Research On Event-Triggered Fault Detection And Fault-Tolerant Control Of Networked Nonlinear Systems

Networked nonlinear systems have been extensively employed in production and life due to their low cost,strong expansibility,and easy maintenance.Since channel bandwidth,communication frequency,and node energy are usually limited,to ensure the successful implementation of cooperative tasks between network nodes,it is necessary to introduce efficient data transmission and interaction mechanisms.The event-triggered scheme is one of the research hotspots in control and communication communities in recent decades.It can effectively reduce the network burden while maintaining system performance.On the other hand,with the increasing scale and complexity of control systems,potential malfunctions of system components or subsystems may lead to instability of the entire system,and even seriously endanger personnel safety of personnel and cause substantial economic losses.Therefore,to realize efficient utilization of resources and smooth operation of networked nonlinear systems,this paper studies the problem of fault detection and fault-tolerant control of networked nonlinear systems under eventtriggered transmission protocols.Specifically,the main research contents of this paper are as follows.(1)The problem of finite frequency fault detection for a class of nonlinear systems with Lipschitz nonlinearities,parameter uncertainties,unknown faults,and bounded disturbances is explored.First,a finite-frequency nonlinear fault detection filter based on an adaptive event-triggered scheme is proposed,and the augmented filter system is transformed into a linear parameter-varying system through the reformulated Lipschitz characteristic.Then,two sufficient conditions are derived to ensure that the linear parameter-varying system has a finite-frequency H_fault-sensitivity performance index and a full-frequency H∞ disturbance-attenuation performance index.Besides,the obtained sufficient conditions are further organized into linear matrix inequalities to solve the filter parameters.(2)A finite frequency fault detection scheme based on an event-triggered mechanism is considered for discrete-time Markovian jump systems with Lipschitz functions,timevarying delays,and packet loss.First,an event-triggered fault detection filter subject to data loss is developed,and the augmented filter system is arranged into a linear parameter-varying system through the reformulated Lipschitz property.Then,two new theorems are proposed to guarantee that the linear parameter-varying system is robust to unknown disturbances and sensitive to finite frequency faults.Moreover,the proposed theorems are transformed into linear matrix inequalities using Finsler’s lemma and Sprocedure.(3)The problem of finite-frequency distributed fault detection for discrete-time nonlinear switched systems with time-varying delays is studied via sensor networks.First,the switching mechanism is described using the sojourn probability approach.Also,a distributed memory event-triggered scheme is employed to reduce the network bandwidth load,and the impact of malicious deception attacks is considered and analyzed.Furthermore,a finite-frequency distributed fault detection filter affected by memory event triggering scheme and cyber attacks is developed for each sensor node.And then,some sufficient conditions are provided to ensure that the augmented filter system is meansquare stable and satisfies the specified H_/H∞,performance indexes,and the desired distributed filter gains are obtained by solving the linear matrix inequalities.(4)A distributed fractional-order active fault-tolerant control paradigm based on an event-triggered strategy is constructed for multiple unmanned aerial vehicles subject to full-state constraints.First,a distributed fault diagnosis scheme based on H∞,technology is designed to realize fault detection and fault estimation.Then,by exploiting the fault information obtained from the fault diagnosis module,a distributed fractional-order active fault-tolerant control protocol is constructed to achieve accelerated convergence within a predefined finite time,and to guarantee that the time-varying asymmetric state constraints are not violated after the system has been running for a period of time.At the same time,an event-triggered mechanism is introduced to reduce the update frequency of control signals.It is verified that all signals of each follower unmanned aerial vehicle are semi-globally uniformly ultimately bounded,and all followers can follow the leader’s attitudes.(5)The problem of distributed fault-tolerant control of multiple unmanned aerial vehicles affected by full-state prescribed performance is discussed.First,a distributed event-triggered observer is designed to reconstruct the leader’s states in finite time,and a Zeno-free dynamic triggering condition is proposed.Next,based on the proposed observer and backstepping control techniques,a distributed fault-tolerant controller is constructed to track the estimated leader states.Additionally,a finite-time performance function is introduced to ensure the tracking error converges to a specified region within a bounded time.The Lyapunov theory states that all closed-loop signals are practical finite-time stable,and the prescribed performance requirements are strictly obeyed.