Parameter Identification And Model-free Current Predictive Control Of PMSM

By 2020, in the Thirteenth Five-Year Planning, establishing a perfect system of electric vehicle power and industrial technology system and realizing the industrialization of all kinds of electric vehicles are put forward. It will promotes new energy vehicles strategic emerging industries into a rapid growth, and also is an important opportunity for our country to achieve corner overtaking of electric car industry. Permanent magnet synchronous motor (PMSM) becomes the main drive motor of electric vehicle drive system for it owns superior performance. It is becoming a hot research direction on the high performance control of PMSM. PMSM drive system not only requires wide speed range, but also requires torque output precision, however, the complicated operating conditions lead to motor parameter uncertainty, the traditional system based on PI control is difficult to achieve precise control, and moreover the inverter nonlinear will also result in output torque fluctuation, current distortion and so on. Therefore, this article in view of the PMSM inverter nonlinear compensation, parameter identification and control strategy for further study.Firstly, this article analyzes the nonlinear influence on the inverter output voltage, and summarizes the current main compensation schemes. On this basis, this paper proposes a forward compensation scheme based on MFAC, feasibility and effectiveness of the scheme is verified by simulation.Secondly, parameter identification is very important for PMSM control system design and real-time monitoring, but in actual operation, motor parameter will change a lot, therefore, the implementation of online identification for motor parameters is needed. After comparing existing parameter identification methods, the PMSM parameter online identification scheme is proposed based on the principle of differential algebraic. The scheme is effective to solve the problem that model owe rank, and the identification algorithm is simpler compared to other methods. Simulation results show the fast convergence speed and high identification accuracy.PMSM usually adopts the traditional PI control. However, PI control is designed based on linear model, it is only suit for a certain range, and engineering practice shows that integral saturation, overshoot and oscillation problems occur frequently under wide range speed control. Although PMSM model predictive control can solve some shortcomings of PI control, but it still depends on the motor parameters, therefore, this article proposed a predictive current control of PMSM based on model free theory to get rid of the dependence on motor parameters. The feasibility and effectiveness of the algorithm is validated through the system simulation.