Adaptive Neural Network Control Of Flexible Joint Manipulator With Time Delay

With the swift development of science and technology,the manipulator has been utilized in diverse industries based on its high efficiency,appropriately price and powerful environmental adaptability.Flexible manipulator gradually replaces the traditional rigid manipulator because of its good flexibility,high environmental adaptability,and comprehensive human-computer interaction.Hence,the research of flexible manipulator is particularly important.In practical application,the working performance of the manipulator is often affected by the internal conditions and external environment of the manipulator.Such as,the control delay caused by the joint backlash of the manipulator,the limitation of the load range of the manipulator,and the unavoidable external environment interference.In view of the above control problems of the manipulator,the adaptive neural network control strategies for several classes of single-link flexible joint manipulator systems with time delay have been presented in this article.(1)Based on Lagrange equation,the rigid manipulator model with motor drive is constructed.Considering the flexible deformation of the manipulator in motion and the influence of the time-delay phenomenon based on the joint backlash of the manipulator on the system,we improved the model to the one driven by motor.(2)After simple mathematical transformation,the improved manipulator model is transformed into partial differential equations suitable for backstepping control method.The Lyapunov-Krasovskii functional is used to deal with the time-delay link in the system of equations.The unknown nonlinear functions of systems can be approximated by utilizing the properties of radial basis function neural network.(3)The three control strategies investigated in this article can be analyzed by Lyapunov stability theory.Meanwhile,the effectiveness of the control methods has been proved.(4)We have carried on the simulation verification to the actual manipulator system.The experimental results show that under the influences of time delay,communication resource limitation and external interference,the three control strategies proposed in this article can still accurately control the manipulator system.