Research Of PMSM Parameters Identification Based On FPGA

Permanent magnet synchronous motor(PMSM) is an attractive candidates for high performance drive systems because of its high efficiency and suitability for wide speed ranges of constant power operation. The performance of PMSM servo system is highly influenced by variation of moment of inertia. In order to enhance the system performances, it is necessary to identify the moment of inertia and adjust the controller's parameters. Based on the traditional algorithm, the improved identification algorithms are proposed in the paper. The performances of the improved identification algorithms are verified in Matlab and System Generator.Firstly, based on the analysis of PMSM mathematical model, a novel fuzzy PI controller is proposed. Simulations based on Matlab/Simulink are carried out and the results show that when comparing to the traditional PI control strategy, systems with the proposed controller characterized by performances of good robustness to disturbances and parameter varieties, quick response and better stability, but it is still affected largely by the variable moment of inertia. Then moment of inertia identification research is done in this paper. The Recursive Least Squares(RLS) algorithm and model reference adaptive identification algorithm are discoursed. On the basis of RLS algorithm, a new identifier based on the Recursive Least Squares algorithm is proposed. A switch controller is introduced to the Forgetting Factor Recursive Least Squares(FFRLS) identifier to overcome the fluctuation. When the moment of the inertia changes, the switch controller initialize the identifier and realize the quick tracking of the parameter. The simulation and experiments show that this method is free from the fluctuation, the identifying is more quickly and high precision, and is benefit for the improvement of the performance of the system. To reduce the fluctuation, on the basis of mode1 reference adaptive identification algorithm, the identification result is feedbacked. In the process of the identification, the larger adaptive gain is chosen to identify the parameter quickly; when the result of identification is steady, the smaller adaptive gain is choosen. Then the improved identification algorithms is used to identify PMSM moment of inertia, the simulation and experiments show that the improved algorithms are free from fluctuation, the identifications are more quickly and high precision. Finally, the models of the improved algorithms are established with the System Generator, and give good validation for the proposed algorithms. These works built the foundation of the hardware realization based on the FPGA.