基于群集蜘蛛算法的神经网络控制器研究  

A Study of Neural Network Controller Based on Social Spider Algorithm

作　　者：孟昕元 范峥 

机构地区：[1]河南工学院,新乡453002

出　　处：《微特电机》2017年第9期33-36,共4页Small & Special Electrical Machines

基　　金：河南省科技厅科技攻关重点项目(162102310167)

摘　　要：研究了群集蜘蛛算法对神经网络控制器参数进行优化的方法。使用神经网络控制器对他励直流电机的转速进行弱磁控制,以期提高闭环控制的性能。构建了MATLAB/Simulink下的仿真模型,分析了在不同的转速设定值下系统的稳态性能和动态响应。仿真结果显示,基于群集蜘蛛算法优化的神经网络控制器具有适应性强、动态过程过渡平稳快捷等优点。The optimization method on parameters of neural network controller using social spider algorithm was stud- ied. The neural network controller was used in controlling the speed of a separately excited DC motor under field weakening condition to improve the performance of closed loop control. By using MATLAB/Simulink, the system was simulated to ana- lyze the steady-state performance and dynamic response of the system under different speed setting values. The simulation resuhs show that the controller has the advantage of strong adaptability and the dynamic process of the system presents smoothly and fast.

关 键 词：他励直流电机 弱磁控制 神经网络控制器 群集蜘蛛算法 

分 类 号：TM33[电气工程—电机]