End edge feedback and RBF neural network based vibration control of flexible manipulator

Abstract

A vision feedback based active vibration control system for flexible manipulator is proposed. Active vibration control is conducted by employing Radial Basis Function (RBF) neural networks to regulate the control parameters of the proportional and derivative (PD) controller. Firstly, the composition of the vision feedback based piezoelectric flexible manipulator system is described. Furthermore, vision based vibration measurement for the flexible manipulator's end side and image processing methods are introduced. Secondly, the system dynamics model is derived by using finite element method, and the equation of the state space representation for the lower mode vibration is provided. Then, the vision feedback based control method is expounded, using RBF neural networks to tune the control parameters dynamically. Finally, simulation and experiments are conducted. The results demonstrate that vision feedback based control system and the RBF neural networks control algorithm can suppress the vibration of the flexible manipulator effectively.