摘要
无刷直流电机(BLDCM)的动力学特性是一个高阶、非线性、强耦合的系统,针对传统PI控制的滞后性和动态响应性能较差等特点,本文提出一种基于动态递归模糊神经网络PI控制的无刷直流电机调速系统速度控制器的实施方案,利用蚁群算法优化递归模糊神经网络的隶属度函数参数和网络权值系数,从而提高系统的动态响应性能。仿真结果表明,该方法响应快,具有较强的抗干扰性和鲁棒性,动、静态特性均优于传统PI控制。
The dynamic characteristics of brushless DC motor (BLDCM) is a high-order, nonlinear and strong coupling system. For the characteristics of the lagging and the dynamic poor response, this paper puts forward a way of dynamic PI control to the speed system of BLDCM. Ant colony algorithm is used to optimize the membership function parameters and the network coefficient of recurrent fuzzy neural network to improve the dynamic performance. Simulation results show that the way has fast response, strong coupling . The dynamic and static characteristics are superior than the traditional PI control.
基金
江苏广播电视大学学术带头人基金资助工程阶段性成果