Research On Robust Consensus For Multiple Euler-Lagrange Systems With Uncertain Parameters

The consensus problem of multi-agent systems,as the most basic and important problem of distributed cooperative control,has received wide attention from scholars in various field,whose relevant theoretical results have been widely used in industrial applications.However,due to the complex system dynamics in real engineering applications,Euler-Lagrange system models can describe the dynamic processes more scientifically than the general linear models,and it is very important to introduce the dynamics described by Euler-Lagrange equation into the cooperative control research of multi-agent systems,so research on the consensus of multiple Euler-Lagrange systems(MELSs)is not only of theoretical value,but also of great significance to engineering applications.At present,the results on consensus of MELSs are very limited.Thus,in this dissertation,based on graph theory and cooperative control theory,with adaptive robust control as the core tool,combined with sliding mode control,event-triggered control,observer-based control,and so on,the robust consensus of MELSs is investigated under the cases of uncertain system parameters,input disturbance,communication link faults(CLFs),incompletely detectable target system,discontinuous communication,and so on.The main contents are as follows:Research on leader-follower robust consensus of MELSs with uncertain parameters subject to CLFs.For the problem of CLFs,by applying the adaptive control method,an adaptive observer with resilient to CLFs is designed,under which the followers can accurately estimate the system matrix,output matrix and state information of the leader.Then,considering the uncertain parameters and input disturbances of follower systems,based on the ideal of sliding mode control,a resilient observer-based robust consensus protocol is proposed and shown to achieve leader-follower consensus for MELSs.Research on distributed observer-based leader-follower robust consensus of MELSs with uncertain parameters under the cases of CLFs and incompletely detectable target system.Based on the adaptive control and detectability decomposition methods,an adaptive distributed observer is designed,where the detectable part is updated with the information observed by itself,while the undetectable part is updated by neighbors.Then,the fulldimensional state of target system can be accurately estimated.Moreover,according to the sliding mode control strategy,a distributed observer-based robust consensus protocol is constructed,and it is proved that the leader-follower consensus of MELSs can be achieved.Research on leader-follower robust consensus of MELSs with uncertain parameters and an incompletely detectable target system via distributed reduced-order observers.According to the adaptive control strategy,idea of designing distributed observer and Luenberger observer theory,distributed reduced-order observers with adaptive coupling gains and fixed coupling gains are designed under the directed fixed graph of CLFs and the jointly connected undirected switching graph,respectively.Then,by using the sliding mode control strategy,based on the above two distributed reduced-order observers,robust consensus protocols are constructed,and it is proved that both the proposed protocols can achieve leader-follower consensus.Research on leader-follower robust consensus for MELSs of an uncertain leader.Under the undirected fixed graph of CLFs,in view of an uncertain leader,that is,no follower can obtain the dynamic parameters of leader system,by using adaptive control strategy,observers of node-based and edge-based adaptive coupling gains are respectively designed.Then,for the follower system of uncertain parameters and input disturbances,based on the ideal of sliding mode control,node-based and edge-based adaptive observer-type robust consensus laws are proposed respectively.Moreover,it is shown that the MELSs can achieve leader-follower consensus.Research on event-triggered robust consensus of MELSs with uncertain parameters on undirected fixed graph.By introducing the event-triggered control and adaptive control strategies,two adaptive event-triggered robust consensus protocols are respectively proposed to solve the leaderless and leader-follower consensus problems of MELSs,where the event-triggered function of each system is only related to its own state,and the system sends information to neighbor systems only when the given event-triggered condition is satisfied.Then,for these protocols,by constructing suitable Lyapunov functions,it is proved that the MELSs can achieve leaderless consensus and leader-follower consensus,respectively,and there will be no Zeno behavior during the whole control processes.