Research On Parameter Identification And Fixed-time Trajectory Tracking Control Of Flexible Joint Robot

In recent years,flexible joint robots have been used widely in human-computer interaction,medical treatment,aerospace and other fields due to the low energy consumption,small self-weight ratio and high flexibility.Flexible joint robots have rigid-flexible coupling characteristics,which makes the traditional rigid robot mathematical model and control algorithm no longer applicable to flexible joint robots.Therefore,two important issues for flexible joint robots,parameters identification and nonlinear control of dynamic model,are studied in this paper.The main work is given as follows:Firstly,aiming at the unknown dynamic model parameters of flexible joint robots with friction terms,an identification strategy based on moth flame-genetic optimization(MF-GO)is proposed.On the basis of the standard moth flame algorithm,the moth and flame position update mechanism is improved,and the speed of early optimization is accelerated.To avoid the moth flame algorithm falling into the local optimal solution,the cross and mutation mechanism of the genetic algorithm is introduced.The higher identification accuracy and faster speed of MF-GO are verified by the simulations.Secondly,a full-order terminal sliding mode controller based on fixed-time high-order disturbance observer(FxTHODO)is designed for trajectory tracking control of flexible joint robots with uncertain disturbances.The feedback linearization method is employed in the controller to linearize the dynamics model of a 4-DOF flexible joint robot with disturbance terms;The disturbance observer is improved by combining the fixed-time theory,and a fixed-time high-order disturbance observer is designed to estimate the disturbance accurately.At the same time,the double-power reaching law is introduced on the basis of the full-order terminal sliding mode control,and the fixed-time full-order terminal sliding mode controller is designed.The simulation results show that the controller has faster response speed and higher control accuracy.Finally,a dynamic surface control method based on fixed-time extended state observer(FxTESO)is designed based on the the system model with uncertain disturbances and considering the dead-zone characteristics of input torque of each joint.FxTESO is used to estimate the state variables and uncertainties of each order subsystem.To avoid the "differential explosion" problem in traditional back-stepping control,a fixed-time first-order filter is introduced.At the same time,in order to complete the fast trajectory tracking task of the system in any initial state,a fixed-time dynamic surface controller is designed by using the estimation result of FxTESO.The simulation experiment completed the verification of the accuracy,rapidity and robustness of the controller.