Permanent magnet synchronous motor position sensorless control method and system

Abstract

The application belongs to the technical field of permanent magnet synchronous motor control, and particularly relates to a permanent magnet synchronous motor position sensorless control method and system. The method comprises the following steps: a mathematical model of back electromotive force of the permanent magnet synchronous motor is established, and a discrete model of stator current of the permanent magnet synchronous motor is obtained through discretization processing; a sliding surface is defined according to current estimation error, and a reference sliding surface is designed based on a super-twisting algorithm; a deadbeat predictive sliding mode observer is designed to realize estimation of the motor back electromotive force; a rotor position estimation error signal and a cost function are established according to the estimated motor back electromotive force; a limited position set phase-locked loop based on golden section search is designed based on the rotor position estimation cost function, the rotor position search interval is updated through iteration, the motor rotor position estimation value is obtained, and the motor speed estimation value is obtained. The application improves the estimation accuracy of the system for the motor rotor position and speed, and ensures that the system has strong robustness when the load is disturbed and the parameters are mismatched.

Description

Technical Field

[0001] This invention belongs to the field of permanent magnet synchronous motor control technology, and particularly relates to a sensorless control method and system for permanent magnet synchronous motors. Background Technology

[0002] In the medium-to-high speed range, permanent magnet synchronous motors typically employ a sensorless control method based on a back EMF model. This method first uses an observer to estimate the back EMF of the permanent magnet synchronous motor, and then extracts the rotor position and speed information contained in the back EMF through an arctangent function or a phase-locked loop.

[0003] Sliding mode observers are widely used for back EMF estimation in permanent magnet synchronous motors due to their simple structure, ease of implementation, and strong robustness. However, traditional sliding mode observers employ discontinuous switching functions, which can easily lead to chattering during back EMF estimation. Furthermore, their dynamic response and convergence speed can affect the accuracy of rotor position and velocity estimation. Predictive observation methods can improve the dynamic response of the observer by utilizing the discrete model of the motor. Among these, the deadbeat prediction method has advantages such as fast response speed and simple structure, but its estimation performance is somewhat dependent on the motor model and parameters, and may amplify noise in the sampled signal.

[0004] After obtaining the back electromotive force estimate, a phase-locked loop (PLL) is typically used to estimate the rotor position and speed. Traditional PLLs often employ fixed-gain proportional-integral (PI) controllers, which are difficult to tune, and their fixed bandwidth struggles to balance dynamic response and steady-state accuracy under different operating conditions. Finite-position set PLLs can transform rotor position estimation into an optimization problem of candidate positions, but existing methods usually require comparing a large number of candidate positions, resulting in significant computational overhead. While iterative optimization methods can reduce the number of candidate positions, some methods are susceptible to the influence of initial values ​​and measurement noise. Summary of the Invention

[0005] To overcome the shortcomings of the prior art, this invention provides a sensorless control method and system for permanent magnet synchronous motors. Addressing the slow convergence speed and chattering issues of traditional sliding mode observers, the invention introduces the super-twisted algorithm into the discrete deadbeat prediction process, designing a deadbeat prediction sliding mode observer to improve the observer's fast and accurate estimation of rotational speed. Furthermore, to address the difficulty in tuning traditional phase-locked loop parameters, a finite position set phase-locked loop based on golden section search is designed, transforming rotor position estimation into a single-peak optimization problem to obtain the rotor position estimate. A tracking differentiator is used to extract motor speed information, which improves the system's estimation accuracy of motor rotor position and speed while ensuring strong robustness under load disturbances and parameter mismatches.

[0006] To achieve the above objectives, one or more embodiments of the present invention provide the following technical solutions: The first aspect of this invention provides a sensorless control method for a permanent magnet synchronous motor.

[0007] A sensorless control method for permanent magnet synchronous motors includes the following steps: A mathematical model of the back electromotive force of a permanent magnet synchronous motor is established, and the stator current discrete model of the permanent magnet synchronous motor is obtained by discretization. The sliding surface is defined based on the current estimation error, and a reference sliding surface is designed based on the super-twisting algorithm; Based on the mathematical model of the back EMF of a permanent magnet synchronous motor and the reference sliding surface, a deadbeat prediction sliding observer is designed based on the deadbeat prediction principle to realize the estimation of the back EMF of the motor. Based on the estimated back electromotive force of the motor, the rotor position estimation error signal and cost function are established; Based on the rotor position estimation cost function, a finite position set phase-locked loop based on the golden section search is designed. By iteratively updating the rotor position search interval, the estimated value of the motor rotor position is obtained, and the estimated value of the motor speed is obtained by using a tracking differentiator, so as to realize sensorless control of the permanent magnet synchronous motor system.

[0008] The second aspect of the present invention provides a sensorless control system for a permanent magnet synchronous motor.

[0009] The sensorless control system for permanent magnet synchronous motors includes: The model building module is configured to: establish a mathematical model of the back electromotive force of the permanent magnet synchronous motor and discretize it to obtain a discrete model of the stator current of the permanent magnet synchronous motor. The sliding surface design module is configured to: define the sliding surface based on the current estimation error and design a reference sliding surface based on the super-twisting algorithm; The observer design module is configured to: based on the mathematical model of the back EMF of the permanent magnet synchronous motor and the reference sliding surface, design a deadbeat prediction sliding mode observer based on the deadbeat prediction principle to realize the estimation of the back EMF of the motor; The position cost function construction module is configured to: establish the rotor position estimation error signal and cost function based on the estimated motor back electromotive force; The phase-locked loop (PLL) design module is configured to: design a finite position set PLL based on the golden section search based on the rotor position estimation cost function; obtain the estimated value of the motor rotor position by iteratively updating the rotor position search interval; and obtain the estimated value of the motor speed by using a tracking differentiator, thereby realizing sensorless control of the permanent magnet synchronous motor system.

[0010] A third aspect of the present invention provides a computer-readable storage medium having a program stored thereon, which, when executed by a processor, implements the steps of the sensorless control method for a permanent magnet synchronous motor as described in the first aspect of the present invention.

[0011] The fourth aspect of the present invention provides an electronic device including a memory, a processor, and a program stored in the memory and executable on the processor, wherein the processor executes the program to implement the steps of the sensorless control method for a permanent magnet synchronous motor as described in the first aspect of the present invention.

