Parameter Identification for SPMSM With Deadbeat Predictive Current Control Using Online PSO
Chuanxun Xie, Shuo Zhang, Xueping Li, Ying Zhou, Yuelin Dong
Abstract
Deadbeat predictive current control (DPCC) can precisely predict current and calculate voltage vector of surface-mounted permanent magnet synchronous motor (SPMSM). But the prediction and calculation ability depends on accurate mathematical model, so the model parameters are significant for DPCC. Since motor parameters vary with operation condition, real-time parameter identification is essential for SPMSM. This paper proposes a parameter identification method using online particle swarm optimization (PSO) to obtain the important motor parameters (such as flux linkage and inductance). First, the principle of online PSO is illustrated compared with traditional PSO. Second, the fitness function is designed based on current prediction equations. Then a fitness modification method is presented to prevent particles from getting stuck in a stagnation state. Finally, a censor algorithm is proposed to ensure that the <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">gbest</i> of the swarm is always correct. Simulation and experiment show that the proposed method can identify the stator inductance and flux linkage simultaneously, and the particles can always converge to the actual value under different working conditions. After parameter identification, the harmonic content and current tracking error caused by parameter mismatch are effectively reduced.