Decentralized Iterative Learning Control For Some Classes Of Singular Large-scale Interconnected Linear Systems

In this thesis,the decentralized iterative learning control for singular large-scale interconnected linear systems is studied.The text is divided into five chapters,and the detailed contents are as follows:In the first chapter,we introduce the theory of iterative learning control,and analyze the significance and research status.Finally,the main work of this paper is presented.In the second chapter,the state tracking problem for a class of singular large-scale interconnected continuous-time systems is put forward.By decomposing singular matrix,the original systems are transformed into a class of differential-algebraic systems equivalently,then a D-P controller is constructed.As a result of this algorithm,the states of the systems can converge uniformly to the desired trajectory.In the third chapter,the state tracking problem for a class of singular large-scale interconnected discrete-time systems is further studied.The systems are divided into a class of difference-algebraic systems equivalently.Under the action of the D-P type algorithm,the output tracking error of the system is convergent along the iterative axis.In the fourth chapter,we study the decentralized iterative learning control of singular large-scale interconnected regular systems.Based on the state tracking problem in the second and third chapters,the decentralized iterative learning control problem for singular large-scale interconnected systems with output is further considered in this chapter.Here,the discrete-time systems and continuous-time systems are obtained.Under the influence of the P-type algorithm,the output of the system can converge to a given desired trajectory along the axis direction.In the fifth chapter,the decentralized iterative learning control problem for a class of large-scale interconnected non-affine discrete-time systems is presented.The P-type learning control algorithm is designed.it is proved that the proposed algorithm can guarantee the convergence of the output tracking errors for each subsystem along the iterative axis.Finally,the conclusions of this paper are given and some problems that need to be further explored are presented.