[0012] The above one or more technical solutions have the following beneficial effects: This invention provides a sensorless control method and system for permanent magnet synchronous motors. Addressing the slow convergence speed and chattering issues of traditional sliding mode observers, this invention introduces the super-twisting algorithm into the discrete deadbeat prediction process. By constructing a reference sliding surface for the next sampling moment and making the predicted sliding surface track the reference sliding surface, a deadbeat prediction sliding mode observer is designed. This observer retains the fast response speed of deadbeat prediction while reducing sliding mode chattering in the back EMF estimation process, thus improving the estimation accuracy and dynamic response performance of the back EMF.

[0013] This invention constructs a rotor position cost function based on the estimated back electromotive force, transforming rotor position estimation into a single-peak optimization problem, and employs a golden section search algorithm to successively narrow the rotor position search interval. Compared with finite position set PLLs that compare all candidate rotor positions one by one, the designed finite position set PLL can reduce the computational load in the rotor position optimization process; at the same time, it eliminates the need for the proportional-integral regulator in traditional PLLs, avoiding the difficulties in proportional-integral parameter tuning and the limitations of fixed bandwidth on position estimation performance.

[0014] This invention also utilizes a tracking differentiator to process the rotor position estimate, reducing the impact of high-frequency noise on the speed estimation result during the position differentiation process while obtaining the speed estimate. Through the synergistic effect of the deadbeat predictive sliding mode observer, the finite position phase-locked loop, and the tracking differentiator, the estimation accuracy of the rotor position and speed of the permanent magnet synchronous motor can be improved, and the dynamic performance of the sensorless control system can be enhanced.

[0015] Advantages of additional aspects of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description

[0016] The accompanying drawings, which form part of this invention, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an improper limitation of the invention.

[0017] Figure 1 This is a flowchart of a sensorless control method for a permanent magnet synchronous motor in an embodiment of the present invention.

[0018] Figure 2 This is a structural block diagram of the beatless prediction sliding mode observer in an embodiment of the present invention.

[0019] Figure 3 This is a flowchart of a finite location set phase-locked loop based on the golden section search algorithm in an embodiment of the present invention.

[0020] Figure 4 This is a system block diagram of a sensorless control method for a permanent magnet synchronous motor in an embodiment of the present invention.

[0021] Figure 5 The graphs show the estimated speed, rotational speed error, and position error of the three sliding mode observers when the expected rotational speed is 200 rpm in the embodiments of the present invention.

[0022] Figure 6 The graphs show the estimated speed, rotational speed error, and position error of the three sliding mode observers when the expected rotational speed is 500 rpm in the embodiments of the present invention.

[0023] Figure 7 The graphs show the estimated speed, rotational speed error, and position error of the three sliding mode observers when the expected rotational speed is 800 rpm in the embodiments of the present invention.

[0024] Figure 8 The graphs show the speed, rotational speed error, and position estimation error curves of the two observers in the embodiments of the present invention during acceleration and deceleration.

[0025] Figure 9 In this embodiment of the invention, when the desired rotational speed is 200 rpm, a sudden increase of 2... The graphs show the speed, rotational speed error, and position estimation error of the three phase-locked loops after the load speed is measured.

[0026] Figure 10 In this embodiment of the invention, when the desired rotational speed is 500 rpm, a sudden increase of 2... The graphs show the speed, rotational speed error, and position estimation error of the three phase-locked loops after the load speed is measured.

[0027] Figure 11 In this embodiment of the invention, when the desired rotational speed is 800 rpm, a sudden increase of 2... The graphs show the speed, rotational speed error, and position estimation error of the three phase-locked loops after the load speed is measured.

[0028] Figure 12 This is a graph showing the speed error and position error obtained by using the control method of this invention when the desired speed is 200 rpm, after changing the speed of the stator inductance and stator resistance.

[0029] Figure 13This is a graph showing the speed error and position error obtained by using the control method of this invention when the desired speed is 500 rpm, after changing the speed of the stator inductance and stator resistance.

[0030] Figure 14 This is a graph showing the speed error and position error obtained by using the control method of this invention when the desired speed is 800 rpm, after changing the speed of the stator inductance and stator resistance. Detailed Implementation

[0031] It should be noted that the following detailed descriptions are exemplary and intended to provide further illustration of the invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0032] It should be noted that the terminology used herein is for the purpose of describing particular implementations only and is not intended to limit the exemplary implementations of the present invention.

[0033] Where there is no conflict, the embodiments and features in the embodiments of the present invention can be combined with each other.

[0034] Example 1 Permanent magnet synchronous motors (PMSMs) possess advantages such as high power density, small size, high efficiency, and high reliability, and have been widely used in new energy vehicles, aerospace, intelligent robots, and industrial drives. Achieving high-performance control of PMSMs requires accurate acquisition of the rotor's position and speed information. Traditional methods typically employ mechanical position sensors such as encoders, resolvers, and Hall effect sensors to measure rotor position. However, mechanical position sensors and their associated signal processing equipment increase the size, cost, and structural complexity of the motor drive system, and are susceptible to external environmental factors such as vibration, high temperature, high humidity, and electromagnetic interference, thereby reducing the reliability of the motor drive system.

[0035] To address the aforementioned problems associated with mechanical position sensors, sensorless control technology for permanent magnet synchronous motors has gradually emerged as a reliable solution. This technology avoids reliance on mechanical sensors, reducing costs and providing accurate position estimation. In the medium-to-high speed operating range, sensorless control methods based on back-EMF models are commonly employed. However, these methods suffer from the problems described in the background section.

[0036] Therefore, improving the convergence speed and dynamic response of back EMF estimation for permanent magnet synchronous motors, and enhancing the estimation accuracy of rotor position and speed, are problems that need to be solved in sensorless control of permanent magnet synchronous motors.

[0037] Based on this, this embodiment provides a sensorless control method for permanent magnet synchronous motors. Specifically, this method is a sensorless control method for permanent magnet synchronous motors based on a deadbeat predictive sliding mode observer and a finite position phase-locked loop. The general idea is as follows: First of all, In a two-phase stationary coordinate system, a mathematical model of the back electromotive force of a permanent magnet synchronous motor is established, and the established mathematical model is discretized to obtain a discrete model of the stator current of the permanent magnet synchronous motor.

