Composite Control Of Flexible Manipulators Based On Adaptive Dynamic Programming

Flexible manipulators have been widely applied in modern industry because of their many advantages such as light weight,fast motion,low cost,lower manufacturing consuming,etc.The motion of flexible manipulators includes macro rigid-body rotation and micro flexible vibration,and the overall dynamic process is subject to external disturbances.In addition,it has complex characteristics such as rigid-flexible coupling,non-linearity,uncertainties etc.The existing methods lack comprehensive consideration of adaptability and optimality.And it is difficult to realize high-precision motion control in complex environment.In this thesis,the problem of trajectory tracking and vibration suppression of flexible manipulators is studied by singular perturbation(SP)and adaptive dynamic programming(ADP)in order to improve the adaptability and optimality of the system.1.A two-time-scale model of flexible manipulators is established.Firstly,by using Lagrange and assumed-mode methods,the dynamic model of flexible manipulators is established.Then,the dynamic model is decomposed in two-time-scales by SP theory.Finally,the validity of the two-time-scale model of flexible manipulators based on SP decomposition is verified by simulation experiments.2.A composite controller of flexible manipulators based on sliding-mode control and ADP is proposed.Firstly,using the slow-time-scale model(slow sub-system),a trajectory tracking controller is designed by using sliding-mode control theory based on quasi-sliding mode.The influence of sliding-mode buffeting is avoided at the same time.Secondly,a vibration suppression controller with inaccurate known parameters is designed based on the fast-time-scale model(fast sub-system)by using ADP.Finally,the performance of the composite controller is verified by simulation in Matlab/Simulink environment.Compared with the existing methods,the controller leads to a better vibration suppression effect and higher tracking accuracy.3.A composite controller of flexible manipulators based on duel ADP is proposed.Firstly,a trajectory tracking controller is designed by ADP based on the measurements of the slow inputs and the position in the slow-time-scale,while the vibration in slow-time-scale is estimated by least square method.Secondly,the vibration in fast-time-scale is reconstructed based on the measurements of vibration and its estimate in slow-time-scale,by which a vibration suppression controller is designed using ADP.Then,stability of the closed-loop system is proved by SP theory.Finally,simulation results are given to show the feasibility and effectiveness of the proposed methods.