基于神经网络逆的永磁同步电机自适应解耦研究

摘要

An decoupling control method based on neural network inverse system and model reference adaptive control for permanent magnet synchronous motor (PMSM) , which was of multivariable nonlinear characteristics, had been proposed. Consequently, two pseudo-linear subsystems were completed by combining the neural network inverse system with the PMSM. PMSM was decoupled into a first-order linear flux subsystem and a second-order linear speed subsystem. Stator resistance changes with temperature, and then affects the two subsystems. Error of stator resistance was used to adjust by model reference adaptive control and reduce the effect to the system. Simulation results showed this method could realize dynamic decoupling control between flus and speed, and guarantee the stability of the system.%针对永磁同步电机多变量、非线性、强耦合等特性,提出一种基于神经网络逆的模型参考自适应解耦控制方法.神经网络逆系统与原系统复合成两个伪线性子系统,一个一阶磁链子系统和一个二阶转速子系统.定子电阻随温度变化而变化,进而影响磁链子系统和转速子系统的解耦控制.模型参考自适应控制方法可以通过误差调节来减小定子电阻变化对系统的影响,仿真试验结果表明该控制策略能够在定子电阻发生变化的情况下实现转速与定子磁链之间的动态解耦,并能保证系统的稳定性.