Research On Fault Detection Methods For Neutral Delay Systems Under Network Environment

The intersection between network technology and modern industrial production processes is deepening,which makes many control system research fields become more intelligent and complicated.As a kind of widely representative system,many practical industrial production processes can be described by neutral delay system,such as distributed network lossless transmission,two-stage dissolving tank,turbojet engine,etc.,so the research on this system has attracted extensive attention in recent years.However,under the limitation of communication resources in the network environment,system data is prone to delay and data loss in the transmission process,which leads to system failure,these faults inevitably affect system security and reliability.Since the 21st century,with the rapid development of various industrial systems,higher requirements have been put forward for them,it is important to study the fault detection technology and detect the fault in advance for the safe and stable operation of the system.As an effective means to improve system performance,fault detection technology has become a key research topic in recent years.In this paper,the fault detection problem of neutral delay system under network environment is studied.The main contents include the following aspects:Firstly,for the neutral delay system in the network environment,considering the delay and packet loss phenomenon of the system,H_∞fault detection filter based on observer is established,the packet loss problem of the system in the process of information transmission is described by Bernoulli distribution,and the stability of the system is analyzed by LMI technology and Lyapunov stability theory.On the other hand,bernoulli distribution is also used to describe the probability distribution of time delay in different intervals.By constructing delay-dependent Lyapunov functional,a performance criterion is established to make the system mean square asymptotically stable with L₁ noise suppression level.Numerical simulation results show the feasibility of this design method.Secondly,considering the neutral delay system with distributed delay and sensor saturation constraint,the fault detection problem of the system with two kinds of different input noise is studied.By using Schur lemma and LMI technology,the design problem of fault detection filter is transformed into the optimization problem of H_∞and L₁ performance,and the fault detection filter is constructed to generate residual signal.Then the residual evaluation function value is compared with the preset threshold value to judge the occurrence of system faults.Thirdly,considering the combination of Markov jump parameters and neutral delay systems,the H∞fault detection problem is studied.By using Ito differential rule and Newton-Leibniz formula,a sufficient condition for the existence of a fault detection filter which makes the filter error system randomly stable and satisfies H_∞performance index is proposed.Finally,numerical simulation shows that the proposed method is feasible and effective.Finally,for the neutral delay system with distributed delay,an event-triggered mechanism is adopted to reduce the transmission of redundant data and the unnecessary waste of resources,the problem of L₁ fault detection for neutral delay systems with event-triggered mechanism is studied.A fault weighting model is introduced to improve the detection performance of the system,and the asymptotic stability condition of the fault detection system is obtained by using Lyapunov stability theory,which is proved by numerical simulation.