基于控制器与辨识器的转台系统设计  

Design of Turntable System Based on Controller and Identifier

作　　者：申洁[1] 王新民[1] 

机构地区：[1]西北工业大学自动化学院,陕西西安710072

出　　处：《计算机仿真》2008年第9期324-326,330,共4页Computer Simulation

摘　　要：为改善三轴转台系统性能,结合传统控制方法与神经网络控制,提出一种基于RBF辨识转台系统的CMAC神经网络与PID并行的复合控制算法。算法采用RBF辨识对象模型,CMAC实现前馈控制,并实现PID控制参数的在线整定和优化。也给出了CMAC控制器算法和系统辨识的RBF网络算法。以某转台模型为对象,仿真结果表明算法具有了传统控制的优点,进一步也证明了算法的可行性和优越性,且具有了更强的适应性和鲁棒性,能更为有效地应用于转台系统中。In order to improve three axis turntable system＇s performance,by combining the traditional control method with the nerve network control,a compound control algorithm based on RBF identification turntable system is proposed.This algorithm recognizes the object model using RBF nerve network,uses the CMAC to realize feed-forward control,and realizes the PID control variable online tuning and optimization.In the article the RBF network algorithm for identifying the system and CMAC controller algorithm is given.The simulation result indicates that this algorithm has the traditional control merit,and the feasibility and superiority of the algorithm are also verified.

关 键 词：转台系统 小脑模型神经网络控制器 神经网络辨识器 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]