Robust Fault Identification Research Of Uncertain Time-delay Systems

Fault detection and diagnosis provides a guarantee for the reliability and security of the modern complex system.At the same time,the time-delay exists widely in the actual system,and the existence of time-delay often leads to the instability of actual control system.Also the system may contain structural parameter uncertainty and model uncertainty caused by external disturbance,noise or model mismatching.So the model combining uncertainty and time-delay is closer to the actual system,which can truly reflect the characteristics of complex system.Therefore,in this paper,robust fault identification for the model including time-delay and uncertainty is researched.Firstly,a novel approach to study the game-theoretic estimation problems for linear continuous time with state delay or output delay is proposed.For different models,the optimal solution and state estimation equations are given.The saddle point is the optimal solution.At the same time,the design method of the estimator satisfying the boundary conditions and Riccati equations is given.By an example,it is demonstrated that the proposed estimator is valid to estimate state and unknown input.What's more,a sensor fault diagnosis approach is presented for a class of time delay nonlinear systems via the use of adaptive updating rules.The considered system is represented by time-varying delayed dynamical state space model,and the system is subjected to a nonlinear vector representing the modelling uncertainty.Firstly,a fault detector observer is constructed to detect fault.Meanwhile,the way to choose the threshold value is given.Further,a fault diagnosis device is constructed to diagnose the fault.The Lyapunov stability theory is used to obtain the required adaptive tuning rules for the estimation of sensor fault.An adaptive diagnostic algorithm is developed to obtain the information of the sensor fault.Finally,a simulated numerical example is included to demonstrate the use of the proposed approach,and the experimental results show that the proposed adaptive diagnosis algorithm can track the fault signal and the proposed method is valid.Finally,this paper studies the problem of actuator fault estimation for linear continuous systems,which is subjected to time-varying interval delay and norm-bounded external disturbance.Based on the fast adaptive fault estimation(FAFE)algorithm,our attention is focused on designing fault estimation filters which guarantee the filtering error system to be asymptotically stable with a prescribed H? performance.A delay-dependent criterion is established to reduce the conservatism of designing procedure.A novel Lyapunov-Krasovskii functional is employed,which includes the information of the upper and the lower bounds of the time delay.An improved sufficient condition for the existence of such a filter is established in terms of the linear matrix inequality(LMI)by the Schur complements and the cone complementary linearization algorithm.In addition,the results for the systems with time-varying interval delay are simplified when the delay is constant.Two numerical examples are presented to illustrate the effectiveness of the proposed algorithm when the delay is time-varying or constant.