Parameter Compensation for the Predictive Control System of a Permanent Magnet Synchronous Motor Based on Bacterial Foraging Optimization Algorithm

Abstract

The accurate identification of permanent magnet synchronous motor (PMSM) parameters is the foundation for high-performance driving in predictive control systems. The traditional PMSM multi-parameter identification method suffers from insufficient rank of the identification equation and is prone to getting stuck in local optimal solutions. This article combines the bacterial foraging optimization algorithm (BFOA) to establish a built-in PMSM predictive control parameter compensation model. Firstly, we analyzed the reasons why the distortion of PMSM motor parameters affects the actual speed and calculated the deviation of d-axis and q-axis currents caused by the distortion. Secondly, parameter compensation was applied to the prediction model, and BFOA was combined to optimize the compensation parameters. This algorithm does not use the traditional voltage equation as the fitness function but instead uses a brand-new set of four equations for parameter iteration optimization. The optimized compensation parameters can reduce current deviation and improve the robustness of the PMSM predictive control system. The proposed model can cover four kinds of motor distortion parameters, including stator resistance, D-axis inductance, Q-axis inductance, and permanent magnet flux linkage. Finally, the traditional PMSM predictive control model is compared with the predictive control model combined with BFOA. The simulation results show that the dynamic and static performance of the compensated system is improved when single or multiple parameters are distorted.