Event-Triggered Bipartite Synchronization Analysis For Cooperation-Competition Networks

This thesis investigates event-triggered bipartite synchronization of two complex networks,including heterogeneous coupled harmonic oscillators and multiagent systems under false data injection attacks.To alleviate communication burden and save limited resources,two asynchronous event-triggered protocols are proposed.Several sufficient criteria are given to ensure bipartite synchronization by utilizing differential equation theory,inequality theory and Lyapunov stability method.The main contents are as follows:(1)Event-triggered bipartite synchronization of heterogeneous coupled harmonic oscillators is investigated.Considering the communication bandwidth and system resources are usually limited in reality,an asynchronous event-triggered protocol is provided.In addition,owing to the heterogeneities of networks,a fully asynchronous event-triggered protocol is designed,in which each oscillator only needs to be sampled at its own triggering times.Based on the inequality theory and so on,it is proved that heterogeneous coupled harmonic oscillators can reach event-triggered quasi-bipartite synchronization.Meanwhile,several sufficient criteria and the upper bound of quasi-bipartite synchronization errors are provided.(2)Bipartite synchronization of multiagent systems under false data injection attacks is addressed.Regard false data as augmented states and construct an extended observer to compensate the impact of attacks.Then an observer-based asynchronous event-triggered controller and an observer-based fully asynchronous event-triggered controller are proposed.By utilizing Lyapunov stability method and so on,bipartite synchronization of multiagent systems under false data injection attacks is achieved.It is illustrated the impact of false data injection attacks on the system stability by comparing the numerical examples and the feasibility of controllers.(3)Based on event-triggered observers,bipartite synchronization of multiagent systems under false data injection attacks is addressed.To avoid the observer utilizing real-time information,a static event-triggered observer and a dynamic event-triggered observer are designed,respectively.The observers receive the state information from the network at the triggering time.Furthermore,the controller is designed based on the estimated values.Combined with Lyapunov stability method and so on,some sufficient criteria are given to ensure bipartite synchronization of multiagent systems under false data injection attacks.