[0038] Secondly, in order to suppress chattering in traditional sliding mode observers and improve the estimation accuracy of the observers, thereby improving the system's estimation accuracy of rotor speed, the present invention introduces the concept of deadbeat prediction and the super-torque sliding mode control method to design a deadbeat prediction sliding mode observer to achieve accurate estimation of the motor's back electromotive force.

[0039] Then, based on the estimated back electromotive force, a rotor position cost function is established, transforming the rotor position estimation into a single-peak optimization problem. This invention also designs a finite position set phase-locked loop based on the golden section search algorithm to obtain the rotor position estimate. Meanwhile, the tracking differentiator is used to extract the speed estimate, thereby further improving the system's estimation accuracy of rotor position and speed.

[0040] Finally, experimental results in this embodiment show that the sensorless control method for permanent magnet synchronous motors proposed in this invention can not only effectively improve the estimation accuracy of rotor position and speed, but also ensure that the system has strong robustness under load disturbances and parameter mismatches.

[0041] The method of the present invention will now be explained in detail with reference to the accompanying drawings.

[0042] like Figure 1 As shown, the sensorless control method for permanent magnet synchronous motors provided in this embodiment includes the following steps: Step 1. In A mathematical model of the back electromotive force of a permanent magnet synchronous motor is established in a two-phase stationary coordinate system, wherein... , The directions of the two stationary coordinate axes are respectively represented, and the back electromotive force mathematical model is discretized to obtain the stator current discrete model of the permanent magnet synchronous motor. Step 2. Design the sliding surface based on the estimated and actual values ​​of the stator current of the permanent magnet synchronous motor, and design the reference sliding surface for the next sampling moment based on the super-twist algorithm; Step 3. Based on the stator current discrete model established in Step 1 and the reference sliding surface designed in Step 2, construct the cost function between the prediction sliding surface and the reference sliding surface, and design a deadbeat prediction sliding observer based on the deadbeat prediction principle to realize the estimation of the back electromotive force of the permanent magnet synchronous motor. Step 4. Based on the back electromotive force estimate obtained in Step 3, establish the rotor position equivalent error and rotor position cost function, and transform the estimation of the permanent magnet synchronous motor rotor position into a single-peak optimization problem; Step 5. Design a finite position set phase-locked loop based on the golden section search algorithm, obtain the rotor position estimate of the permanent magnet synchronous motor by iteratively updating the rotor position search interval, and obtain the speed estimate of the permanent magnet synchronous motor by using a tracking differentiator. Step 6. Using the deadbeat predictive sliding mode observer designed in Step 3 and the finite position phase-locked loop designed in Step 5, perform sensorless control of the permanent magnet synchronous motor system.

[0043] The following describes the detailed steps of the sensorless control method for permanent magnet synchronous motors proposed in this invention.

[0044] Step 1. In A mathematical model of the back electromotive force of a permanent magnet synchronous motor is established in a two-phase stationary coordinate system.

[0045] exist In the two-phase stationary coordinate system, a mathematical model of the back electromotive force of the permanent magnet synchronous motor as shown in formula (1) is established, and the mathematical model of the back electromotive force is discretized to obtain a discrete model of the stator current of the permanent magnet synchronous motor: (1) in, , In the two-phase stationary coordinate system respectively axis, Stator voltage of the shaft. This is the stator resistance. , In the two-phase stationary coordinate system respectively axis, Stator current of the shaft. It is the stator inductance. Indicates time. , They are respectively The back electromotive force of the motor in the two-phase stationary coordinate system of the shaft, and satisfying: (2) in, It is a permanent magnet flux linkage. This indicates the electric angular velocity of the motor. This indicates the electrical angle of the motor.

[0046] The forward Euler method is used to discretize formula (1), and the discrete model of the stator current of the permanent magnet synchronous motor is obtained as follows: (3) in, The sampling period; Indicates the current sampling time; Indicates the next sampling time; and The first Each sampling time shaft and Shaft stator current; and The first Each sampling time shaft and Shaft stator current; and The first Each sampling time shaft and Shaft-stator voltage; and The first Each sampling time shaft and Axis back electromotive force. Indicates stator inductance; This indicates the stator resistance.

[0047] Step 2. Design the sliding surface based on the estimated and actual values ​​of the stator current of the permanent magnet synchronous motor, and design the reference sliding surface for the next sampling moment based on the super-twisting algorithm.

[0048] To improve the estimation accuracy of back EMF and reduce chattering during the observation process, this invention introduces a super-twisting algorithm into the beatless predictive observer.

[0049] by Taking the shaft as an example, the error between the estimated stator current and the actual sampled value is defined as the sliding surface: (4) in, Indicates the first Each sampling time Shaft sliding surface; Indicates the first Each sampling time Estimated value of shaft stator current; Indicates the first Each sampling time Shaft stator current; This indicates the sequence number of the discrete sampling time.

[0050] No. The predicted sliding surface at each sampling time is represented as: (5) in, Indicates the first Each sampling time Shaft prediction sliding surface; Indicates the first Each sampling time Estimated value of shaft stator current; Indicates the first Each sampling time Actual value of shaft stator current; This indicates the sequence number of the discrete sampling time.

[0051] To reduce chattering caused by high-frequency switching of the sign function and improve the convergence speed of the sliding surface, this invention designs the first... The reference sliding surface at each sampling time is: (6) in, Indicates the first Each sampling time Shaft reference sliding surface; Indicates the first Each sampling time Shaft sliding surface; This represents the first sliding mode gain in the super-twist algorithm; This represents the second sliding mode gain in the super-twist algorithm; For the first The state variables of the super-twisted algorithm at each sampling time. For the first The state variables of the super-twisting algorithm at each sampling time; Indicates the sampling period; Represents a symbolic function; This indicates the sequence number of the discrete sampling time.

[0052] Using the same method as formulas (4) to (6), construct The sliding surface of the shaft and its reference sliding surface at the next sampling moment.

[0053] Step 3. Based on the stator current discrete model established in Step 1 and the reference sliding surface designed in Step 2, construct the cost function between the prediction sliding surface and the reference sliding surface, and design a deadbeat prediction sliding observer based on the deadbeat prediction principle to realize the estimation of the back electromotive force of the permanent magnet synchronous motor.

