Research On Adaptive Neural Network Control Of Rigid-Flexible Manipulator System Based On PDE

In recent years,with the rapid development of industry,aerospace and other fields,people are gradually facing some extremely challenging tasks,such as deep sea development,space exploration,etc.In order to solve this problem,the robotic arm has gradually entered the human field of vision.The traditional rigid manipulator brings great convenience to people in handling,grasping and other work.Flexible materials have received extensive attention in various fields due to their light weight,high flexibility,low energy consumption,and safety of human-computer interaction,and are very suitable for robotic arm systems.However,due to its flexibility,elastic vibration and even damage of the manipulator are often caused in the movement process of the manipulator or in the presence of external environmental disturbances.Therefore,it is an important and challenging problem to suppress the vibration of the flexible part while ensuring the operational motion of the manipulator.In this paper,the position tracking and vibration suppression of a coupled rigid-flexible manipulator system under external disturbances,input or output constraints are studied.Firstly,the distributed parameter model of the coupled rigid-flexible manipulator system is derived by using the Hamilton's principle.In order to avoid the overflow problem caused by ignoring higher-order modes,the finite-dimensional approximation of the model is not performed,and the controller is designed directly for the infinite-dimensional system model.Then the energy analysis of the flexible link part in the system is carried out.Secondly,the adaptive neural network control method thinking about the external disturbance and parameter uncertainty in the system is carefully thought about.Based on the energy analysis of the flexible links of the system,a coupled sliding surface is designed which consists of the joint angle,angular velocity,strain force and shear force at the root of the flexible link.Considering external disturbances and parameter uncertainty,an adaptive neural network sliding mode control method based on PDE is designed,which effectively compensates the external disturbances and has good nonlinear approximation characteristics.The adaptive method is used to estimate the uncertain parameter in the system The asymptotic stability of the closed-loop system is proved by semigroup theory and La Salle's Invariance Principle,and the effectiveness of the controller is verified by simulation.Again,consider a direct joint control approach where the system has input constraints and output constraints.An adaptive fault-tolerant control scheme based on asymmetric output constraints and actuator partial failure was proposed by using piecewise barrier Lyapunov function.For manipulator system with input constraints and external disturbances,hyperbolic tangent function and hyperbolic secant function were used to solve the input saturation problem,and direct joint torque control was designed based on disturbance observer to estimate external disturbances,so as to realize joint position control and suppress vibration of flexible link.Finally,the asymptotic stability of the system is proved by semigroup theory and La Salle's Invariance Principle,and the control scheme is verified by simulation.Finally,the content of the full text is summarized,and the future work is planned and prospected based on the research experience of coupled rigid-flexible manipulator system.