Cooperative Control Of Multi-agent Systems With Limited Information Mode

Multi-agent systems are modeling and control algorithm for many lager-scalar systems in nature and engineering applications,including biological neural networks,internet,electricity networks system and sensor networks.In these systems,by designing the rule of information exchanging,agent reaches the decision objective of the group by a cooperative way.With growing volume of control information and signal transmutation in communication network of modern control systems,multi-agent systems are inevitably influenced by many factors,such as model uncertainty,instable network environment,faulty and/or limited actuators.With these factors,the design and application of cooperative controllers of multi-agent systems become challenging problem.Therefore,the cooperative control of multi-agent systems with limited information mode is crucial for large-scale complex systems.According to the desired value or set point of the multi-agent systems being determined by collative or particle agents,the cooperative control protocol in our works is divided into tracking cooperative control and regulation cooperative control.We study the protocol design for both tracking and regulation cooperative control under the limited information mode.The main research works and contributions are summarized as follows:1).For limited communication protocol for multi-agent systems,a class of joining consensus tracking cooperative control protocols is proposed via pinning control.Firstly,a weighted average consensus protocol of weighted direct networks is developed to achieve the collective decision of multi-agent systems by interacting with the local state information.Then,a novel joining consensus protocol is designed to realize the intervention of the weighted decision-making for the static and dynamics weighted directed networks of multiagent systems under the desired objective.The design approach of finite control Lyapunov controllers and efficient joining consensus protocol are proposed,the control gain for achieving the coordinate objective intervention is effectively calculated,and consensus tracking cooperative control with limited communication protocol is achieved.2).For limited communications resources for multi-agent systems,the consensus tracking cooperative control protocols with dynamic interaction jointly connected topology and event-triggered mechanism are proposed.Firstly,a concept of mirroring nodes for different roles of the agents in systems evolution and a criterion to determine if a network is of a three-layer network are introduced.By combining the state information of agents in different layers,linear and nonlinear leader-follower consensus protocols are designed with bounded consensus speed.Then,we propose an event-triggered consensus strategy that relaxes the assumptions in existing works that the control gain and local inputs must be continuous,to achieve the consensus tracking cooperative control with limited communications resources.3).For limited actuator health of multi-agent systems,the consensus tracking cooperative control protocols with distributed intermittent communication topology and dynamics fault tolerance mechanism are proposed.Firstly,for actuators with occasional failure inputs and communication resources constraints,a novel distributed intermittent communication framework is proposed via an adaptive approach.By introducing a compensation inputs strategy of actuator fault,a combination of robust consensus tracking protocol is designed by different adaptive feedback controllers.Then,time-varying fault parameters for multi-agent uncertain systems with a directed network topology are considered.The cooperative control law is updated according to the proposed fault tolerant control strategy,to achieve the consensus tracking cooperative control with limited actuator health.4).For limited actuator ability for heterogeneous dynamics networks of multi-agent systems,the synchronization regulation cooperative control protocols with explicit synchronization and robust adaptive explicit synchronization are proposed.Firstly,for the heterogeneous dynamics networks of multi-agent systems with three-layer network structure,we propose an explicit synchronization algorithm.By constructing a three-layer nodes model,an explicit synchronization protocol is designed based on the states of agents in different layers.Further,the design is extended to the cases in which the switching topologies are only frequently but not always strongly connected.Then,we propose the explicit adaptive synchronization protocol that depends on less control parameters and achieves complete synchronization between the agents of heterogeneous uncertain dynamic networks and the virtual target,to achieve the synchronization regulation cooperative control with limited actuator ability.