[0054] Specifically, in this embodiment, based on the sliding surface designed in step 2, the process of designing a deadbeat prediction sliding mode observer using the deadbeat prediction principle is as follows: In sliding mode predictive control, make the first Predicted sliding surface at each sampling time Tracking reference sliding surface The cost function is established as follows: (7) in, Let $\frac{ ... Indicates the first Each sampling time Shaft reference sliding surface; Indicates the first Each sampling time Axis prediction of sliding surface.

[0055] Based on the flawless prediction principle, let the cost function... Then the first The sliding surface at each sampling time satisfies .

[0056] Considering the first The sampling time cannot directly obtain the first... The actual value of the stator current at the sampling time is adopted. Stator current at each sampling time replace .

[0057] Combining equations (3), (5), and (7), we obtain... The estimated value of the shaft back electromotive force is: (8) in, Indicates the first Each sampling time Back EMF estimate; Indicates the first Each sampling time Shaft-stator voltage; Indicates stator inductance; Indicates the sampling period; Indicates the first Each sampling time Shaft stator current; Indicates stator resistance. Indicates the first Each sampling time Estimated value of shaft stator current.

[0058] Substituting the back electromotive force estimate shown in formula (8) into the stator current discrete model shown in formula (3), we obtain the first... Each sampling time The estimated value of the shaft stator current is: (9) in, Indicates the first Each sampling time Estimated value of shaft stator current.

[0059] From formulas (8) and (9), it can be seen that the deadbeat predictive sliding mode observer calculates the back electromotive force estimate based on the stator voltage, stator current, stator current estimate, and reference sliding surface at the current sampling time, and updates the stator current estimate at the next sampling time using the back electromotive force estimate. Similarly, it can be obtained that... The back electromotive force estimate of the shaft and the first Each sampling time The estimated value of the shaft stator current is used to obtain the permanent magnet synchronous motor's current at the time of operation. Estimate of back electromotive force in a two-phase stationary coordinate system.

[0060] This invention determines the reference sliding surface at the next sampling time using a super-twisting algorithm and directly calculates the back EMF estimate based on the deadbeat prediction principle. This approach retains the fast response characteristics of deadbeat prediction while reducing chattering during the back EMF estimation process. The block diagram of the designed deadbeat prediction sliding mode observer is shown below. Figure 2 As shown.

[0061] Furthermore, after completing the design of the deadbeat prediction sliding mode observer, in order to ensure the finite-time stability of the deadbeat prediction sliding mode observer, Taking the axis as an example, the observer gain is selected as follows.

[0062] In the actual operation of a permanent magnet synchronous motor, the impact of parameter variations, model uncertainties, and measurement noise on the stator current estimation error is defined as lumped disturbance. The lumped disturbance satisfies: (10) in, This represents the upper bound coefficient of the lumped perturbation. Based on the finite-time convergence condition of the super-twisting algorithm, the sliding mode gain... and satisfy: (11) Under the lumped disturbance condition shown in formula (10) and the sliding mode gain condition shown in formula (11), the designed deadbeat prediction sliding mode observer satisfies the finite-time stability condition. The stability of the shaft can be obtained similarly.

[0063] Step 4. Based on the back EMF estimate obtained in Step 3, establish the rotor position equivalent error and rotor position cost function, and transform the estimation of the permanent magnet synchronous motor rotor position into a single-peak optimization problem.

[0064] To obtain the rotor position of the permanent magnet synchronous motor based on the back electromotive force estimate, a system is established... The axis is used to estimate the rotating coordinate system, where, shaft and The axes represent the direct and quadrature axes of the estimated rotating coordinate system, respectively. The rotor position estimation error is defined as: (12) in, This indicates the position estimation error of the motor. This indicates the actual electrical angle of the motor. This indicates the estimated electrical angle of the motor.

[0065] (13) Correspondingly, shaft and The equivalent positional error between the shafts is: (14) in, express shaft and Equivalent positional error between axes express shaft and Equivalent positional error between axes. and They are respectively shaft and Estimated back electromotive force of the shaft. This represents the estimated magnitude of the back electromotive force. This indicates the estimated electrical angular velocity of the motor. This indicates the position estimation error of the motor. This indicates the actual electrical angle of the motor. This indicates the estimated electrical angle of the motor.

[0066] Based on formulas (13) and (14), the rotor position cost function is established as follows: (15) in, This represents the rotor position cost function. express shaft and Equivalent positional error between axes express shaft and Equivalent positional error between axes. This represents the estimated magnitude of the back electromotive force.

[0067] As can be seen from formula (15), within the rotor position estimation error range Within this function, the rotor position cost function has a unique zero. When the rotor position estimation error... When the rotor position cost function reaches its minimum value, the estimated rotor electrical angle is equal to the actual rotor electrical angle.

[0068] Therefore, by searching for the minimum value of the rotor position cost function within the rotor position estimation error interval, the rotor position estimate of the permanent magnet synchronous motor can be determined, and the rotor position estimation is transformed into an optimization problem of a single-peak function.

[0069] Step 5. Design a finite position set phase-locked loop based on the golden section search algorithm. By iteratively updating the rotor position search interval, obtain the rotor position estimate of the permanent magnet synchronous motor, and use the tracking differentiator to obtain the speed estimate of the permanent magnet synchronous motor.

[0070] Specifically, according to formula (15), the rotor position cost function is in the interval It has a unique minimum value. The golden section search algorithm is used to iteratively optimize the rotor position cost function to obtain the rotor position estimate of the permanent magnet synchronous motor. Let the first... The rotor position search interval for the next iteration is The initial search range is set as follows: (16) in, and These represent the left and right endpoints of the initial search interval, respectively. Indicates the first i The left endpoint of the rotor position search interval in the next iteration. Indicates the first i The right endpoint of the rotor position search interval in the next iteration. This represents the left endpoint of the initial search interval. This represents the right endpoint of the initial search interval.

