改进综合学习粒子群算法的PMSM参数辨识

摘要

为了解决永磁同步电机( permanent magnet synchronous motor, PMSM)多参数辨识问题,提出一种改进综合学习粒子群优化算法。针对综合学习粒子群算法后期搜索效率低的缺陷,所提算法引入反映粒子状态的增长率算子,通过该算子动态调整综合学习粒子群算法的关键参数,并根据增长率算子判断种群中粒子所处状态,对处于停滞状态的粒子实施高斯扰动,使粒子能在解空间中进行有效搜索。将所提改进算法应用于永磁同步电机多参数辨识,该方法仅需采样电机的定子电流、电压和转速信号。实验结果表明,改进综合学习粒子群优化方法能够准确地辨识PMSM的定子电阻、d轴和q轴电感和永磁体磁链等参数。%An improved compressive particle swarm optimization algorithm( ICLPSO) for identification of the permanent magnet synchronous motor( PMSM) parameters was proposed. Aiming at the drawback of CLPSO, in the proposed algorithm the growth operator was introduced, with which changed the value of acceleration coefficient dynamically and judged the statue of particles. Using Gaussian disturbance, the particles in stagnant state effectively search in the solution space. The ICLPSO was employed to identify the parameters permanent magnet synchronous motor( PMSM) . The proposed method needs only to sam-ple and to store the data of the motor stator current, voltage and rotor speed. Experimental results on PMSM electrical parameter identification show that the proposed method accurately identifies the stator re-sistance, d-axis inductance, q-axis inductance and the permanent magnet flux.