The Predictive Current Control Of Pmsm Based On Disturbance Observation

In AC drive servo system, the advantages of permanent magnet synch ronous motor(PMSM) make it become the first choice of industrial control occasions with high precision and research focus in recent years. Since the current loop determines the transient and steady-state performance of AC servo system, PMSM predictive current control owing an excellent transient performance has become a research hotspot. But its disadvantages are obvious at the same time, including dependence of a precisive motor model and static current error caused by open-loop control.Firstly, this dissertation briefly discribes the principle of PWM predictive current control(PPC). Through theoretical analysis, the stability condition of the PPC is deduced and the reasen of static current error is proved to be caused by the motor parameter errors. Several simulations are done to show that current divergence and static current error exist in the PPC under different situations including the resistance error, inductance error and magnet flux linkage error in the system separately, which verify the previous theoretical analysis.In terms of the drawbacks in PPC, this dissertation constructs two different state observers to estimate the parameter errors and parameter disturbance in t he servo system. Firstly, Luenberger Observer, a widely used observer, is built and its stability is deduced by the pole assignment. Then an Extended State Observer is constructed and also the stability condition is derived. The merits and faults are compared between these two disturbance observers. Proposed disturbance obserers make the robust predictive current control method based on the dist urbance compensation possible.According to the different defination of system disturbance in two observers, two different improved predicitve current control methods are proposed, i.e the predicitve current control based on the Luenberger disturbance observation and Extended State Observer separately. This dissertation ananlyses the stability of the proposed method and proves that the robustness has been improved. The simulation results illuminate that the two methods could observe the system disturbance and make a real-time compensation of disturbance. In this case, static current error is avoided amd the current divergence is inhibited. At last, the characteristics of two proposed methods are listed and concrete suggestions are offered for the application.Finally, PMSM experimental platform is designed. The experiment results prove the effectiveness of the proposed algorithm.