基于RBF神经网络的翻曲机器人自适应滑模变结构控制

摘要

Aiming at the problems of the uncertainty of joint chattering and control in a wine turning robot,a linear feedback sliding control strategy based on RBF neural network is proposed combined with the robot parts structure,materials and movement of the actual situation from the dynamic point of view.By using the linear feedback RBF,the sliding mode controller is compensated without uncertainty,and the control effect is further improved.The results show that the proposed linear feedback RBF neural sliding mode Variable structure control based on position control strategy has good accuracy and high convergence speed,and effectively reduces chattering.%针对酿酒翻曲机器人关节抖振和控制存在的不确定性等问题,从动力学角度出发,结合机器人部件结构、材料和运动实际情况,提出一种基于RBF神经网络的线性反馈滑模变结构控制策略.通过线性反馈RBF神经网络系统实现对滑模变控制器不确定性的补偿,进一步改善控制效果.研究结果表明,文中提出的基于线性反馈RBF神经网络神经的滑模变结构控制策略,具有良好的位置控制精度和收敛速度,且能有效地削弱抖振.