直接转矩控制系统中自学习控制器的研究

Research of Self-learning Controller in Direct Torque Control System

交流电动机是一非线性、多变量、强耦合的控制对象,其数学模型较难精确建立,为了克服常规模糊控制器隶属函数不变的缺陷,本文设计了一种具有自学习功能的控制器。在模糊原理的基础上,本控制器借助自学习机制,实现非线性PID控制器参数的实时优化调整,直到系统性能达到预期性能要求为止。把该控制器应用于直接转矩控制系统,通过自学习模糊推理,获得速度环的自适应控制。同常规PID控制器相比,它可以在一定程度上提高系统的动态控制性能。

AC motor is a control object that is nonlinear, multivariable and strong coupling, so it is difficult to build the exact model for AC motor. To overcome the shortcoming of invariable membership function in traditional fuzzy controller, we design a controller with the ability of self-learning, which can realize the real-time optimization of parameters of nonlinear PID controller based on fuzzy theory until the performance of the system conforms to requirement. When the controller is applied to DTC system, self-adaptive control of speed loop can be accomplished by self-learning fuzzy reasoning. Compared with traditional PID controller, it can promote the dynamic performance of the system at some degree.