Intelligent Decentralized Control For A Class Of Constrained Nonlinear Interconnection Models With Extended Lower Triangular Structure

With the rapid development of modern industry,the nonlinear interconnection models with extended lower triangular structure exist widely in engineering systems.In view of the complex structure of such models,the analysis,design and intelligent control for them have become a challenging task.At the same time,actual control systems usually need to consider environmental factors,performance requirements and safety so that the controlled systems often have different types of input and output constraints.Therefore,the constrained control of system models has become a widely concerned research topic in the whole wide world,and there are a lot of design problems that need to be solved urgently.In this thesis,for a class of nonlinear interconnection models with extended lower triangular structure,under different assumptions of interconnection terms,design schemes are proposed for different types of constraints via combining the Backstepping design method with the adaptive neural network technology.In addition,the corresponding intelligent decentralized control algorithms and the stability of the closed-loop systems are studied.The specific work is as follows:A decentralized adaptive barrier-function-based intelligent control scheme is proposed for the output tracking problem of a class of nonlinear interconnected systems with extended lower triangular structure and asymmetric constant output constraint requirements.The proposed barrier function is universal and thus can be applied to systems with asymmetric constraints,as well as systems with symmetric constraint or unconstrained requirements without the need to modify the control strategy.The obstacles caused by the existence of both the extended lower triangular structure and the unknown interconnections are overcome by taking advantage of the inherent properties of the Gaussian functions in neural networks and introducing a smooth function,respectively.Moreover,the designed intelligent tracking controller guarantees that all signals of the overall closed-loop system are uniformly ultimately bounded,which is independent of the strength of the interactions.Moreover,the output tracking error of each subsystem in the interconnected system can be convergent to a predetermined and arbitrarily small range while satisfying the asymmetric constant constraint requirement of the output tracking error.Finally,simulation results for double inverted pendulum show the validity of the proposed intelligent control schemeFor the considered nonlinear interconnected systems with extended lower triangular structure,time-varying asymmetric output constraints and dead-zone inputs,the adaptive decentralized intelligent control problem is investigated.The existence of the extended lower triangular structure,time-varying asymmetric output constraints and dead-zone inputs makes the construction of the intelligent controller of each subsystem very difficult.Therefore,under designed intelligent control scheme,radical basis function(RBF)neural networks are used to approximate the unknown nonlinear system functions,the extended lower triangular structure of the system is skillfully solved via combining the Backstepping design method and the inherent property of the basis function in neural networks.Furthermore,Lyapunov stability analysis shows that all the signals of the closed-loop system are ultimately bounded,and each subsystem output can converge to an arbitrarily small and predefined time-varying range with the corresponding constraint is always satisfied by employing a barrier Lyapunov function(BLF)Finally,simulation results based on a practical example prove the effectiveness of the proposed design strategy.