Variable parameter model reference adaptive system positionless switching control method

Abstract

This invention discloses a sensorless switching control method based on a variable parameter model reference adaptive system, relating to the field of permanent magnet synchronous motor (PMSM) control technology. Traditional switching control methods for PMSMs suffer from problems such as large speed fluctuations, severe voltage and current overshoot, high computational resource consumption, and complex engineering implementation when the position sensor fails. This invention employs a sensorless algorithm based on a variable parameter model reference adaptive system and designs a staged switching control strategy. By optimizing key parameters such as the adaptive law PI parameters, demodulation value weights, and switching delay time, a smooth switching between sensor-on and sensorless control is achieved. This invention significantly reduces speed fluctuations, voltage and current overshoot, and improves motor operation stability and reliability while requiring relatively little processor memory and computational resources. It is applicable to PMSM drive systems in civilian and military fields.

Description

Technical Field

[0001] This invention relates to the field of permanent magnet synchronous motor (PMSM) control technology, specifically to a positionless segmented switching control method based on a variable parameter model reference adaptive system (MRAS), which is applicable to permanent magnet synchronous motor drive systems in civilian, military and other fields that require both high control accuracy and high reliability. Background Technology

[0002] Permanent magnet synchronous motors (PMSMs) have significant advantages in high power density and high energy conversion efficiency, and are widely used in civilian and military fields such as industrial production, new energy vehicles, and aerospace. The vector control strategy of PMSMs relies on rotor position and speed information, which is typically obtained through demodulation by position sensors (such as resolvers) or estimated using sensorless algorithms.

[0003] Position sensor demodulation offers high control accuracy, but its system structure is complex and costly. Sensorless algorithms are low-cost and highly reliable, but their control accuracy is lower and they are sensitive to motor parameters. When a position sensor fails, the control mode must be smoothly switched from sensor-based control to sensorless control; otherwise, speed and angle information will be lost, potentially leading to serious consequences such as motor runaway.

[0004] Existing switching control methods have many shortcomings: direct switching methods can lead to excessive speed fluctuations, voltage overshoot, and current overshoot, which may exceed the protection values ​​of the devices; complex sensorless algorithms based on extended Kalman filters, high-order sliding mode observers, etc., can improve estimation accuracy, but involve a large number of matrix operations, which consume a lot of processor memory and computing resources, making them unsuitable for the engineering implementation of multi-mode converters; improved switching strategies based on weighted functions may have poor universality or rely on complex intelligent algorithms, which also suffer from high computational resource consumption.

[0005] Therefore, developing a method that can achieve smooth switching between control with and without position sensors while consuming relatively few computing resources is of great significance for improving the reliability and stability of permanent magnet synchronous motor systems. Summary of the Invention

[0006] To address the problems of large speed fluctuations, high computational resource consumption, and complex engineering implementation in existing technologies, this invention provides a position-free segmented switching control method for permanent magnet synchronous motors based on variable parameter MRAS, which enables smooth switching when the position sensor fails, reduces system fluctuations, and minimizes computational resource consumption.

[0007] To achieve the above objectives, the present invention adopts the following technical solution: a position-free segmented switching control method for a permanent magnet synchronous motor based on a variable parameter model reference adaptive system, comprising the following steps: Step S1: Select a sensorless control algorithm based on Model Reference Adaptive System (MRAS), and construct a reference model and an adjustable model ( Figure 2 As shown in the figure, an adaptive law is designed based on Popov's superstability theory to obtain the estimated motor speed. and rotor position estimate ; Step S2, configure the adaptive law PI parameters of the variable parameter MRAS according to the different speed domains of the permanent magnet synchronous motor. This enables high-precision estimation under different working conditions. Step S3: When the system is running normally, a weighted fusion strategy is adopted. The feedback values ​​of motor closed-loop control speed, rotor position and electric angular velocity are obtained by weighted fusion of position sensor demodulation value and MRAS estimation value to improve MRAS estimation accuracy. Step S4: After detecting a position sensor malfunction, initiate the first stage of switching ( Figure 3 As shown): The reference and feedback values ​​of the outer speed loop are immediately switched to MRAS estimated speed. The rotor position step transformation used for the inner current loop coordinate transformation is switched to MRAS estimated position. The electric angular velocity of the motor in the feedforward section remains unchanged from its final value before the fault. Step S5: After the first stage switching delay time t, the second stage switching is initiated: the outer loop reference value of the speed is estimated by MRAS, and the speed ramp rises or falls to the target speed. Restore closed-loop speed regulation and complete a smooth transition.

