Containment Control Of Hybrid Multi-Agent Systems

In recent years,the coordinated control of multi-agent systems has received extensive attention of many scholars.Containment control is a basic issue of coordinated control,which aims to drive followers into the convex hull spanned by leaders by designing some effective distributed protocols and utilizing distributed network information.In the exploration of containment control,the dynamic models of agents have attracted considerable attention,which include first-order,second-order,continuous-time,discrete-time,and so forth.The development of computer science and robotics provides conditions for the interaction between continuous-time and discrete-time individuals.Therefore,the containment control of hybrid multi-agent systems composed of continuous-time and discrete-time dynamic agents is investigated in this thesis.The main contributions are as follows:1.The containment control of two types of hybrid multi-agent systems with first-order dynamics is studied.When followers have first-order continuous-time dynamics and leaders have first-order discrete-time dynamics,three efficient distributed protocols are designed by analyzing the interaction modes between followers and leaders.By means of Gers?gorin circle theorem and stability theory,some necessary and sufficient conditions are given to solve the containment control.When followers have first-order discrete-time dynamics and leaders have first-order continuous-time dynamics,an efficient distributed protocol is devised,and the sufficient and necessary criteria to achieve containment control are presented.Finally,the effectiveness of the proposed protocols is verified by some simulations.2.The containment control of hybrid multi-agent systems composed of second-order and first-order dynamics is investigated.When followers have second-order continuous-time dynamics and leaders have first-order discrete-time dynamics,three efficient distributed protocols with absolute velocity information are proposed.For the first two,all agents with different dynamics can be transformed into first-order discrete-time dynamic agents through the system transformation method,so as to obtain the conditions related to the feedback gains and the degree matrix.For the third,the necessary and sufficient conditions for the system to achieve containment control are established by constructing the error system and analyzing its stability.An efficient distributed protocol with absolute velocity information is proposed when followers have second-order discrete-time dynamics and leaders have first-order continuous-time dynamics.By adopting the bilinear transformation method,the Schur stability of the discrete-time system is transformed into the Hurwitz stability of the continuoustime system,so that the necessary and sufficient criteria to achieve containment control are derived.Finally,some simulations are provided to demonstrate the validity of the theoretical results.3.In practice,velocity and acceleration measurements are more difficult to acquire than position measurements.Therefore,an efficient distributed protocol using only position measurements is proposed to achieve containment control of the hybrid multi-agent system,consisting of second-order continuous-time followers and first-order discrete-time leaders.A new variable is introduced in the protocol to estimate the velocity information of the followers,and the derivative of this variable is only related to the position information of itself and all agents.By employing the Lyapunov direct method and La Salle’s invariance principle,sufficient conditions are established to solve the containment control,which are related to the communication topology and feedback gains.Finally,the effectiveness of the proposed protocols is verified by some simulations.