[0071] In the In this iteration, two candidate rotor positions are selected within the search interval according to the golden ratio: (17) in, and They represent the first The first and second candidate rotor positions during the next iteration. It is the golden ratio, and Substituting the two candidate rotor positions into formula (15) respectively, we obtain the corresponding rotor position cost function values. and The rotor position search interval is updated based on the comparison results of the two cost function values: (18) in, Indicates the first The rotor position search interval for the next iteration. This represents the rotor position cost function value corresponding to the first candidate rotor position. This represents the rotor position cost function value corresponding to the second candidate rotor position. Indicates the first i+1 The left endpoint of the rotor position search interval in the next iteration. Indicates the first i+1 The right endpoint of the rotor position search interval in the next iteration.

[0072] When the cost function value corresponding to the first candidate rotor position is greater than the cost function value corresponding to the second candidate rotor position, the retention interval is... Conversely, retain the interval. The search interval is repeatedly updated according to formula (18) so that the search interval is gradually reduced by a fixed ratio, and the minimum value of the rotor position cost function is always located within the updated search interval.

[0073] go through After the last iteration, the midpoint of the final rotor position search interval is taken as the rotor position estimate of the permanent magnet synchronous motor: (19) in, This indicates the preset number of iterations for the golden section search algorithm; and They represent the first After the next iteration, the left and right endpoints of the rotor position search interval are determined.

[0074] Using formulas (16) to (19), the rotor position estimation is transformed into a cost function optimization process within a finite search interval. The golden section search algorithm is then used to progressively narrow the search interval, thereby obtaining the rotor position estimate. The flowchart of the finite position set phase-locked loop based on the golden section search algorithm is shown below. Figure 3 As shown.

[0075] Since the output of the finite-position phase-locked loop (PLL) is a rotor position estimate, directly differentiating this estimate would amplify high-frequency noise during the differentiation process. To obtain a smooth rotational speed estimate, this invention employs a tracking differentiator to process the rotor position estimate. The tracking differentiator is expressed as: (20) in, This indicates the rotor position that tracks the output of the differentiator; This represents the estimated electrical angular velocity of the motor. It represents the first derivative of the estimated electric angular velocity of the motor. Indicates the gain of the tracking differentiator, and . This indicates the estimated electrical angle of the motor.

[0076] The tracking differentiator tracks the rotor position estimate obtained from a finite-position phase-locked loop, thus obtaining a speed estimate while reducing the impact of high-frequency noise. Increasing the gain of the tracking differentiator... It can improve the dynamic response speed of speed estimation, but it will reduce the ability to suppress high-frequency noise. Therefore, the gain of the tracking differentiator needs to be selected according to the dynamic response and noise suppression requirements of speed estimation.

[0077] Step 6. Using the deadbeat predictive sliding mode observer designed in Step 3 and the finite position phase-locked loop designed in Step 5, perform sensorless control of the permanent magnet synchronous motor system.

[0078] like Figure 4 The system block diagram of the sensorless control method for permanent magnet synchronous motor is shown. The present invention uses a deadbeat predictive sliding mode observer and a finite position set phase-locked loop based on the golden section search algorithm to realize sensorless control of the permanent magnet synchronous motor system.

[0079] Figure 4 middle, Indicates the preset speed of the motor. Indicates the estimated speed. Indicates preset q Shaft reference current, Indicates motor q Actual shaft current, Indicates preset d Shaft reference current, Indicates motor d Actual shaft current, express q Shaft reference voltage, express d Shaft reference voltage, and They represent shaft and Shaft stator voltage, and They represent shaft and Shaft stator current, Indicates the output pulse signal. Indicates DC voltage. The coordinate systems are represented by the direct axis and the quadrature axis. SVPWM represents space vector pulse width modulation, and PMSM represents permanent magnet synchronous motor.

[0080] according to Figure 4 As can be seen, the method of this invention utilizes the deadbeat predictive sliding mode observer designed in step 3 to estimate the back electromotive force (EMF) in the motor, utilizes the finite position phase-locked loop designed in step 5 to extract the rotor position information contained in the back EMF, and utilizes the tracking differentiator to process the rotor position estimate to obtain the rotor speed estimate. That is, the deadbeat predictive sliding mode observer is used to estimate the back EMF of the permanent magnet synchronous motor, and the finite position phase-locked loop and tracking differentiator are used to estimate the rotor position and speed of the permanent magnet synchronous motor. In this embodiment, both the speed loop and the current loop use the PI control algorithm. The d-axis uses... =0 control to ensure constant magnetic flux.

[0081] To better verify the accuracy of the control method designed in this invention for rotor position and speed estimation, as well as its robustness under load disturbances and parameter mismatches, the deadbeat predictive sliding mode observer proposed in this invention is compared with the super-torsion sliding mode observer and the deadbeat predictive observer. The finite position set phase-locked loop proposed in this invention is also compared with the proportional-integral phase-locked loop and the extended state observer.

[0082] First, to verify the performance of the fractional-order high-order sliding mode observer proposed in this invention, measurements were taken at 200... 500 With 800 Three sets of comparative experiments were set up. The first set used a super-torsional sliding mode observer, the second set used a deadbeat predictive observer, and the third set used the deadbeat predictive sliding mode observer proposed in this invention. All three sets of experiments used the finite-position phase-locked loop proposed in this invention to extract rotor position and speed information from the back electromotive force.

[0083] Experiment 1 involved the motor running stably at 200... 500 800 Experiments with different sliding mode observers.

[0084] The experimental results under the three different conditions are described below: The back electromotive force (EMF) of the permanent magnet synchronous motor (PMSM) was estimated using a super-torsional sliding mode observer, a deadbeat predictive observer, and the deadbeat predictive sliding mode observer proposed in this invention, respectively. Furthermore, a finite-position phase-locked loop (PLL) was used to extract the rotor position and velocity information contained within the back EMF. The estimated velocity, velocity estimation error, and rotor position estimation error of different back EMF observers are shown below. Figures 5 to 7 As shown.

[0085] like Figure 5 As shown, when the permanent magnet synchronous motor operates at the desired speed of 200... At that time, the estimated velocity fluctuations of the super-torsional sliding mode observer, the deadbeat prediction observer, and the deadbeat prediction sliding mode observer proposed in this invention were 39.2%. 26.7 20.5 The speed error fluctuation was 48.6%. 32.6 26.6 The rotor position estimation error fluctuation was 0.45%. 0.29 0.21 .

