永磁同步直线电机自组织概率型模糊神经网络控制  

Permanent magnet linear synchronous motor self-organizing probabilistic fuzzy neural network control

作　　者：张圳 王丽梅[1] Zhang Zhen;Wang Limei(School of Electrical Engineering,Shenyang University of Technology,Shenyang 110870)

机构地区：[1]沈阳工业大学电气工程学院,沈阳110870

出　　处：《电气技术》2020年第12期1-5,16,共6页Electrical Engineering

基　　金：国家自然科学基金资助项目(51875366)。

摘　　要：针对永磁同步直线电机易受参数变化和外部扰动等非线性因素影响的问题,本文采用了一种自组织概率型模糊神经网络控制方法来提高伺服系统的控制性能。概率型模糊神经网络(PFNN)可以有效地对系统中的不确性因素进行估计,且相比于神经网络(NN)有较强的鲁棒性,但是结构固定、隶属度低的节点对当前系统的控制力较差,难以调整系统动态过程中的稳态误差,因此本文在此基础上采用了一种自组织概率型模糊神经网络控制器(SOPFNN)。在对参数学习的情况下,同时采用了一种结构学习算法,来保证控制过程中每一节点都能发挥最大作用,进一步提高系统的跟踪性能。仿真结果表明,自组织概率型模糊神经网络控制不仅改善了系统的位置跟踪性能,还提高了系统的鲁棒性。In this paper,a self-organizing probabilistic fuzzy neural network control method is used to improve the control performance of permanent magnet linear synchronous motor(PMLSM).Probabilistic fuzzy neural network(PFNN)can effectively to estimate of uncertainty factors in the system of,and compared with NN has strong robustness,but the structure is fixed,low membership degree of node control to the current system is poorer,difficult to adjust the system steady-state error in the process of dynamic,so this article on the basis of the adopted a self-organized probabilistic fuzzy neural network controller(SOPFNN).In the case of parameter learning,a structure learning algorithm is adopted to ensure that each node can play the maximum role in the control process and further improve the tracking performance of the system.Simulation results show that the self-organizing probabilistic fuzzy neural network control not only improves the position tracking performance of the system,but also improves the robustness of the system.

关 键 词：永磁同步直线电机 自组织概率型模糊神经网络 跟踪误差 鲁棒性 

分 类 号：TP13[自动化与计算机技术—控制理论与控制工程] TM359.4[自动化与计算机技术—控制科学与工程]