基于支持向量机广义逆的永磁同步电机模型参考自适应控制

Model reference adaptive control of PMSM based on support vector machines generalized inverse

针对永磁同步电机多变量、非线性、强耦合的特点,提出了基于支持向量机广义逆的模型参考自适应控制策略.在分析永磁同步电机广义可逆性的基础上,利用支持向量机回归理论辨识出系统逆模型并引入线性环节构造其广义逆系统,进而与原系统串联构成广义伪线性复合系统,实现被控对象的线性化重构,简化系统结构和外围控制器的设计难度.结合模型参考自适应控制策略,将支持向量机广义逆作为主控制器,设计出新型可调增益的闭环自适应控制器,提高了系统的鲁棒性.仿真与实验表明,该方法不依赖于被控对象的数学模型,系统抗干扰能力强,减小超调的同时提高了响应速度,实现了转速的高精度跟踪控制.

To deal with the problems of multivariable,nonlinearity and strong coupling of the permanent magnet synchronous motor ( PMSM ) ,a strategy of model reference adaptive control ( MRAC) based on support vector machines ( SVM) generalized inverse,termed as MRAC-SVM generalized inverse,is proposed. After analyzing the generalized inverse of PMSM,SVM regression theory is used to identify the generalized inverse model by adopting linear link. Then the nonlinear system is transformed into a pseudo-linear system by connecting generalized inverse and original unit to simplify external controller design. Finally,the SVM generalized inverse is introduced into MRAC,and an adjustable gain of the closed-loop adaptive control system is designed,in which the main controller is an SVM generalized inverse unit. Both simulation and experimental results confirm that the proposed method incorporates the merits of model-free learning,high-precision speed tracking feature,strong anti-interference capability,low overshoot and high response rate.