A Robust Model Predictive Current Closed-Loop Control With Parameter Estimation Strategy Based on Immune Chaotic Antipredator PSO for PMSM
Yang Zhang, Sicheng Li, Wenjing Yi, Yiping Yang, Kun Cao, Bing Luo
Abstract
Aiming at the issues of the current fluctuation and dependence of the control performance on model parameters accuracy in model predictive current control (MPCC) for a permanent magnet synchronous motor (PMSM), a robust model predictive current closed-loop control strategy with multiparameter estimation based on immune chaotic antipredator particle swarm optimization (RMPCC-ICAPSO) algorithm is proposed. The motor parameters are estimated by the ICAPSO algorithm and updated parameters are fed into the prediction model in real time. To begin with, the different single-phase voltage vectors are selected to synthesize the reference voltage vector, and the zero vector is split equally for redistribution. In addition, the immune chaotic theory is employed in updating the population of the PSO algorithm, learning factors as well as inertia weights. The ability of the algorithm to explore potentially better regions is enhanced, the problem of the local optimality is escaped, and the particle prematurity is suppressed. Moreover, an antipredator algorithm is introduced to the PSO, the worst solution position is excluded, the more uncharted regions are explored, and optimization of search accuracy is further enhanced. Finally, the RMPCC-ICAPSO strategy is experimentally demonstrated with a 3-kW motor. The correctness and effectiveness of the proposed RMPCC-ICAPSO strategy are proved by the experimental results.