[0008] Furthermore, the speed estimation formula for the adaptive law mentioned in step 1 is: The rotor position estimation formula is: In the formula: To estimate the electric angular velocity, , For the adaptive law PI parameters, , These are the actual currents along the d and q axes. , Estimate the current for the d and q axes. For motor magnetic flux, For stator inductance, To estimate the initial value of the electric angular velocity, This is the estimated rotor position.

[0009] Furthermore, the proportional coefficient of the adaptive law described in step 2 and integral coefficient The motor speed is adjusted in segments, specifically including: Determined based on the system's open-loop transfer function ratio: in, For the stator resistance of the motor, , , These are the steady-state operating conditions. , Shaft current and electric angular velocity values; Combined with Bode Figure 3 Frequency band theory and closed-loop root locus analysis were used to select different speed ranges. and The value ensures that the system has optimal dynamic performance and steady-state accuracy within the corresponding speed range.

[0010] Furthermore, the weighted fusion formula described in step 3 is: Speed ​​fusion: Location fusion: Electric angular velocity fusion: In the formula: , , These are the rotational speed, rotor position, and electrical angular velocity demodulated from the position sensor, respectively. , , These represent the rotational speed, rotor position, and electrical angular velocity estimated by MRAS, respectively. The demodulated value weight.

[0011] Furthermore, the weight of the demodulated value The methods for determining this include: Establish the open-loop transfer function for the weighted outer-loop vector control of rotational speed: in, This is the outer ring proportional coefficient for rotational speed. For extreme logarithms, , The outer ring integral coefficient for rotational speed. For rotational inertia, It is an equivalent small time constant; Establish the transfer function of the estimated speed error with respect to the speed reference value, and analyze different step responses. The system settling time and estimation accuracy under various values ​​are comprehensively selected to achieve the optimal balance between system dynamic performance and estimation accuracy. value.

[0012] Furthermore, the position sensor fault detection described in step 4 is achieved by comparing the continuity and rationality of the sensor demodulation signal. When the demodulation signal exceeds a preset threshold or is disconnected, it is determined to be a fault.

[0013] Furthermore, the delay time mentioned in step 5 The methods for determining this include: Establish the current inner loop control block diagram during the first stage switching ( Figure 4 (as shown), derivation right The closed-loop transfer function is obtained from the unit step response. Adjustment time ; based on Based on the step response characteristics, it is approximated that the actual motor speed undergoes a step change. A transfer function of the estimated speed error to the actual speed error is established, and the settling time for the estimated speed to stabilize is obtained through the unit step response. ; Taking into account different operating conditions of the motor and allowing for margin, the delay time is determined. .

[0014] Compared with the prior art, the present invention has the following beneficial effects: 1. Excellent switching smoothness: Speed ​​fluctuation, voltage overshoot, and current overshoot are all far superior to the direct switching method; 2. Low computational resource consumption: It only involves linear changes in four parameters, without the need for complex matrix operations or intelligent algorithms, and is suitable for the limited resources of embedded controllers; 3. Simple engineering implementation: The MRAS algorithm has a simple structure, the initial position can be arbitrarily chosen, and the key parameters have been optimized through theoretical calculations, requiring no complex debugging; 4. High reliability: During normal operation, the system improves the MRAS estimation accuracy through weighted fusion, and quickly switches in case of failure to ensure continuous and stable operation of the motor. Attached Figure Description

[0015] Figure 1 Functional diagram of converter + permanent magnet synchronous motor system; Figure 2 Schematic diagram of the basic structure of MRAS; Figure 3 Flowchart of the switching process after a fault occurs; Figure 4 PMSM current inner loop control block diagram; Figure 5 Simulation waveforms of the switching strategy under typical operating conditions; Figure 6 Experimental waveform diagram of the switching process under typical operating conditions. Detailed Implementation

[0016] The present invention will now be described in further detail with reference to the accompanying drawings and specific embodiments.

