Font Size: a A A

Fault Detection For Discrete Stochastic Systems And Its Application In Networked Environments

Posted on:2018-08-21Degree:DoctorType:Dissertation
Country:ChinaCandidate:Y Q WangFull Text:PDF
GTID:1360330593950000Subject:Oil and Natural Gas Engineering
Abstract/Summary:PDF Full Text Request
With the increasing demands of reliability and safety requirements for control systems,fault detection problem becomes the research hot spot in control field.Fault detection is to identify possible fault using the measurable information in control system.The rapid development of communication and network technology enables information in the system to be transmitted through network,however,due to the limitation of channel bandwidth in the network environment,network induced phenomena occur inevitably,such as network congestion,transmission delay,data drop packet,signal quantitation,channel fadings and randomly occurring nonlinearity.At present,people have not fully understood the occurrence of these network induced phenomena,and their impact on fault detection has become a difficult problem to be solved.Under this background,models of nonlinear stochastic systems with network induced phenomena in network environment are constructed,and the corresponding fault detection problems are studied.Aiming to the fault detection problem of networked systems with successive packet loss under networked environments,a model under consideration of parameter uncertainty and randomly occurred nonlinearity is set up.Based on this,the fault detection problem is studied.A sufficient condition for the existence of the fault detection filter is obtained according to Lyapunov stability theory and stochastic analysis techniques and the fault detection problem is solved by transforming into convex optimization problem.Non-fragile H_?filtering problem for Markovian jump system with mode-dependent time delays and quantization is studied.Mode switching is considered not only in the system parameters but also in the time delays.Stochatic variables and norm bound uncertainty are utilized to indicate the filter gain variation.A non-fragile H_?filter is designed which ensures that the filtering error dynamics is stochastically stable with a prescribed H_?performance and the filtere parameters are robust.The fault detection and fault estimation problem is investigated for scochastic delayed systems with channel fadings.Fault detection filter is designed for a class of discrete-time stochastic systems with channel fadings,randomly occurring multiple communication delays,and infinitely distributed delays so that the filtering error dynamics is exponentially stable in the mean square with a prescribed disturbance attenuation level.Moreover,in order to solve the fault estimation problem for a class of stochastic delayed systems subjected to channel fadings,by constructing new Lyapunov function,the sufficient condition for the existence of the fault estimator is obtained through stochastic analysis techniques.After the feasible solution of LMI is solved,the expression of desired estimator gain is acquired.The fault detection problem for T-S fuzzy system with channel fadings and randomly occurred multiple time-delay is studied.Using a series of mutually independent random variables which obey to Bemoulli distribution to describe the randomly occurred properties of multiple time-varying delays.The sufficient condition for the existence of the fault detection filter is obtained which ensures the fault detection error dynamics is exponentially stable in the mean square.Based on the above result,the T-S fuzzy model is used to approach dimensionless mechanism model for continuous stirring reactor system with time-delay,uncertainty and channel fadings and the fault detection problem is studied.This is an attempt of model-based fault detection method in engineering applications and new theoretical guidance is provided for engineering practice.The fault detection scheme for a class of time-varying dynamic systems is proposed.Randomly occurring uncertainties,randomly occurring nonlinearities,quantization effects and packet dropouts is considered.The influence of the above network induced phenomena on the performance of the fault detection dynamic system is analyzed deeply in the process of signal transmission.A kind of fault detection filter algorithm is put forward which is based on recursive matrix inequality and the filtering error system satisfies the given H_?performance...
Keywords/Search Tags:Fault detection, Stochastic systems, Networked-induced phenomena, H_? filtering, Recursive matrix inequality
PDF Full Text Request
Related items