Research On Control Algorithm Of Flexible Joint Manipulator With Input And Output Constraints

China's manufacturing industry continues to develop towards high-end,intelligent,and informatization.Due to the intricacies of the work environment and the increasing frequency of human-computer interaction,various types of robots have emerged to adapt to the different environments,a type of flexible joint manipulators with high load ratio,small quality and reliable safety but accompanied by accuracy and chattering problems have received extensive attention.Because the actuator inevitably has nonlinear characteristics such as input saturation and dead zone;at the same time,with the advancement of science and technology,the requirements for transient and steady-state control performance are increasing,and the system output needs to be constrained;and because of the existence of flexibility,it is very difficult to design a controller that achieves high precision,no overshoot,and no chattering performance on the basis of ensuring trajectory tracking.For the flexible joint manipulator with input and output constraints,the singular perturbation method is used to realize the decomposition of rigidity and flexibility through time scale transformation,and the flexible joint manipulator model is decomposed into two subsystems of fast(flexible)and slow(rigid)and the control rate is designed separately.The total control rate is a simple linear superposition of two subcontrollers,at the same time,the actual requirements of input and output constraints are considered,and the precise tracking of the trajectory and the suppression of flexibility are realized.1.Aiming at the trajectory tracking problem of a flexible joint manipulator system with model uncertainties under the condition of limited input,an backstepping sliding mode controller based on singular perturbation method is proposed.First,design a speed difference feedback controller for the fast(boundary layer)system to suppress the vibration of the joint flexibility.And for the slow(reduced order)subsystems,design a class of nonlinear integral sliding mode surface to increase the system's steady-state accuracy and the introduction of the global sliding mode surface increases the robustness,thereby improving the control performance of the system.Among them,the radial basis basis(RBF)neural network is used to realize the compensation of system uncertainty and actuator saturation,and the adaptive rate of the neural network weights is designed.Based on the Lyapunov theory,the stability of the system and the convergence to the error are proved.The simulation results show that the designed control rate excellent good speed and accuracy,and the chattering phenomenon is basically eliminated.2.Aiming at the trajectory tracking problem of the flexible joint manipulator system with output constraints,based on the singular perturbation model,the slow subcontroller is designed by combining the Log-barrier Lyapunov function(BLF)with the backstepping method,and the constrain of tracking error is realized.At the same time,a linear sliding mode fast subcontroller is designed to suppress the flexible chattering,and a complete proof of stability is given.Finally,the effectiveness of the method is proved by digital simulation.3.For the flexible joint manipulator system with output constraints,the traditional controller based on Tan-BLF has the shortcoming that the error convergence domain is constant.Based on the singular perturbation method,a globally convergent piecewise control strategy is proposed,which extends the convergence domain to the whole world and solves the problem of excessive the initial error will cause the system to diverge;then a linear quadratic optimal controller is proposed for the fast subsystem.The effectiveness of the controller is verified by simulation.4.For the flexible joint manipulator system with output constraints and model uncertainty,a fixed-time prescribe performance tracking control method based on singular perturbation method is proposed.For the slow subsystem,a new type of time-varying constraint boundary is first constructed,which broadens the scope of application of the time-varying tangent barrier Lyapunov function.Then,a class of nonlinear disturbance observer is used to accurately estimate the exogenous disturbance.Secondly,a composite controller based on non-singular fast integral terminal sliding mode is constructed to make any initial tracking error converge in a fixed time.So far,the transient and steady-state performance constraints of the tracking error of the manipulator are guaranteed.For the fast subsystem,a controller that guarantees the suppression of flexible vibration within a fixed time and has a superior performance is also proposed.Finally,combined with the Lyapunov stability theory,the stability of the closed-loop system is strictly proved and the effectiveness of the controller is verified by simulation comparison.