Consensus And Formation Control Of Multi-agent Systems With Constraints

Multi-agent systems have been greatly impressed by researchers due to its applications in computer network, artificial intelligence, communication engineering, formation control, and search and rescue and so on. Consensus control is not only one of the basic problems of multi-agent research but also is the premise of other studies of multi-agent. However many unpredictable factors that exist in the actual environment often directly affect whether the multi-agent system can achieve the desired consistency. So it is meaningful to make the multi-agent systems achieve expected consensus and formation with constraints.Firstly, the consensus control protocol of general linear multi-agent systems based on the distributed event driven is proposed which is under the constraint of input saturation. The distributed event trigger algorithm we proposed can effectively reduce the number of control updates, reduce the agent’s energy consumption. Further combining the events trigger conditions, the consensus control protocol is put forward with input saturation. Finally the convergence condition of the event trigger control protocol is analyzed by using the method of S-procedure.Secondly, because in the real environment, the communication delay between agents often cannot be avoided, we studied the event-based saturated consensus problem for linear multi-agent systems under the constraint of input time delays. In addition, on the base of the previous text the convergence condition of the event trigger control protocol is analyzed by using the development method of S-procedure. Eventually all the agents can achieve the desired consistency.At last, based on consensus research, combined the coverage control problem of sensor networks, the leader-following coverage formation control protocol of mobile sensors is designed in the process of movement. A gravitational potential function is designed for the virtual leader to get to the target position. Under the constraint of unpredictable leader information, the velocity and acceleration observer-based control law is presented for all mobile sensors tracking the leader and achieving expected formation. In our control law, mobile sensors can form the coverage formation rapidly before achieve the target coverage area which greatly enhance the effect and flexibility of coverage with less mobile sensors.