Adaptive Control For Nonlinear Systems With Unknown Control Direction And Abnormal Input

In recent years,many experts and scholars have put forward effective control schemes for uncertain nonlinear systems.However,most of these control methods are designed under the assumption that the control direction is known,and the control model is mainly for single-input single-output nonlinear systems.In addition,the abnormal input characteristics,such as actuator failure,input dead zone,input saturation and so on,are inevitable in the actual engineering system.These abnormal input characteristics will have an unpredictable impact on the performance of the controlled system.Therefore,aiming at three kinds of MIMO uncertain nonlinear systems,this thesis studies the tracking control of MIMO uncertain nonlinear systems with unknown control direction and abnormal input by using adaptive control,Nussbaum gain technology,neural network approximation principle and proportional integral(PI)control as the main tools.Firstly,an adaptive neural network control algorithm is designed for a class of MIMO uncertain non-affine systems with unknown control direction.First,the non-affine system is transformed into affine system by mean value theory.Second,Nussbaum gain technology is used to solve the problem of unknown control direction of the system.For the two cases of the controlled system being square system and non-square system,the matrix decomposition technology is used for analysis and design.The nonlinear terms in the system are approximated by neural network technique.Then with the help of Lyapunov stability theory and Nussbaum function,the stability analysis is carried out.Finally,the effectiveness of the proposed control algorithm is verified by numerical simulation.Secondly,an adaptive neural network fault-tolerant PI control algorithm is designed for a class of MIMO uncertain nonlinear systems with unknown control direction and undetectable actuator faults.First,virtual matrix and actuator "health factor" are introduced to solve the time-varying and unpredictable actuator failure problems in the system.Second,the PI control is cleverly introduced into the controller to make the controller structure more simple.Moreover,the PI controller designed has adaptive PI gain adjustment.Finally,through simulation analysis and comparison with traditional PI control,the proposed control algorithm has good tracking performance.Thirdly,for a class of MIMO uncertain nonlinear systems with unknown control direction,undetectable actuator fault and asymmetric non-smooth saturation,an adaptive neural network fault-tolerant PI control algorithm based on barrier Lyapunov function(BLF)is designed.Firstly,for the problem of asymmetric and non-smooth saturation,a kind of smooth function is used to approximate the input saturation function.Secondly,the neural network is used to approximate the nonlinear terms in the system,and the barrier Lyapunov function is used to ensure that the input of the neural network is strictly in a compact set.Finally,the effectiveness of the proposed control algorithm is verified by numerical simulation.