| The performance of an electric vehicle as a whole is greatly influenced by the motor,which is the heart of the drive system.The permanent magnet vernier motor,among other types of motors,can generate high torque at low speeds thanks to the magnetic gear effect and the technique of magnetic field modulation.This motor also has a straightforward design and high efficiency,and it has a promising future as an electric vehicle motor.Unfortunately,the motor vector control system that is currently in use has subpar robustness and control precision,a sluggish reaction time,and a heavy reliance on mechanical sensors,all of which contribute to the system’s expensive cost,poor environmental adaptability,and lack of anti-interference capabilities.Model predictive control and a model-referenced adaptive observer are investigated for permanent magnet vernier motors to address the shortcomings of the aforementioned control systems.A model predictive control system incorporating the model-referenced adaptive observer is then proposed,and the system’s viability and superiority are confirmed in the ensuing simulation analysis.First,the structure and performance of the permanent magnet vernier motor were analyzed.Using electromagnetic analysis software,a two-dimensional simulation model of the permanent magnet vernier motor is made with the goal of analyzing its electromagnetic properties.A mathematical model was also derived to serve as the foundation for the execution of the control algorithm below.Second,by extending the inverter voltage vector,creating dual-vector,and improving its cost function,a dual-vector model predictive controller based on virtual vector was created to address the drawbacks of the PI controller’s subpar control performance in the vector control system.Designing a model-referenced adaptive observer for position sensorless control enables the detection of rotor information.In order to create a model predictive control system with a model reference adaptive observer,the vector control system’s initial PI controller and mechanical sensor are respectively replaced by the model predictive controller and the model reference adaptive observer.Finally,by comparing the deviation of the actual values of speed and rotor position with the estimated values and the overall operation effect,a simulation model of the control system was built in the MATLAB/Simulink software to test the viability of the model predictive control system with a model-referenced adaptive observer.Then,the performance of the system was examined in terms of response time,stability,and torque pulsation to compare the speed,torque,and current waveforms with the traditional vector control system under various speed and load conditions,respectively.The results show that the developed model predictive control system with a model reference adaptive observer is capable of monitoring rotor information under a variety of working conditions with a small observation error and seamless system operation.With a response speed that is more than 35% faster,a stable torque output,a torque pulsation coefficient that is roughly 17% lower,a high sinusoidal three-phase current waveform,and better control accuracy and stability,the new vector control system outperforms the conventional vector control system in terms of speed regulation.Based on the above simulation results,it can be concluded that the new control system designed in this paper has better control performance compared with the traditional vector control system and can better realize the control of the permanent magnet vernier motor. |
