Stability Analysis And Filtering Of Time-Delay Systems

In recent years,the development of science and technology has promoted the progress in the fields of power,aviation,communications and manufacturing,which formed various large-scale and complex-structured systems.This kind of systems generally has the phenomenon of time delay and external interference,which leads to the degradation of systems performance and the deviation of the received measurement signal even of instability.In order to solve such problems,stability analysis and filtering can be performed on the time-delay system.Among the many time-delay systems,there are two special time-delay systems.The first is the neutral time-delay system,which has the time-delay term in the state variable and its derivative,so it is more complex than the general time-delay system and has more practical research value.The second is the network time-delay system,which describes the phenomenon of time-delay caused by network intervention and communication resource limitation in the actual network environment.It is very important in the industrial production of gradually networked and intelligent.Taking of the above factors into consideration,this paper takes neutral mixed time-delay systems,nonlinear network time-delay systems and aircraft network time-delay systems as research objects.Using the Lyapunov stability theory and linear matrix inequality(LMI)technology,the progressive stability of the filter or fault detection filter error system under the H? performance index is proved,and these three categories are solved.A series of problems existed in the time-delay systems,such as mixed time-delay,partially unknown transfer probability,packet loss and sensor failure.The main results achieved in this:(1)The problems of uncertain parameters and partially unknown Markov transition probability for neutral mixed time-delay systems are studied.In combination with Lyapunov stability theory,LMI technology and integral inequality method,the requirement for upper and lower bounds of time delay is relaxed and the computational complexity is reduced.Finally,the reliability of the design is verified by numerical simulation.(2)Research on nonlinear network control systems with time-varying delays has problems such as network packet loss,sector boundedness,waste of resources,and sensor failure.Firstly,the single-channel packet loss phenomenon described by Bernoulli distribution,an event trigger mechanism is proposed to reduce the network occupancy rate.Secondly,an H? fault detection filter is designed,and the stability analysis of the corresponding fault filtering error system is performed.Finally,two examples of spring system and motor servo control system are used to further verify the effectiveness and practicability of the design.(3)Research on the network control system of the aircraft has the problems of multi-channel transmission packet loss,partially unknown transfer probability,waste of resources and sensor failure.Firstly,aiming at the multi-channel packet loss phenomenon described by Markov chain,an event trigger mechanism is proposed to reduce the network occupancy rate.Secondly,the system's sensitivity to faults is improved by weighting faults,combined with cone complementarity linearization(CCL)algorithm design.The H? fault detection filter is obtained.Finally,when the transition probability is partially unknown,the simulation obtains residual fault information,trigger weight matrix and trigger interval,etc.which further verifies the reliability of the design.