Finite-time Control Research Of Several Class Of P-normal Form Nonlinear Systems

Finite-time control problem has caused extensive attention of domestic and overseas scholars.Because p-normal form nonlinear systems models have wider applicability in actual engineering,the paper researches five problems related to finite-time control for several classes of p-normal form nonlinear systems based on backstepping technique.The main research contents are as follows:1)The new neural adaptive exact finite-time stabilisation method is proposed for a class of p-normal form strict-feedback nonlinear systems.The proposed method extends the assumption for unknown nonlinear function of systems in other paper about exact finite-time control.And the impact of the neural network approximate error for steady-state performance is removed,such that the controlled system can converge into the equilibrium rather than the neighbourhood of the equilibrium within finite time,and the extimation error of the adaptive parameter is bounded.The simulation studies compared with other method show that the control method proposed has better transient control performance,which proves the effectiveness and the superiority of the proposed method.2)The fast exact finite-time H? sabilisation problem with output constraint is studied for a class of p-normal form nonlinear systems with external disturbances.The fast exact finite-time H? control method is first applied in p-normal form system.The output-constrained fast exact finite-time H? controller is designed,which can guarantee that the considered system is fast exact finite-time stable,and the system output is constrained by a given bound.At the same time,the influence of external disturbances can be attenuated.The simulation comparison studies demonstrate the effectiveness and the superiority of the proposed method.3)The practical finite-time prescribed performance H? tracking control problem is studied for a class of non-strict feedback p-normal form nonlinear systems with external disturbances.A new design thought of prescribed performance controller is proposed,which avoids the shortcomings of classic prescribed performance control design.Not only the designed controller can guarantee that the tracking error of the controlled system converge to a predetermined neighborhood of the equilibrium within the arbitrarily given settling time,but also the impact of external disturbances can be attenuated by the H? performance.Specially,the settling time is irrelevant to the initial states of the system.Simulation results show the effectiveness and the advantage of the proposed method.4)The decentralized neural adaptive exact finite-time connective stabilization control method is first proposed for a class of p-normal form large-scale nonlinear systems.The decentralized exact finite-time controller is designed for every subsystem in the large-scale system,respectively.The designed controllers can guarantee that all of the large-scale system states are exact finite-time stable,and the large-scale system is connectively stable.Simulation results are provided to show the effectiveness and the advantage of the proposed method.5)A decentralized connectively practical finite-time stabilization control method is proposed for a class of p-normal form nonlinear interconnected large-scale systems with expanding construction.Namely,a new subsystem is added to the original system online.Under the premise that the original controllers are kept to be unchanged,only the control laws and the adaptive laws for the newly added subsystem need to be designed,which can guarantee that both newly added subsystem and new resultant expanded large-scale system are connectively practically finite-time stable.Not only that,the singularity problem existing in practical finite-time control is solved.Simulation results show the effectiveness and the advantage of the proposed method.