Robust Filtering For Distributed Netwoked Control System Based On Local Information

DNCS(Distributed networked control system)is the closed-loop feedback control system whose sensors, actuators, and controllers communicate through a shared network. the control structure of control has reformed fundamentally by the development of the computer communication technology and the control subject. As an outcome of the reformation, DNCS has broad application prospect ,not only in the area of industy and military,but also in Sociology, Medicine and Economics.As NCS become more large-scale,more independent and more intelligent,networked nodes within a limited range can exchange their local information and influence each other to complete a common task,besides performming their tasks independently. On the other hand, as a result of the involvement of communication networks, NCS will be influenced by data missing and noise inevitably.In this paper, based on the application trends of DNCS modeling,the concept of Local Information in artificial intelligence is introduced to modelling a kind of DNCS. The data missing in the communication process between the central control unit and the network node is considered. Robust H2, H∞and L2-L∞filtering algorithm are given respectively for this distributed networked control system, and the relationship between the filtering effect and the success rate of data transmission is preliminarily studied.Finally, a comparison of the applicability of these three algorithms is made simply.The work of this paper focused on three areas: 1) for a class of DNCS in which network nodes influence each other when measuring the presence of feedback interaction within the local area, a model is established considering data missing in the process of information transmission between the central control unit and network nodes; 2) According to the previous modeling analysis, the stability of the argumented filtering error system is analyzed using Lyapunov stability analysis and the conversion technique of the uncertainty factor in linear matrix inequalities .The H₂,H_∞and L₂-L_∞performance indicators are given to design the corresponding filters respectively,. 3) the Matlab LMI toolbox is used to solve the linear matrix inequalities,and experiments are made to similate the designed filters.