superscript 's' indicating that it represents the sampled voltage. These represent the transient voltage components along the d and q axes, respectively.

4. The motor field weakening control method based on finite set model predictive control according to claim 3, characterized in that: d-axis reference current vector The q-axis reference current vector is obtained from the output torque command of the speed loop during the k-sampling period using the maximum torque-to-current ratio method. The output torque command of the speed loop is obtained by proportional-integral regulation and current amplitude limitation based on the k sampling period.

5. The motor field weakening control method based on finite set model predictive control according to claim 1, characterized in that: Step S4 is executed as follows: Determine whether the reference current vector is within the predicted current boundary range of the finite control set model predictive control; If the reference current vector exceeds the predicted current boundary range, the transient voltage component is dynamically limited to obtain the limited transient voltage component; if the reference current vector is within the predicted current boundary range, the original value of the transient voltage component is retained.

6. The motor field weakening control method based on finite set model predictive control according to claim 5, characterized in that: Step S4, determining whether the reference current vector is within the predicted current boundary range of the finite control set model predictive control, specifically includes the following steps: In the predicted current vector of the next sampling period, select the two predicted current points that are closest to the reference current vector to determine the sector where the reference current vector is located; A linear equation is constructed based on the two selected predicted current points. The predicted current vector and the reference current vector corresponding to the zero vector are substituted into the linear equation to obtain two sign judgment calculation values. The reference current vector is then judged to be within the predicted current boundary range based on the obtained sign judgment calculation values. If the signs of the two sign judgment calculation values ​​are consistent, the reference current vector is judged to be within the predicted current boundary range. If the signs are inconsistent, the reference current vector is judged to be outside the predicted current boundary range.

7. The motor field weakening control method based on finite set model predictive control according to claim 6, characterized in that: In step S4, the two predicted current points closest to the reference current vector are selected as follows: and ; A straight line is constructed based on the two selected predicted current points. The equation of the constructed straight line is: ； Substituting the predicted current vector and the reference current vector into the linear equation, we obtain the sign determination values ​​d1 and d2, which are expressed as: ； Where d1 and d2 are the calculated values ​​for sign determination. and It refers to the d-axis and q-axis components of the predicted current vector at one of the selected predicted current points n1 during the sampling period k+1. and These are the d-axis and q-axis components of the predicted current vector at another selected predicted current point n2 during sampling period k+1. and It is the predicted current vector at sampling period k+1, which is the zero vector. and It is the reference current vector in the sampling period k+1.

8. The motor field weakening control method based on finite set model predictive control according to claim 7, characterized in that: In step S4, the transient voltage component is dynamically limited, specifically as follows: The reference current vector is pulled back within the limiting value r by proportionally shortening it. The limiting value r is calculated based on the predicted current point corresponding to the two predicted current points and the zero vector. The value of the limiting value r is determined by the following formula: ； in, and It is the predicted current vector of one of the selected predicted current points n1 during the sampling period k+1. and It is the predicted current vector at another selected predicted current point n2 during sampling period k+1. and It is the predicted current vector at sampling period k+1, which is a zero vector. For reference current points that require dynamic limiting processing Reference current vector , The formula for proportional reduction is as follows: ； Where r is the amplitude limit value, and It is the reference current point after limiting. The reference current vector is used to adjust the transient voltage component based on the reference current vector after the amplitude is limited.

9. The motor field weakening control method based on finite set model predictive control according to claim 1, characterized in that: In step S5, the closed-loop feedback controller is a proportional-integral controller, and the maximum output voltage amplitude of the inverter is based on the voltage modulation coefficient m and the DC bus voltage. Confirmed, the calculation formula is: IN max =m·(2u dc / π), The value of the voltage modulation coefficient m corresponds to the ratio of the inscribed radius to the circumscribed radius of the voltage vector hexagon.

10. The motor field weakening control method based on finite set model predictive control according to claim 2, characterized in that: In step S6, the formula for the evaluation function of the finite set model predictive control is as follows: ； In the formula, g is the evaluation function value. Based on the predicted current vector obtained in step S2 and the updated reference current in step S5, the evaluation function value g is calculated according to the formula of the evaluation function. The voltage vector that minimizes the value of g is selected from the voltage vector set as the optimal voltage vector. The inverter drives the motor to operate through the optimal voltage vector. , The predicted current vector for the next sampling period. and These are the reference current vectors in the dq-axis coordinate system with a sampling period of k+1.

11. A motor drive system, comprising an inverter, a motor, and a controller, characterized in that, The controller is configured to perform the motor field weakening control method based on finite set model predictive control as described in any one of claims 1 to 10.

12. A computer-readable storage medium having a program stored thereon, characterized in that: When the program is executed by the processor, it implements the motor field weakening control method based on finite set model predictive control as described in any one of claims 1 to 10.

13. A computer program product comprising a computer program / instructions, characterized in that, When the computer program / instructions are executed by the processor, they implement the steps of the method described in any one of claims 1 to 10.

Images