Adaptive Control For Nonlinear Systems With Time-Varying Constraints

Modern industrial processes tend to be more and more large-scale and complicated,presenting a high degree of non-linearity and serious uncertainty.Moreover,due to the influence of actual mechanical mechanism,safety and other indicators,the constraint problem inevitably appears in the actual systems.If these constraints are not handled or mishandled,it is bound to affect the stability of the system and even cause serious safety accidents.Therefore,this thesis will construct different adaptive controllers for several classes of uncertain nonlinear systems by combining Barrier Lyapunov Functions(BLFs),backstepping control design and neural networks.The main contents are as follows:(1)An adaptive tracking control approach is proposed for nonlinear pure-feedback systems with time-varying full state constraints.The pure-feedback systems will be transformed into strict feedback forms by using the mean value theorem.The neural networks are employed to approximate unknown items in the newly generated system.The appropriate time-varying BLFs are constructed based on error variables to avoid the violation of full state constraints in the process of designing adaptive controller by backstepping.The stability analysis is carried out by using Lyapunov stability theorem.Finally,two simulation examples reveal the performance of the proposed control approach.(2)An adaptive tracking control method is proposed for nonlinear strict feedback systems with time-varying full state constraints.The neural networks are used to model unknown items in the system.The new type of time-varying integral type BLFs are constructed,and an adaptive controller is designed in combination with the backstepping,which realizes the direct constraint of state in the system,that is,to overcome the conservative restriction of transforming the original state constraint into of tracking error constraints in the existing constraint control methods.The stability is analyzed by using Lyapunov stability theorem.Finally,the effectiveness of the proposed control method is illustrated through a simulation example.(3)An adaptive neural network control approach is developed for stochastic nonlinear strict feedback systems with time-varying full state constraints.The neural networks are used to online identify unknown functions in systems.The time-varying tangent type BLFs of the fourth power are constructed,which integrate constraint analysis into a common method which can deal with both constrained and unconstrained systems.And the adaptive controller is designed with the backstepping.The Lyapunov stability theory is used to prove the stability of the system.Finally,a simulation example exhibits the effectiveness of the proposed control approach.(4)A finite-time adaptive control method is proposed for nonlinear strict feedback systems with time-varying asymmetric full state constraints.The approximation properties of neural networks are used to deal with the unknown terms in the systems.The asymmetrical time-varying BLFs are introduced to ensure that the time-varying full state constraints are not violated.The adaptive controller is designed by combining the backstepping.It is based on the Lyapunov stability theory to prove that the asymmetrical full state constraints are not violated,all the closed-loop signals are semi-global uniformly ultimately bounded,and the error converges to small neighborhood of zero in finite-time.Finally,a simulation example is given to verify the effectiveness of the proposed finite time control method.