Research On Control Problems Of Several Types Of Uncertain Nonlinear System

With the continuous development of the industrialization and urbanization,the scale and complexity of the nonlinear systems are also gradually increasing in practical engineering,which makes the performance requirements of the controlled system are constantly improving.As is known to all,the adaptive intelligent control based on neural network(NN)and fuzzy system is an effective way to solve the control issues of uncertain nonlinear systems,however,in the adaptive intelligent control field,it will be more difficult to research the better control quality,such as finite-time control and optimal control,that is,to achieve the stability and convergence of the controlled system in a shorter time or small energy consumption.Now,researching on the adaptive intelligent finite-time and optimal control issues for uncertain nonlinear systems have just started,and there exist several control problems that need to be further studied.Under this background,this thesis will study the adaptive intelligent finite-time and optimal control design problems for several typical classes of complex nonlinear systems based on fuzzy system and NN theories,and give the proof methods of the stability and convergence for the closed-loop system.The main contents of this thesis are as follows:1.Adaptive intelligent finite-time control design problem is studied for uncertain nontriangular structural nonlinear systems.In the considered nonlinear system,since there exist the odd powers in virtual and actual control inputs,with the help of adding a power integrator technique,by constructing the Barrier Lyapunov function(BLF),and establishing the nonlinear filter via odd power,an adaptive fuzzy finite-time control design method is developed,which can ensure the output does not beyond the constraint bound,and all signals of the controlled system are finite-time bounded.On this foundation,researching on the robust adaptive fuzzy finite-time control problem for nonlinear system with unmodeled dynamics.By introducing the dynamic signal function,a robust adaptive fuzzy finite-time control design method is developed to suppress the external disturbances and unmodeled dynamics,and ensured the controlled system is input-to-state practically stable,and all signals of the controlled system are finitetime bounded.Simulation results are given to further verify the effectiveness of the developed control method.2.Adaptive intelligent fixed-time control design problem is investigated for uncertain nonlinear strict-feedback systems.By introducing smooth projection operator,a novel adaptive law is designed.Based on adding a power integrator technique,constructing the nonlinear filter via fractional power and adaptive parameters,a global adaptive NN finite-time control method is presented,which can guarantee the controlled system is global finite-time stable,and the output can track the desired signal within finite time.On this foundation,researching on the adaptive fuzzy fixed-time decentralized control design problem for nonlinear interconnected large-scale systems.With the help of fixed-time stability and stochastic stability theories,a robust adaptive fuzzy fixed-time decentralized control design method is proposed,the proof method of the stability of the closed-loop system and convergence of the errors are given.Simulation results are given to further verify the effectiveness of the developed control method.3.Adaptive intelligent optimization prescribed performance control problem is studied for stochastic nonlinear systems with input and state constraints.By constructing homeomorphic mapping of the state constraints,an equivalent fuzzy constraint system is established.Establishing feedforward fuzzy compensator,the influence raised by input constraint can be offset.Based on this system,by constructing identify-actor-critic NNs construction,an identifier-based robust adaptive fuzzy optimization performance constraint control design method is developed,which can guarantee all signals in controlled system are uniformly ultimately bounded(UUB),and the transient and steady-state performance constraint of tracking errors can be realized.On this foundation,the effectiveness of the developed adaptive intelligent optimal performance constraint control method can be further verified by simulation of the third-order nonlinear vehicular platoon system.4.Adaptive intelligent consensus optimal fault-tolerant control problem is studied for uncertain nonlinear multi-agent systems with actuator failures.By constructing identify-actorcritic NNs construction,an identifier-based robust adaptive NN optimal fault-tolerant control design method is developed,which can ensure all signals of the closed-loop system are bounded in probability,and the followers’ states can track the states of the leader.On this foundation,researching on the robust adaptive fuzzy optimal full state constraint control problem for nonlinear multi-agent systems with nonlinear faults.By introducing the law-pass filter,the influence caused by nonlinear fault can be solved.Based on BLF,constructing barrier-type cost function,combining the backstepping control technique,a robust adaptive fuzzy optimal full state constraint control method is presented,which can ensure all signals of the controlled system are bounded,all states do not beyond their constraint bound.And each subsystem can realize optimal.Simulation results are given to further verify the effectiveness of the developed control method.5.Adaptive intelligent finite-time optimization prescribed performance control problem is studied for uncertain nonlinear strict feedback systems.By constructing the transfer function,tracking and virtual errors are transformed,and BLF based on the transformation variable is established.By the aid of adding a power integrator technique and BLF,combining the backstepping control,a reinforcement learning(RL)-based adaptive NNs finite-time optimization performance constraint control design method is developed,which can guarantee all signals of the controlled system are finite-time bounded,all state errors do not beyond the preset region.And the minimization of the cost function can be ensured.On this foundation,the effectiveness of the developed adaptive intelligent finite-time optimal performance constraint control method can be further verified by simulation of the underactuated autonomous underwater vehicles(AUV).