基于 RBF 神经网络整定 PID 的PMLSM 控制研究.PDFVIP

安徽农业大学学报, 2012, 39(1): 154-158 Journal of Anhui Agricultural University [DOI]CNKI：34-1162/S1123.001 网络出版时间：2011-12-30 11:23:17 [URL]/kcms/detail/34.1162.S1123.001.html 基于RBF神经网络整定 PID 的PMLSM 控制研究 1 1 2 曹文霞 ，汪永华 ，钱多德 (1. 安徽水利水电职业技术学院机械工程系，合肥 231603 ；2. 安徽江淮汽车股份有限公司技术中心，合肥 230601) 摘 要：介绍了直线永磁同步电机（permanent magnet linear synchronous motor ，PMLSM ）的数学模型。由于 直线永磁同步电机具有非线性、耦合性和负载扰动等特性，常规 PID 控制器不能达到理想的控制效果，而径向基 函数（radial basis function ，RBF ）神经网络具有学习速度快、无局部极小和良好的逼近能力，提出了将RBF 网络 与增量式 PID 控制结合在一起应用的方法。matlab 仿真结果表明：该控制系统具有良好的逼近能力、较强的适应 性和抗干扰能力。 关键词：直线永磁同步电机；RBF 神经网络；PID 控制；matlab 仿真 中图分类号：TM359.4 文献标识码：A 文章编号：1672−352X (2012)01−0154−05 RBF neural network- based setting PID in controlling PMLSM 1 1 2 CAO Wen-xia , WANG Yong-hua , QIAN Duo-de (1. Department of Mechanical Engineering, Anhui Technical College of Water Resources and Hydroelectric Power, Hefei 231603; 2. RD center, Anhui Jianghuai Automobile Co.,Ltd, Hefei 230601) Abstract: In this paper, a math model of permanent magnet linear synchronous motor (PMLSM) was intro- duced, Because PMLSM has characteristics of nonlinear, coupling and load disturbance, common PID controller can’t reach an ideal control effect, while radial basis function (RBF) neural network has the ability of studying quickly, without local minimums, and better approaching, so the RBF was combined with PID in controlling PMLSM. Matlab simulation results demonstrated that the control system has better approaching ability, better ad- aptation and preventing disturbance ability. Key words: PMLSM; RBF neural network; PID control; matlab simulation PID 控制由于算法简单，鲁棒性好及可靠性高，