[0017] 1. Experimental Platform The experimental platform consists of a multi-port converter, a three-phase permanent magnet synchronous motor, a lithium battery energy storage system, a position sensor (rotary transformer), and a controller. Figure 1 (As shown), the specific parameters are as follows: 1) Multi-port converter: Input 380V AC power to realize rectification, inversion, charging and drive functions; 2) Controller: An embedded processor configured with the switching control algorithm described in this invention; 3) Permanent magnet synchronous motor parameters: number of pole pairs stator resistance Direct-axis inductor quadrature axis inductance Magnetic Link ; 4) The configuration rules for the adaptive law PI parameter of MRAS are as follows: When the given rotational speed is 200 r / min to 800 r / min, , ; When the given rotational speed is 800 r / min to 1200 r / min, , ; When the given rotational speed is greater than 1200 r / min, , .

[0018] 5) Weights for demodulated values ; 6) Delay time .

[0019] 2. Control Algorithm Implementation Steps Step 1: System Initialization After the controller is powered on, initialize the motor parameters, MRAS adaptive law PI parameters (preset according to the speed domain), and demodulation value weights. Switching delay time The position sensor and MRAS algorithm are started to work in parallel. The speed, position and electric angular velocity feedback values ​​are obtained through weighted fusion, and vector control with position sensor is performed.

[0020] Step 2: Fault Detection The demodulated signal of the position sensor is monitored in real time. When the demodulated signal exceeds the preset threshold (speed fluctuation ±5%, position signal interruption), it is determined to be a position sensor fault and the switching process is triggered.

[0021] Step 3: Phased Execution Switching Phase 1 (0~0.2s):  Rotational speed outer ring: Reference value Both the feedback value and the estimated rotational speed are updated to MRAS values. When the input to the outer loop PI controller is 0, the output remains unchanged. The current reference value is stable;  Inner current loop: The rotor position used for coordinate transformation is determined by... Step switch to ;  Feedforward section: The electrical angular velocity remains unchanged from its final value before the fault.

[0022] Second stage (>0.2s):  The outer ring reference value of the rotational speed is determined by The engine speed is ramped up / down at a rate of 0.5 r / (min·ms) to the target speed of 1500 r / min; The controller adjusts according to the speed error. The current is used to stabilize the motor speed at the target speed, thus completing the switching.

[0023] 3. Simulation and Experimental Results Simulation results A simulation model was built in Simulink, with a target rotational speed of 1500 r / min and a load torque of 130 N·m. A position sensor failure was simulated at the 16th second. The simulation results are as follows. Figure 5 As shown: • Speed ​​fluctuation: Maximum 1506 r / min, fluctuation range 0.4%; Bus voltage: Maximum 565V, overshoot 0.9%; Three-phase current: peak 75A, overshoot 15%.

[0024] Experimental results The verification was performed on the constructed experimental platform, with experimental conditions consistent with the simulation. The results are as follows: Figure 6 As shown:  Speed ​​fluctuation: Maximum 1518 r / min, fluctuation range 1.2%; Bus voltage: Maximum 550V, overshoot 0.9%; Three-phase current: peak 70A, overshoot 15%.

[0025] Results Analysis Simulation and experimental results show that the control method described in this invention achieves smooth switching after a position sensor failure, and the fluctuations in speed, voltage and current are all within the safe range of the device. Moreover, the computational resource consumption is only 15% of that of the extended Kalman filter algorithm, which meets the multi-operating mode requirements of multi-port converters.

[0026] The slight differences between the simulation and experimental results stem from fluctuations in the actual load torque (caused by the characteristics of the hydraulic pump) and nonlinearity of the motor parameters, but their trends are consistent, verifying the feasibility and effectiveness of the method of the present invention.

[0027] Matters not covered in this invention may be adjusted based on the conventional understanding of those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this invention should be included within the protection scope of this invention.

