基于扰动补偿的电机模拟器电流控制策略

Mechanical load dynamometer systems are commonly used to test and develop motor drive systems. However, traditional mechanical load testbeds for electric motors have several drawbacks, such as large volume, high cost, and difficulties-to-implementing fault testing, which fail to meet the requirements of efficient electric drive system testing. An electric motor emulator (EME) based on power hardware-in-the-loop (PHIL) can simulate the current and voltage characteristics of real motor ports using power electronic devices and control algorithms, making it a novel solution for testing the performance of motor controllers. The current control strategies are key to affecting the accuracy of EME. However, the low-pass filtering nature of conventional PI control limits the dynamic control effect of EME, and traditional open-loop control strategies are susceptible to parameter perturbations. Therefore, an open-loop current control strategy based on sliding mode perturbing observers is proposed. First, the PI control defect and low-pass filtering properties are analyzed by deriving the transfer function and using the Bode diagram. The theoretical analysis shows that PI control cannot simulate the amplitude-frequency properties of the target motor in the full bandwidth. Second, the open loop control voltage equation without differential calculation is derived based on the mathematical model of the target motor and the interface circuit model. Moreover, the perturbation effects of the resistance and inductance parameters on the control policy are analyzed. The inductance perturbation mainly affects the high-frequency features, which are more distorted as the degree of perturbation increases, leading to a decrease in the accuracy of the simulation of the dynamic response. Resistance perturbations mainly affect the low-frequency features. As the degree of mismatch increases, the distortion of the low-frequency features and the steady-state error become larger. Finally, robust sliding-mode perturbation observers are designed for perturbation observation and feed-forward compensation. Simulation results show that under 35% resistance disturbance and 25% resistance disturbance, the traditional open-loop control exhibits phase current simulation error as high as 3 A in the dynamic response, and the steady-state error is about 2 A. The PI control has the largest phase current simulation error during the acceleration stage, about 2 A, and the steady-state error is about 0.8 A. The error of the proposed control strategy is within 0.7 A in both dynamic and steady-state responses. Experimental results show that traditional open-loop control experiences amplitude errors and phase perturbations after parameter perturbations. The proposed control strategy has higher precision and accuracy than the motor simulator system based on the traditional PI control and open-loop control strategy, regardless of the transient or steady-state condition when there is parameter disturbance. Under 25% inductance disturbance and 35% resistance disturbance, the maximum phase current simulation error is reduced by 60% compared to the traditional open-loop control and 10% compared to the PI control. Meanwhile, the harmonic content of phase current is reduced by 2.48% compared to PI control.