Research On Finite-time Consistency Control Method Of Multi-agent System Under Constraint And Network Attack

Multi-agent system is a kind of intelligent system with complex communication relationship in the system and individual computing decision-making ability.The system is usually oriented to complex application scenarios to meet specific functional requirements.At the same time,it adopts distributed design ideas to deal with possible system problems.In the scene of formation and cluster which depend on cooperative control,because of the advantages of independent of central control and global communication,cooperative control method based on multi-agent system has been widely used,and consistency control is a particularly important part.Combining with the specific scenario requirements of the consistency control of the multi-agent system,this thesis has carried out the theoretical model research of the multiagent system and designed different non-smooth controllers for different work scenarios,to realize the multi-agent system for different scenarios within a finite time.Finite time consistency control of the agent system.For the case of input saturation and speed constraint,this thesis designs a saturation function and introduces a speed constraint parameter.Based on the theory of homogeneity,a finite-time protocol is constructed to realize the finite-time consistency control of multi-agent system under constraint.For the case that the dynamics of some agents in the system changes due to network attack,the finite time consensus control is realized by switching system modeling,combining the consensus control algorithm and neural network method.The above control algorithms are verified by numerical simulation.At the same time,to solve the problem of large amount of calculation and long execution time of multi-agent system simulation,this thesis has carried out research on multi-agent simulation framework and explored a set of remote multi-agent system simulation scheme based on MATLAB.The main innovations of this thesis are as follows:(1)For the multi-agent system with input saturation and speed constraint,the finitetime consistency algorithm designed in this thesis has faster convergence speed and antiinterference performance than the traditional method and can meet the constraints of input saturation and speed constraint at the same time.(2)For the multi-agent system with dynamic change under network attack,the finitetime consistency algorithm designed in this thesis can still make the system reach agreement quickly in the case of dynamic asynchronous switching of multi-agents,which is more in line with the actual scene than the synchronous switching system.(3)Aiming at the remote simulation needs of multi-agent system,this thesis constructs a set of remote multi-agent system simulation software based on MATLAB.With the help of the tools provided by MATLAB,the software can be easily deployed on the local or server to meet the simulation needs of different users.