Claims

1. A position-free segmented switching control method based on a variable parameter model reference adaptive system, applied to a three-phase permanent magnet synchronous motor system, characterized in that, Includes the following steps: Step S1: Select a sensorless control algorithm based on Model Reference Adaptive System (MRAS), construct a reference model and an adjustable model, design an adaptive law based on Popov hyperstability theory, and obtain the estimated motor speed. and rotor position estimate ; Step S2, configure the adaptive law PI parameters of the variable parameter MRAS according to the different speed domains of the permanent magnet synchronous motor. This enables high-precision estimation under different working conditions. Step S3: When the system is running normally, a weighted fusion strategy is adopted. The feedback values ​​of motor closed-loop control speed, rotor position and electric angular velocity are obtained by weighted fusion of position sensor demodulation value and MRAS estimation value to improve MRAS estimation accuracy. Step S4: After detecting a position sensor fault, initiate the first stage of switching: the reference and feedback values ​​of the outer speed loop are immediately switched to MRAS estimated speed. The rotor position step transformation used for the inner current loop coordinate transformation is switched to MRAS estimated position. The electric angular velocity of the motor in the feedforward section remains unchanged from its final value before the fault. Step S5, first stage switch duration delay time t Then, the second stage of switching is initiated: the outer loop reference value of the speed is estimated by MRAS, and the speed ramp rises or falls to the target speed. Restore closed-loop speed regulation and complete a smooth transition.

2. The control method according to claim 1, characterized in that, The speed estimation formula for the adaptive law mentioned in step 1 is: The rotor position estimation formula is: In the formula: To estimate the electric angular velocity, , For the adaptive law PI parameters, , These are the actual currents along the d and q axes. , Estimate the current for the d and q axes. For motor magnetic flux, For stator inductance, To estimate the initial value of the electric angular velocity, This is the estimated rotor position.

3. The control method according to claim 1, characterized in that, The proportional coefficient of the adaptive law in step 2 and integral coefficient The motor speed is adjusted in segments, specifically including: Determined based on the system's open-loop transfer function ratio: in, For the stator resistance of the motor, , , These are the steady-state operating conditions. , Shaft current and electric angular velocity values; Combining Bode plot three-band theory and closed-loop root locus analysis, different frequency ranges were selected respectively. and The value ensures that the system has optimal dynamic performance and steady-state accuracy within the corresponding speed range.

4. The control method according to claim 1, characterized in that, The weighted fusion formula mentioned in step 3 is: Speed ​​fusion: Location fusion: Electric angular velocity fusion: In the formula: , , These are the rotational speed, rotor position, and electrical angular velocity demodulated from the position sensor, respectively. , , These represent the rotational speed, rotor position, and electrical angular velocity estimated by MRAS, respectively. The demodulated value weight.

5. The weighted fusion formula according to claim 4, characterized in that, The weight of the demodulated value The methods for determining this include: Establish the open-loop transfer function for the weighted outer-loop vector control of rotational speed: in, This is the outer ring proportional coefficient for rotational speed. For extreme logarithms, , The outer ring integral coefficient for rotational speed. For rotational inertia, It is an equivalent small time constant; Establish the transfer function of the estimated speed error with respect to the speed reference value, and analyze different step responses. The system settling time and estimation accuracy under various values ​​are comprehensively selected to achieve the optimal balance between system dynamic performance and estimation accuracy. value.

6. The control method according to claim 1, characterized in that, The position sensor fault detection described in step 4 is achieved by comparing the continuity and rationality of the sensor demodulation signal. When the demodulation signal exceeds the preset threshold or is disconnected, it is determined to be a fault.

7. The control method according to claim 1, characterized in that, The delay time mentioned in step 5 The methods for determining this include: Establish the current inner loop control block diagram during the first stage switching and derive the... right The closed-loop transfer function is obtained from the unit step response. Adjustment time ; based on Based on the step response characteristics, it is approximated that the actual motor speed undergoes a step change. A transfer function of the estimated speed error to the actual speed error is established, and the settling time for the estimated speed to stabilize is obtained through the unit step response. ; Taking into account different operating conditions of the motor and allowing for margin, the delay time is determined. .

8. A permanent magnet synchronous motor drive system applying the control method described in any one of claims 1-7, characterized in that, It includes a multi-port converter, a permanent magnet synchronous motor, an energy storage system, a position sensor, and a controller. The controller is equipped with the switching control algorithm described in any one of claims 1-7 to realize rectification, inversion, motor drive, power generation, and fault switching functions.

Images