摘要
本文针对机器人动力学系统的非线性、强耦合和时变等特性,设计出了基于神经网络算法的轨迹跟踪控制系统。在控制算法的实现方面,以BP网络为基础,加入反馈信号和偏差单元,生成了内部回归神经网络。克服了BP网络学习收敛速度慢、容易陷入局部极小点等缺点,大大提高了学习速率,并且保证了学习过程的稳定性。
Aimed at the high nonlinear, coupling and time-varying characteristics of robot dynamics, it designs a trajectory tracking control system based on neural network in this paper.In the control algorithm realization, it joins the feedback signal and the deviation unit and produces Inner Regression Neural Network based on BP network.The control system overcomes the shortcomings of BP neural network such as slow convergence and fall into the partly extreme minimum value easily and so on,increases the learning speed greatly and guarantees the stability in the learning process.
出处
《微计算机信息》
2010年第23期136-137,150,共3页
Control & Automation
关键词
机器人
神经网络
轨迹跟踪
robot
neural network
trajectory tracking
