基于双极性电压注入和PSO拟合的同步磁阻电机磁链自学习方法

Self-commissioning method of flux linkage for synchronous reluctance motor based on bipolar voltage injection and PSO fitting

同步磁阻电机(SynRM)存在明显的磁路饱和现象。如果在电机控制算法中使用固定电感值,无法实现理想的控制性能。针对该问题,提出一种磁链自学习方法。首先,详细分析了SynRM的交叉饱和现象,并采用恒定转速方法建立了SynRM的磁饱和模型;然后,设计了双极性电压注入的自学习给定方式,通过分区域拟合磁链曲线来后处理不均匀分布的原始磁链数据,并使用粒子群优化(PSO)算法求解拟合函数系数;最后,在实验平台对提出的磁链自学习策略进行验证。实验证明,相较于直接插值方法,分区域拟合磁链曲线的后处理方法对原始数据点的分布没有要求,因此可以显著降低磁链曲面的学习误差。

The synchronous reluctance motor(SynRM)exhibits significant flux saturation phenomenon.If a fixed inductance value is used in the motor control algorithm,it cannot achieve the desired control performance.To address this issue,a flux self-commissioning method is proposed.Firstly,the cross-saturation phenomenon of SynRM is analyzed in detail,and a flux saturation model of SynRM is established using the constant speed method.Then,a self-commissioning reference method with bipolar voltage injection is designed.The raw flux data with uneven distribution is post-processed by fitting flux curves in different regions,and the coefficients of the fitting function are obtained using the particle swarm optimization(PSO)algorithm.Finally,the proposed flux self-commissioning strategy is validated on an experimental platform.The results show that compared to direct interpolation methods,the post-processing method of fitting flux curves in different regions does not require a specific distribution of the raw data points,thus significantly reducing the error of the flux surface.