[0086] like Figure 6 As shown, when the permanent magnet synchronous motor operates at the desired speed of 600... At that time, the estimated velocity fluctuations of the super-torsional sliding mode observer, the deadbeat prediction observer, and the deadbeat prediction sliding mode observer proposed in this invention are 10.4. 6.5 4.4 The speed error fluctuation was 21.8%. 20.1 19.2 The rotor position estimation error fluctuation was 0.21. 0.12 0.07 .

[0087] like Figure 7 As shown, when the permanent magnet synchronous motor operates at the desired speed of 1000... At that time, the estimated velocity fluctuations of the super-torsional sliding mode observer, the deadbeat prediction observer, and the deadbeat prediction sliding mode observer proposed in this invention were 7.2. 4.1 3.4 The speed error fluctuation was 28.1%. 24.6 22.9 The rotor position estimation error fluctuation was 0.25%. 0.11 0.08 .

[0088] As can be seen from the results of Experiment 1 above, the sensorless control method for permanent magnet synchronous motors of the present invention can effectively suppress chattering and improve the estimation accuracy of rotor position and speed. Compared with the super-twisted sliding mode observer and the deadbeat predictive observer, the deadbeat predictive sliding mode observer proposed in this invention has smaller estimated speed fluctuations, speed estimation errors and rotor position estimation errors, and has better steady-state estimation performance.

[0089] Secondly, to verify the dynamic performance of the deadbeat predictive sliding mode observer proposed in this invention, a comparison was made between the deadbeat predictive observer and the deadbeat predictive sliding mode observer proposed in this invention. Both observers utilize the finite-position phase-locked loop proposed in this invention to extract the rotor position and velocity information contained in the back electromotive force.

[0090] Experiment 2 is an experiment using different observers during the acceleration and deceleration of a permanent magnet synchronous motor.

[0091] During the experiment, the desired motor speed was 300. -600 -900 -600 –300 Changes. The estimated velocity, velocity estimation error, and rotor position estimation error obtained by the deadbeat predictive observer and the deadbeat predictive sliding mode observer proposed in this invention are as follows: Figure 8 As shown.

[0092] like Figure 8 As shown, both observers can track the motor speed when the desired motor speed changes. The maximum position error of the deadbeat predictive observer during acceleration and deceleration is 0.36. 0.29 The deadbeat predictive sliding mode observer proposed in this invention achieves smaller overshoot and faster convergence speed, demonstrating better dynamic performance.

[0093] The results of Experiment 2 above show that the deadbeat predictive sliding mode observer proposed in this invention can accurately estimate the rotor position and speed during motor acceleration and deceleration, and has better dynamic estimation performance than the deadbeat predictive observer.

[0094] To verify the disturbance rejection performance of the proposed sensorless permanent magnet synchronous motor control method, a sudden load application experiment was also designed in this embodiment. During the experiment, the load torque suddenly changed from 0 to 2. It becomes 0 again after 5 seconds. .

[0095] Experiment 3 is a sudden load test.

[0096] Figures 9 to 11 The permanent magnet synchronous motors are given respectively when operating at the desired speed of 200. 500 800 Results of a sudden load application experiment. Specifically, during the experiment, the load torque suddenly changed from 0 to 2. It becomes 0 again after 5 seconds. .

[0097] like Figure 10 As shown, when the permanent magnet synchronous motor operates at the desired speed of 200... During a sudden load application, the maximum rotor position estimation errors of the proportional-integral phase-locked loop, the expanded state observer, and the finite-position collection phase-locked loop proposed in this invention are 0.37 and 0.37, respectively. 0.33 0.11 During the sudden load unloading process, the maximum rotor position estimation errors of the three methods were 0.16 and 0.16, respectively. 0.19 0.10 .

[0098] like Figure 11 As shown, when the permanent magnet synchronous motor operates at the desired speed of 500... During a sudden load application, the maximum rotor position estimation errors of the proportional-integral phase-locked loop, the expanded state observer, and the finite-position collection phase-locked loop proposed in this invention are 0.15, respectively. 0.16 0.12 During the sudden load unloading process, the maximum rotor position estimation errors of the three methods were 0.08 and 0.08, respectively. 0.10 0.04 .

[0099] like Figure 9 As shown, when the permanent magnet synchronous motor operates at the desired speed of 800... During a sudden load application, the maximum rotor position estimation errors of the proportional-integral phase-locked loop, the expanded state observer, and the finite-position collection phase-locked loop proposed in this invention are 0.08, respectively. 0.09 0.06 During the sudden load unloading process, the maximum rotor position estimation errors of the three methods were 0.09 and 0.09, respectively. 0.08 0.05 .

[0100] The results of Experiment 3 above demonstrate that, during sudden load application and unloading, the finite position phase-locked loop (PLL) proposed in this invention can still accurately estimate the rotor position and speed of the permanent magnet synchronous motor. Compared with proportional-integral (PIL) PLLs and extended state observers, the finite position PLL proposed in this invention has a smaller rotor position estimation error and a faster convergence speed, exhibiting good resistance to load disturbances.

[0101] Finally, parameters such as stator resistance and stator inductance of the permanent magnet synchronous motor inevitably change. Mismatches in these parameters will lead to reduced rotor position estimation accuracy, thereby degrading the overall system performance. To verify the robustness of the permanent magnet synchronous motor system controlled by the method of this invention under conditions of parameter uncertainty, experiments were conducted with stator inductance and resistance mismatched, with the stator inductance and resistance at a range of 0.5... Up to 1.5 Variations within the range.

[0102] Experiment 4 is a variable parameter experiment.

[0103] Figures 12 to 14 The sensorless control method for permanent magnet synchronous motors using the present invention was demonstrated in 200 500rpm, 800 At the desired rotational speed, after changing the rotational speed of the stator inductance and stator resistance, i.e., when the stator inductance and stator resistance are at 0.5... Up to 1.5 The experimental results show the changes between them. Based on... Figures 12 to 14 It can be seen that when the stator inductance and stator resistance change, the sensorless control method for permanent magnet synchronous motors of the present invention can still maintain a small speed fluctuation in the permanent magnet synchronous motor system and still have high rotor position and speed tracking accuracy.

[0104] The results of Experiment 4 above demonstrate that the sensorless control method for permanent magnet synchronous motors of the present invention has strong robustness in dealing with parameter mismatch.

