基于Markov链随机脉宽调制的永磁同步电机高频边带谐波与声振响应抑制

Investigation into Markov-chain random modulation for suppression high-frequency sideband vibro-acoustics in permanent magnet synchronous motor

基于Markov链算法的随机脉宽调制(RPWM)技术,以电动汽车驱动用永磁同步电机(PMSM)为研究对象,对高频边带电流谐波及声振响应的抑制效果进行研究。首先,基于dSPACE半实物仿真系统和12极/10槽永磁同步样机,对空间矢量脉宽调制(SVPWM)所引入的边带电流谐波与径向电磁力进行解析与频谱阶次分布特征识别。其次,对常规RPWM的随机数性能进行分析,提出多状态Markov链随机数产生原理与实现方法,并结合粒子群(PSO)算法对转移概率和随机增益两种随机参数进行快速寻优。最后,进行稳态工况与不同转速工况的实验验证,相比于RPWM结果表明,经过多状态Markov链优化后所生成的随机数分布更加均匀且对称,边带电流谐波与声振响应均呈现出更好的抑制效果。

An investigation associated with the sideband vibro-acoustics suppression was presented with multi-state Markov-chain random pulse-width modulation(RPWM)in permanent magnet synchronous motor(PMSM),together with the optimization method for the random parameters.Firstly,the feature identifications of sideband current harmonics and the radial electromagnetic force,including spectral and order distributions introduced by the space vector pulse-width modulation(SVPWM),were undertaken on a 12-pole/10-slot prototype PMSM and its dSPACE hardware-in-the-loop control system.Then,the principle and implementation of different states Markov-chain were presented,while the random performances of conventional RPWM were conducted.The key parameters,i.e.transition probability and random gain,were optimized by using the particle swarm optimization(PSO)method.At last,the sideband current harmonics and vibro-acoustic results were conducted and evaluated under steady state and different speed conditions.Since the random numbers generated by the m Markov chain are evenly and continuously,the suppression effects are better than those in conventional RPWM.