A Novel Distributed and Compound Control of an Adaptive Neural Network and PD for Manipulators

2nd International Conference on Robotics and Mechantronics

计算机科学;无线电电子学

Zhao, Pu^1 ; Zhou, Yunfei^1

School of Mechanical Science and Engineering, Huazhong University of Science and Technology, 1037 Luoyu Road, Wuhan

430074, China^1

关键词: Adaptive neural networks;    Back propagation neural networks;    Compound control;    Distributed control;    Parameter-tuning;    Proportional-differential controllers;    Reference modeling;    Tracking errors;   

Others  :  https://iopscience.iop.org/article/10.1088/1757-899X/517/1/012008/pdf
DOI  :  10.1088/1757-899X/517/1/012008

学科分类：计算机科学（综合）

来源: IOP

【 摘 要 】

In this report, a new compound control scheme is proposed for manipulators based on a proportional-differential (PD) controller and a back-propagation neural network. By modifying the dynamic model of manipulators, a second-order function of tracking errors is obtained and approximated by the neural network. A PD-type reference model is introduced to update the weights of the neural network. By using additional PD controllers, rapid response of the control scheme can be realized. Simulation results highlight performance of the controller to compensate the nonlinear terms and its simpleness in the parameter tuning process. To be concluded, the controller is suitable for distributed control and can be used as supplementary of traditional PD controllers.