基于神经网络的3自由度移动机器人跟踪方法研究

Tracking Method of 3-Degrees-Freedom Mobile Robot Using Neural Networks

针对具有2个可操舵驱动轮的3自由度移动机器人的轨迹跟踪问题,利用神经网络的任意非线性逼近能力辨识出3自由度移动机器人的逆运动学模型,使系统的3个独立输入控制线性化,将一般非线性系统转化为仿射非线性系统.根据反馈控制理论,设计了跟踪控制器,使系统具有良好的全局稳定性,实现了移动机器人的位置和方位的独立跟踪.最后,通过仿真验证了算法的可行性和有效性.

To solve the problem of path tracking for 3degreesfreedom mobile robots with two drivable and steerable wheels, the inverse kinematics model is identified using capability to approximate arbitrary nonlinear function of neural networks, which makes control inputs linearized and transforms the general nonlinear system into affine nonlinear system. Based on feedback control theory, a trajectorytracking controller is designed, so the tracking performance of system has the global stability and the independent control of both position and orientation for a mobile robot is realized. Simulation shows the feasibility and efficiency of the algorithm.