Adaptive Decentralized Control Of Two Interconnected Systems

In the actual control engineering,many complex systems can be modeled as nonlinear interconnected systems.For example,multi-mechanical dynamical systems,electromechanical systems,coupled inverted pendulums,mass-spring-damped systems and large smart grid systems,and tank reactors.However,stochastic,time-varying and other nonlinear factors are prevalent in many fields of industry,science and technology.Therefore,further development and in-depth discussions are urgently needed for such a class of nonlinear systems,both from the theoretical development and from the demand of practical applications.Based on the development status at home and abroad as well as the above-mentioned problems,this paper investigates the corresponding adaptive decentralized adaptive control scheme for several types of nonlinear interconnected systems by combining Backstepping technology and adaptive neural network control technology.The main contents of this dissertation are as follows:(1)For nonlinear interconnected systems with stochastic terms,an improved adaptive decentralized control strategy is designed in Chapter 2.In the process of designing the controller,combined with Backstepping technology,the nonlinear items derived in the derivation process are processed by using Young's inequality and the random items in the system are processed by using Ito stochastic differential equation,the unknown functions in the system are treated by using the neural network approximation principle,and the quadratic Lyapunov function is constructed to analyze its stability.Finally,the effectiveness of the scheme is illustrated by examples.(2)For nonlinear interconnected systems with time-varying performance constraints,an effective adaptive decentralized control scheme is designed in Chapter 3.By giving pre-defined performance conditions,an error transformation mechanism is established to make the system with prescribed performance.Two performance functions are constructed in order to compensate for the effects of unknown nonlinear interactions between subsystems.Subsequently,quadratic Lyapunov functions are constructed to analyze their stability,and the tracking performance of all subsystems is verified,with tracking errors having a prescribed rate of convergence and maximum overshoot,independent of the parameters regarding the speed and the strength of the interactions.Finally,the reliability of the scheme is illustrated by examples.