基于LS-SVM的异步电机解耦控制方法

最小二乘支持向量机（least square support vector machines，LS—SVM）是一种基于结构风险最小化的机器学习方法，能逼近任意非线性函数，具有无局部极小等优点。利用其构造异步电机这一多变量、强耦合系统的逆系统，将转子磁链与转速解耦成两个独立的伪线性子系统，并设计了相应的闭环控制器。仿真结果表明，该逆系统能有效地实现电机转速与磁链的动态解耦。通过合理地选择控制器参数，整个控制系统对负载扰动有较强的鲁棒性。

Decoupling Control of Asynchronous Motor Based on Least Square Support Vector Machines

Least square support vector machines is a type of machine learning method with structure risk minimization.It can approximate any nonlinear function with the advantage of non-local minimization.It is applied to construct the inverse system of asynchronous motor which is a system with multiple variable and strong coupling,and then the rotor flux and speed is decoupled into two independent pseudo-linear subsystems.The relative close-loop controller is designed subsequently.The results show that the inverse sy...