[0105] The above four sets of experimental results demonstrate that the sensorless control method for permanent magnet synchronous motors proposed in this invention can not only improve the steady-state estimation accuracy of rotor position and speed of permanent magnet synchronous motors, but also has good dynamic estimation performance and can ensure that the system has strong robustness under load disturbances and parameter mismatches.

[0106] In this embodiment, a deadbeat predictive sliding mode observer is first designed based on the discrete mathematical model of the permanent magnet synchronous motor (PMSM) and the super-twisted algorithm to estimate the back electromotive force (EMF) of the PMSM. Then, a rotor position cost function is established based on the back EMF estimate, and a finite position set phase-locked loop (PLL) is designed using the golden section search algorithm to obtain the rotor position estimate. Simultaneously, a tracking differentiator is used to obtain the rotational speed estimate. Experimental results show that the sensorless control method for PMSM proposed in this invention can achieve accurate estimation of rotor position and speed, and has good dynamic performance and robustness.

[0107] Example 2 This embodiment provides a sensorless control system for a permanent magnet synchronous motor, which is based on the same inventive concept as the sensorless control method for a permanent magnet synchronous motor in Embodiment 1.

[0108] Specifically, the sensorless control system for this permanent magnet synchronous motor includes the following modules: Model building module, used in In the two-phase stationary coordinate system, a mathematical model of the back electromotive force of the permanent magnet synchronous motor is established, and the mathematical model of the back electromotive force of the permanent magnet synchronous motor is discretized to obtain a discrete model of the stator current of the permanent magnet synchronous motor. The sliding surface design module is used to design the sliding surface based on the estimated and actual values ​​of the stator current of the permanent magnet synchronous motor, and to design the reference sliding surface for the next sampling moment based on the super-twist algorithm; The observer design module is used to establish the cost function between the sliding surface and the reference sliding surface at the next sampling time based on the discrete model of the stator current of the permanent magnet synchronous motor and the reference sliding surface, and to design a deadbeat prediction sliding observer based on the deadbeat prediction principle to realize the estimation of the back electromotive force of the permanent magnet synchronous motor. The position cost function construction module is used to establish the rotor position equivalent error and rotor position cost function based on the back EMF estimate obtained from the deadbeat prediction sliding mode observer, and to transform the estimation of the permanent magnet synchronous motor rotor position into a unimodal optimization problem of the rotor position cost function. The phase-locked loop (PLL) design module is used to design a finite-position set PLL based on the golden section search algorithm. By comparing the cost function values ​​corresponding to different candidate rotor positions, the rotor position search interval is iteratively updated to obtain the rotor position estimate of the permanent magnet synchronous motor. The tracking differentiator is then used to process the rotor position estimate to obtain the speed estimate of the permanent magnet synchronous motor. The system also includes a sensorless control module for estimating the back electromotive force of the permanent magnet synchronous motor using a deadbeat predictive sliding mode observer, obtaining rotor position and speed estimates of the permanent magnet synchronous motor using a finite position set phase-locked loop and a tracking differentiator, and using the rotor position and speed estimates for vector control of the permanent magnet synchronous motor to achieve sensorless control of the permanent magnet synchronous motor.

[0109] Example 3 The purpose of this embodiment is to provide a computer-readable storage medium.

[0110] A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the sensorless control method for a permanent magnet synchronous motor as described in Embodiment 1 of this disclosure.

[0111] Example 4 The purpose of this embodiment is to provide an electronic device.

[0112] An electronic device includes a memory, a processor, and a program stored in the memory and executable on the processor. When the processor executes the program, it implements the steps in the sensorless control method for a permanent magnet synchronous motor as described in Embodiment 1 of this disclosure.

[0113] The steps and methods involved in the apparatuses of Embodiments 2, 3, and 4 above correspond to those in Embodiment 1. For specific implementation details, please refer to the relevant description section of Embodiment 1. The term "computer-readable storage medium" should be understood as a single medium or multiple media including one or more instruction sets; it should also be understood as including any medium capable of storing, encoding, or carrying an instruction set for execution by a processor and enabling the processor to perform any of the methods in this invention.

[0114] Those skilled in the art will understand that the modules or steps of the present invention described above can be implemented using general-purpose computer devices. Optionally, they can be implemented using computer-executable program code, thereby allowing them to be stored in a storage device for execution by a computer device, or they can be fabricated as separate integrated circuit modules, or multiple modules or steps can be fabricated as a single integrated circuit module. The present invention is not limited to any particular combination of hardware and software.

[0115] While the specific embodiments of the present invention have been described above in conjunction with the accompanying drawings, this is not intended to limit the scope of protection of the present invention. Those skilled in the art should understand that various modifications or variations that can be made by those skilled in the art without creative effort based on the technical solutions of the present invention are still within the scope of protection of the present invention.

Claims

1. A sensorless control method for a permanent magnet synchronous motor, characterized in that, Includes the following steps: A mathematical model of the back electromotive force of a permanent magnet synchronous motor is established, and the stator current discrete model of the permanent magnet synchronous motor is obtained by discretization. The sliding surface is defined based on the current estimation error, and a reference sliding surface is designed based on the super-twisting algorithm; Based on the mathematical model of the back EMF of a permanent magnet synchronous motor and the reference sliding surface, a deadbeat prediction sliding observer is designed based on the deadbeat prediction principle to realize the estimation of the back EMF of the motor. Based on the estimated back electromotive force of the motor, the rotor position estimation error signal and cost function are established; Based on the rotor position estimation cost function, a finite position set phase-locked loop based on the golden section search is designed. By iteratively updating the rotor position search interval, the estimated value of the motor rotor position is obtained, and the estimated value of the motor speed is obtained by using a tracking differentiator, so as to realize sensorless control of the permanent magnet synchronous motor system.

2. The sensorless control method for a permanent magnet synchronous motor as described in claim 1, characterized in that, exist A mathematical model of the back electromotive force of a permanent magnet synchronous motor is established in a two-phase stationary coordinate system, wherein... , The directions of the two stationary coordinate axes are represented respectively, and the forward Euler method is used to discretize the mathematical model of the back electromotive force of the permanent magnet synchronous motor, so as to obtain the discrete model of the stator current of the permanent magnet synchronous motor.

