Research On Adaptive Control Of Uncertain Manipulator With Considering Convergence Time And Constraints

The robot industry has entered a new period of rapid development as a result of the deep integration of information technology,biotechnology,new material technology,and so on.The nonlinearity and uncertainty of robotic system pose challenges to controller design.Considering the input constraints formed by the controller components,transient and steady-state performance requirements,or constraints to ensure safe operation,which increases the difficulty of controller design.In some application scenarios,it is necessary to consider the convergence time of the system to meet the response time requirements,improve work efficiency,achieve rapid target interception,and so on.For example,the spacecraft stabilizes to the target attitude within the predefined time,the palletizing robot is required to complete the work within the predefined time,the interceptor missile system is required to intercept the maneuvering target within the predefined time,and so on.Therefore,the convergence time should be considered when designing the controller.Taking the backstepping method,sliding mode control,and neural network as the core techniques,combined with prescribed performance control,nonlinear mapping,funnel control,temporal scale transformation,fixed-time stability theory,and predefined-time stability theory,this article focuses on the adaptive control method for uncertain manipulator with convergence time constraints and input-output constraints.The following are the primary research contents addressed in the entire paper:1.For the uncertain manipulator system with input and output constraints,the fixed-time adaptive neural network controller design scheme is proposed.Firstly,for the uncertain manipulator system subject to input constraints and operating performance requirements,the prescribed performance control is used to make the system meet the performance requirements and a fixedtime auxiliary dynamic system is designed to deal with the input saturation constraints,combined with the RBF neural network and sliding mode control,a fixed-time adaptive neural network tracking control algorithm is designed.On this basis,for an uncertain manipulator system subject to input and asymmetric time-varying output constraints,the nonlinear mapping is used to deal with the asymmetric time-varying output constraints and the RBF neural network is used to estimate the uncertainty of the system,combined with a fixed-time auxiliary dynamic system and the backstepping method,a fixed-time adaptive neural network tracking control algorithm is designed.Through Lyapunov stability analysis,it is proved that the proposed tracking control algorithms can make the tracking error converge to the neighborhood of origin within a fixed time and satisfy the input and output constraints.Finally,the efficacy of the proposed control methods is verified by comparison simulation.2.For the uncertain manipulator system with input constraints,the predefined-time adaptive neural network controller design scheme is proposed.Firstly,for the uncertain manipulator system subject to dead zone input constraints,by simplifying the dead zone model to the superposition form of the linear part and the disturbance part and using the RBF neural network to approximate the lumped uncertainty of the system,combined with the backstepping method and temporal scale transformation,a predefined-time adaptive neural network tracking control algorithm is designed.On this basis,for the uncertain manipulator system subject to input saturation constraints,by designing an auxiliary dynamic system to deal with the input saturation constraints,combined with the RBF neural network,sliding mode control,and temporal scale transformation,a predefined-time adaptive neural network tracking control algorithm is designed.Using Lyapunov stability theory,it is proved that the designed control algorithms can achieve convergence of tracking error within a predefined time without violating the input constraints.Finally,the effectiveness of the proposed control methods is verified by simulation research.3.For the uncertain manipulator system with constraints and external disturbances,the predefined-time robust adaptive controller design scheme is proposed.Firstly,for the uncertain manipulator system subject to operating performance requirements and external disturbances,the funnel control is used to make the system meet the performance requirements,combined with the RBF neural network and the backstepping method,a predefined-time adaptive neural network robust tracking control algorithm is designed.On this basis,for an uncertain manipulator system subject to input constraints and external disturbances,a predefined-time disturbance observer is designed to estimate the aggregate uncertainty of the system and a predefined-time auxiliary dynamic system is designed to deal with the input saturation constraints,combined with the backstepping method,a robust adaptive tracking control algorithm based on the predefinedtime disturbance observer is designed.Through Lyapunov stability analysis,it is proved that the designed robust adaptive tracking control algorithms can ensure the system output tracks the desired trajectory within a predefined time under constraints.Finally,the simulation results show that the designed control algorithms can achieve the desired control objective.4.For the uncertain manipulator system with input constraints and in contact with the external environment,the fixed-time and the predefined-time adaptive neural network force/position hybrid controller design scheme is proposed.Firstly,a fixed-time auxiliary dynamic system is designed to deal with the input saturation constraints,combined with the RBF neural network and sliding mode control,a fixed-time adaptive neural network force/position hybrid control algorithm is designed.On this basis,a new predefined-time Lyapunov stability condition is proposed and a new predefined-time auxiliary dynamic system is designed to deal with the input saturation constraints,combined with the RBF neural network and the backstepping method,a predefined-time adaptive neural network force/position hybrid control algorithm is designed.Using Lyapunov stability theory,it is proved that in the presence of input constraints,the designed force/position hybrid control algorithms not only ensure the boundedness of force tracking error but also realize the fixed-time or predefined-time convergence of position tracking error.Finally,simulation results verify the effectiveness of the proposed control algorithms.