摘要
探测车的运动控制是一个相当复杂和困难的控制问题,由于运动路面特性与探测车的大惯性,传统的PID算法无法满足控制要求。建立了高低精度模型,将自适应模型与增广神经网络结合,并应用于探测车的速度和转向控制。控制器设计包含两个子控制器:(1)基于线性化反馈的控制器;(2)在线自适应神经网络调节的比例积分(PI)的反馈的控制器。用于对速度及转向角的控制。仿真实验表明,该方法对于车辆速度、转向角控制有良好的适应性,控制精度高,稳定性好。
Exploration vehicle＇s motion control is a very complex and difficult problem.In view of the exploration vehicle＇s inertia characteristics and operating surface characteristics,the PID control couldn＇t meet the control precision.The high-fidelity model and low-fidelity model are established,and the control arithmetic is brought out when the adaptive control combines with the neural network command augmentation,and the arithmetic is applied to the control of velocity and steering on the exploration vehicle.The control design consists of two parts of controllers：（1）a feedback linearization based controller;（2） a proportional-integral（PI）feedback that is augmented by an online adaptive neural network.The controllers are used to the velocity and wheel angle.The simulation results illustrate that the arithmetic has flexibility to the velocity and wheel angle,high precision control,and good stability.
出处
《机械设计与制造》
北大核心
2012年第11期117-119,共3页
Machinery Design & Manufacture
基金
引进人才基金资助课题(YKJ1017)
关键词
探测车模型
自适应
增广神经网络
转向控制
Exploration Vehicle Model
Adaptation
Neural Network Augmentation
Wheel Angle Control
