基于模糊神经网络补偿控制器的H型平台同步控制  

Synchronous Control of Type H Table Based on Fuzzy Neural Network Compensation Controller

作　　者：张佳媛[1] 王丽梅[1] 

机构地区：[1]沈阳工业大学电气工程学院,沈阳110870

出　　处：《电气工程学报》2015年第9期54-61,共8页Journal of Electrical Engineering

基　　金：辽宁省高等学校创新团队项目(201334068);辽宁省高等学校优秀人才支持计划资助项目(LR2013006)

摘　　要：直线电机驱动的H型数控平台系统在加工零件时,负载扰动、外部干扰和两电机安装的差异与机械耦合会影响单轴的跟踪精度且会产生同步误差。针对此问题,本文首先用拉格朗日方法给H型平台建模,然后提出一种改进的非奇异终端滑模控制(NTSMC)来进行位置控制器的设计,在不失滑模控制鲁棒性的情况下,有效地削弱了该控制所产生的抖振问题,提高了单轴的跟踪精度。在两轴间采用Sugeno型模糊神经网络(SFNN)补偿控制器来动态补偿H型平台的同步误差。通过模糊神经网络以任意精度逼近非线性系统的能力使同步误差在有限时间内趋近于零,以满足H型平台数控系统的高精度加工要求。仿真结果表明,所设计的控制系统能够有效提高系统的同步控制精度和鲁棒性。Load disturbance, external disturbances, difference in the installation of the two motors and the mechanical coupling affect single axis tracking precision and produce synchronous error for linear motors-drive type H table in the processing parts. Aiming at this problem, a lagrangian equation dynamic model for type H table is derived. Then, the improved nonsingular terminal sliding mode control（NTSMC） is adopted for single axis to design position controller. The system not only has strong robustness but also weaken the effects of chattering and improves tracking precision of the single axis. Then, between the two axis using type Sugeno fuzzy neural network（SFNN） compensation controller to dynamic compensate synchronous error of the type H table. Through fuzzy neural network can approximate arbitrary precision of nonlinear system to make the synchronous error tend to be zero in limited time for meeting high precision machining of the H tables. Simulation results show that the designed control system made the type H table possess high synchronous control precision and robustness.

关 键 词：H型平台 非奇异终端滑模控制 Sugeno型模糊神经网络补偿控制器 负载扰动 

分 类 号：TM306[电气工程—电机]