单级倒立摆系统的神经网络逆模控制  

Inverse Control Based on Neural Network for Single Inverted Pendulum

作　　者：张丹红[1] 王勤[1] 

机构地区：[1]武汉理工大学自动化学院,湖北武汉430070

出　　处：《武汉理工大学学报（信息与管理工程版）》2006年第5期28-31,共4页Journal of Wuhan University of Technology：Information & Management Engineering

基　　金：武汉市青年科技晨光计划资助项目(20035002016-05)

摘　　要：给出了单级倒立摆的一种神经网络逆模控制方案。首先,采用PD控制器控制倒立摆,采集倒立摆的输入输出数据;然后,以这些数据为基础训练倒立摆的神经网络逆模型;再采用比例控制器与逆模型串联构成逆模控制器。整体结构上单级倒立摆系统是一种闭环的逆模控制结构。仿真结果表明,所采用的神经网络逆模控制比传统的PD控制具有更短的调节时间和更小的超调,能更好地实现倒立摆的稳定控制。An inverse controller based on neural network for single inverted pendulum is proposed. Firstly, a PD controller is used to stabilize the inverted pendulum, and the inverted pendulum＇s input and output are sampled. Then the neural network based inverse model of the inverted pendulum is trained. Finally, a proportional controller is connected in series with the inverse model to construct the inverse controller. The inverse control system is a closed -loop control system. Simulation results show that this control strategy has shorter settling time and smaller overshoot and it can control the inverted pendulum better.

关 键 词：倒立摆 PD控制器 神经网络 逆模控制 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]