平面二自由度五杆机构BP神经网络整定PID控制

Design PID Control based on BP Neural Network of 2-DOF Five-bar Mechanism

针对平面二自由度五杆机构的轨迹跟踪问题,设计了一种基于BP神经网络整定的PID控制方法。该控制方案利用BP神经网络学习辨识机构系统的未知非线性动态,可以在线调整PID控制参数,以实现高精度控制。仿真结果显示该控制策略可以精确实现对于五杆机构的轨迹跟踪控制,该方法具有较好的自适应性和鲁棒性。

This paper designs adaptive Proportion Integral Differential（PID）control algorithm based on BP neural network for trajectory tracking of 2-DOF five-bar mechanism. In this scheme,a BP neural network is used to approximate the unknown nonlinear dynamics of the mechanism,then the PID parameters can be adjusted online and high precision can be obtained. Simulation results show that the control algorithm can accurately track trajectories of a 2-DOF five-bar mechanism. The results also indicate that controller has better robustness and adaptability.