3. The sensorless control method for a permanent magnet synchronous motor as described in claim 1, characterized in that, The error between the estimated stator current and the actual sampled value is defined as the sliding surface: ； in, Indicates the first Each sampling time Shaft sliding surface; Indicates the first Each sampling time Estimated value of shaft stator current; Indicates the first Each sampling time Shaft stator current; Indicates the discrete sampling time sequence number; Design based on super-twisting algorithm The reference sliding surface at each sampling time is: ； in, Indicates the first Each sampling time Shaft reference sliding surface; Indicates the first Each sampling time Shaft sliding surface; This represents the first sliding mode gain in the super-twist algorithm; This represents the second sliding mode gain in the super-twist algorithm; For the first The state variables of the super-twisted algorithm at each sampling time. For the first The state variables of the super-twisting algorithm at each sampling time; Indicates the sampling period; Represents a symbolic function; This indicates the sequence number of the discrete sampling time.

4. The sensorless control method for a permanent magnet synchronous motor as described in claim 1, characterized in that, The deadbeat predictive sliding mode observer calculates the back electromotive force estimate based on the stator voltage, stator current, stator current estimate at the current sampling time and the reference sliding surface at the next sampling time, and uses the back electromotive force estimate to update the stator current estimate at the next sampling time. The design process of a deadbeat prediction sliding mode observer includes: In sliding mode predictive control, make the first Predicted sliding surface at each sampling time Tracking reference sliding surface Establish the cost function; According to the principle of errorless prediction, if we set the cost function to zero, then the... Predicted sliding surface at each sampling time With reference sliding surface same; Adopting the first Stator current at each sampling time Instead of the Stator current at each sampling time ; And thus obtain The estimated value of the shaft back electromotive force is: ； in, Indicates the first Each sampling time Back EMF estimate; Indicates the first Each sampling time Shaft-stator voltage; Indicates stator inductance; Indicates the sampling period; Indicates the first Each sampling time Shaft stator current; Indicates stator resistance; Indicates the first Each sampling time Estimated value of shaft stator current; No. Each sampling time Shaft stator current estimate for: 。 5. The sensorless control method for a permanent magnet synchronous motor as described in claim 1, characterized in that, Define the rotor position estimation error as: ； in, This indicates the position estimation error of the motor; This indicates the actual electrical angle of the motor; Indicates the estimated electrical angle of the motor; Establish the rotor position cost function: ； in, Represents the rotor position cost function; This indicates the position estimation error of the motor.

6. The sensorless control method for a permanent magnet synchronous motor as described in claim 5, characterized in that, The rotor position cost function is within the rotor position estimation error range. The rotor position cost function has a unique minimum value, and when the rotor position estimation error is zero, the rotor position cost function reaches its minimum value, thereby transforming the estimation of the permanent magnet synchronous motor rotor position into a single-peak optimization problem of the rotor position cost function.

7. The sensorless control method for a permanent magnet synchronous motor as described in claim 1, characterized in that, Based on the rotor position estimation cost function, a finite position set phase-locked loop based on the golden section search is designed. By iteratively updating the rotor position search interval, the estimated value of the motor rotor position is obtained, specifically including: The golden section search algorithm is used to optimize the rotor position cost function. Let the first part be... The rotor position search interval for the next iteration is The initial search range is set as follows: ； in, Indicates the first i The left endpoint of the rotor position search interval in the next iteration. Indicates the first i The right endpoint of the rotor position search interval in the next iteration; This represents the left endpoint of the initial search interval. Indicates the right endpoint of the initial search interval; In the During this iteration, the first candidate rotor position and the second candidate rotor position are selected within the rotor position search interval according to the golden ratio: ； in, and They represent the first The first and second candidate rotor positions during the next iteration; It is the golden ratio; Calculate the rotor position cost function values ​​corresponding to the first candidate rotor position and the second candidate rotor position, respectively. and The rotor position search interval is updated based on the comparison results of the two rotor position cost function values: ； in, Indicates the first The rotor position search interval for the next iteration; This represents the rotor position cost function value corresponding to the first candidate rotor position; This represents the rotor position cost function value corresponding to the second candidate rotor position; Indicates the first i+1 The left endpoint of the rotor position search interval in the next iteration; Indicates the first i+1 The right endpoint of the rotor position search interval in the next iteration; go through After the iteration, the midpoint of the final rotor position search interval is taken as the position estimate of the permanent magnet synchronous motor. The rotor position estimate is processed using a tracking differentiator, which is expressed as follows: ； in, The first derivative representing the motor position output by the tracking differentiator. This represents the estimated electrical angular velocity of the motor. The first derivative of the estimated electric angular velocity of the motor. This indicates the position of the motor that tracks the output of the differentiator; Indicates the gain of the tracking differentiator; This indicates the estimated electrical angle of the motor.

8. A sensorless control system for a permanent magnet synchronous motor, characterized in that, include: The model building module is configured to: establish a mathematical model of the back electromotive force of the permanent magnet synchronous motor and discretize it to obtain a discrete model of the stator current of the permanent magnet synchronous motor. The sliding surface design module is configured to: define the sliding surface based on the current estimation error and design a reference sliding surface based on the super-twisting algorithm; The observer design module is configured to: based on the mathematical model of the back EMF of the permanent magnet synchronous motor and the reference sliding surface, design a deadbeat prediction sliding mode observer based on the deadbeat prediction principle to realize the estimation of the back EMF of the motor; The position cost function construction module is configured to: establish the rotor position estimation error signal and cost function based on the estimated motor back electromotive force; The phase-locked loop (PLL) design module is configured to: design a finite position set PLL based on the golden section search based on the rotor position estimation cost function; obtain the estimated value of the motor rotor position by iteratively updating the rotor position search interval; and obtain the estimated value of the motor speed by using a tracking differentiator, thereby realizing sensorless control of the permanent magnet synchronous motor system.

9. A computer-readable storage medium having a program stored thereon, characterized in that, When the program is executed by the processor, it implements the steps in the sensorless control method for permanent magnet synchronous motors as described in any one of claims 1-7.

10. An electronic device comprising a memory, a processor, and a program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the steps in the sensorless control method for permanent magnet synchronous motors as described in any one of claims 1-7.

Images