Research And Application On Robust Filtering Method Of Networked Control Systems With Uncertain Parameters

Networked control systems(NCSs) cover various technical fields like control technology,computer network technology and information technology, and have found fruitful applications in a broad range of areas such as aircraft control systems, robot control systems and industrial control systems. Network can brought convenience and reliability to the control system, however, many special problems are also brought to the analysis and design of the system. Because of the insertion of communication network, network-induced delay, data packet dropout, quantisation and other unfavorable factors which affect system performance should be considered during transmission of signals. In addition, changes in the actual project environment, the presence of external interference, will lead to uncertainty of the controlled object in the networked control systems. Therefore, the research of networked control systems with parameter uncertainties has certain theoretical significance and practical value.Firstly, the mathematical model of norm bounded parameter uncertainty networked control systems with random delays is constructed, and the H_∞ filter design is based on this model. random variables are considered to be introduced to describe random network-induced delay which satisfy Bernoulli distributed sequence. By selecting the delay range associated Lyapunov functional and incorporating with integral inequality processing method, the delay dependent sufficient condition for guaranteeing the filtering error system to be asymptotically stable in mean square with a prescribed H_∞ performance is obtained. Then, the solving method of the filter parameters is proposed.Secondly, the robust filtering algorithm for linear parameter-varying systems under networked environment is investigated. By developing the parameter-dependent Lyapunov functional and using parameter linear matrix inequalities technique, the sufficient condition for guaranteeing the filtering error system to be asymptotically stable with a prescribed H_∞and2L L￥- performance is obtained. The desired filter parameter matrices are obtained by solving a class of convex optimization problem in the form of parameterized linear matrix inequalities, which is transferred from the sufficient condition.Then, the robust quantized filering problem for NCSs with random packet dropout and discrete distributed delay is investigated. the process of random packet dropout is described by Bernoulli random distribution, by using the Sector bound method to deal with the quantization error which is transformed into parameter uncertainty. According to Lyapunov functional, the sufficient condition for guaranteeing the filtering error system to be asymptotically stable in mean square with a prescribed H_∞ and2 L L￥- performance is obtained, and filter parameter matrices are obtained by solving a convex optimization problem in the form of LMI.Finally, this chapter is concerned within the context of actual aerospace engineering, the wireless network is introduced into the rendezvous and docking system, and give full consideration to the parameter uncertainty in the rendezvous and docking system model with non Gauss noise input, the improved C-W equation is used to describe the dynamic model of the rendezvous and docking system. Based on the reconstruction dynamic model, the robust filtering strategy for rendezvous and docking system with continuous-time and discrete-time cases are discussed. The relative position and velocity estimation error of spacecraft rendezvous and docking is obtained by numerical simulations, which realizing the robust filter design taking the background of the project.