Adaptive Nn Control For Uncertain Nonlinear Systems With Prescribed Performance

In recent years,with the continuous development of modern science and technology,people have gradually realized that there are inevitably nonlinear and uncertain factors in the modeling of many practical engineering systems.Therefore,it is of great significance for the control research of this kind of uncertain nonlinear system model which has the advantages of both nonlinearity and uncertainty.On the one hand,in the operation of many practical systems such as robotic systems,electromechanical systems and helicopter systems,there are generally nonlinear factors such as fuzzy dead zone,fault tolerance,and time-varying external disturbances.The existence of such nonlinear factors not only seriously affects the expected performance of the system,but also directly leads to system instability.Therefore,the control research of such uncertain nonlinear systems with these uncertain effects has great practical value.On the other hand,in order to achieve the desired performance indicators,the state,output,and error of the controlled system are often restricted to a certain extent,and their constraint boundaries may also be time-varying in the actual industrial procedure.Therefore,for this type of system,how to design an effective controller while considering the influence of these uncertain factors in the procedure of designing the controller is not only to stabilize the controlled system,but also to ensure the prescribed performance of the system.Although many effective control strategies have been proposed for the research of this type of uncertain nonlinear systems,there are still many under-considered control problems to be solved.Based on the current research status at home and abroad,this paper will combine the adaptive prescribed performance control method and neural network backstepping technology to carry out research work on a class of uncertain nonlinear systems as follows:First,for an uncertain nonlinear system subject to fuzzy dead zone input and prescribed performance,an observer-based adaptive neural network control scheme is designed.In the procedure of designing backstepping control,performance function and conversion error are introduced,and an effective observer,fuzzy controller and adaptive law are constructed.Combined with the Lyapunov stability criterion,it is proved that the proposed control scheme can ensure the stability of the system while the tracking error strictly evolves within the predefined boundary.And,the simulation results verify the feasibility of the proposed control strategy.Secondly,the problem of adaptive neural network fixed-time control of state-constrained nonstrict-feedback nonlinear systems with event-triggering and prescribed performance is investigated.In order to meet the specified performance and state constraints,the conversion performance function and the asymmetric barrier Lyapunov function are introduced in the controller design procedure.For the unmeasurable states in the system,a fuzzy observer is designed to observe them.For the algebraic loop problem of nonstrict-feedback nonlinear systems,this chapter introduces an adaptive parameter to overcome the difficulty of unavailability of the next state variables of the current step in the form of nonstrict-feedback in the backstepping procedure.Moreover,in order to avoid the "explosion of complexity" problem,this chapter also introduces dynamic surface control technology.At the same time,combined with the relative threshold strategy control technology,an event-triggered adaptive neural network fixed-time controller is designed.Through fixed-time stability analysis,the closed-loop all signals are bounded within a fixed time and the tracking error can evolve in a predefined area,and all states do not violate the constraint boundary in a fixed time.And,the effectiveness of the method is tested through two simulation examples.In addition,a command filter-based effective neural network decentralized adaptive control strategy for nonlinear large-scale systems with time-varying asymmetric state constraints and prescribed performance is proposed,considering infinite time-varying actuator failures.By introducing the performance functions and the nonlinear transformed functions,the problems of prescribed performance control and time-varying asymmetric state constraints are solved,and the feasibility condition imposed on the virtual control signals in the control design is removed.In order to solve the problem of infinite time-varying actuator failure,the projection technology is introduced.Furthermore,the introduction of command filtering technology reduces the computational burden in the design procedure.According to Lyapunov's stability theory,it is verified that the proposed control algorithm can guarantee that all system signals are bounded,and all constraints are not violated under the condition of being free from the feasibility conditions imposed on virtual control signals,and the tracking errors can be kept within a tight set with specified performance limits even if it is allowed to occur in the case of an infinite time-varying actuator faults.Finally,the simulation results further verify the availability of the proposed control scheme.