Font Size: a A A

Research On Finite-Time Consensus Of Multi-Agent Systems

Posted on:2020-09-22Degree:MasterType:Thesis
Country:ChinaCandidate:K LiFull Text:PDF
GTID:2428330602452478Subject:Control theory and control engineering
Abstract/Summary:PDF Full Text Request
In recent years,distributed coordinative control theory of multi-agent systems have attracted considerable attention among many experts and scholars around the world.It's not only because of its widespread application and prospects for development in unmanned aerial vehicles,robots,sensor networks,but also due to its many advantages can be well utilized such as robustness,economy,flexibility and so on,which the traditional and centralized control method don't have.The consensus problem,the basic topics in this field,is to design a control protocol under which agents can reach a global agreement by using local information shared among neighboring agents.In the dynamic systems,convergence speed is one of the important performance index for assessing the control systems,how to make the systems have a fast convergence speed has a great value for research in practice application.Therefore,the research on finite-time multi-agent systems gradually be taken seriously.Based on the existing research result,the finite-time consensus problem of multi-agent systems is studied in this dissertation,the main work and contents are as follows:First of all,the research status and the basic principles of multi-agent systems are set forth.The related theory for analyzing multi-agent systems are introduced in a brief.Some representative research results in this field are reviewed.Secondly,a novel finite-time consensus tracking control protocol is designed under undirected communication topology.For the systems with first-order integral dynamics,the control protocol can drive all followers of the system to reach an agreement with the leader state at a preset time.Moreover,the convergence time is completely independent of the initial conditions and other parameters of the system.The effectiveness of the mentioned control protocol is testified by strictly theoretical analysis and proof.Then,the control strategy is extended to the systems with second-order integral dynamics,and the finite-time consensus tracking problem of second-order multi-agent systems is solved.Thirdly,a faster finite-time consensus tracking control protocol of multi-agent systems with undirected and switching communication topology based on sign function,by introducing a distributed linear consensus term,is proposed.Then,the finite-time convergence of systems is theoretically shown by using the Lyapunov function and the La Salle invariant principle.Finally,some numerical simulations with MATLAB are also given to further illustrate the effectiveness of the obtained results.
Keywords/Search Tags:multi-agent systems, consensus tracking, finite-time, convergence speed, switching topology
PDF Full Text Request
Related items