Research On Consensus Control Algorithm Of Heterogeneous Multi-agent Systems

Multi-agent systems(MAS)are gradually applied in the fields of biology,physics,and engineering,and have become a research hotspot in the field of interdisciplinary science and control.Generally,the macroscopic multi-agent systems consist of a certain number of unmanned vehicles such as drones or robots,and the abstract multi-agent system is composed of network node systems.In the research of the above-mentioned multi-agent system control method,the consensus algorithm has gradually developed into an important multi-agent control method,and it is also fully applied in various control modes in the multi-agent system.In recent years,the control objects of multi-agent systems have gradually shifted from isomorphism to heterogeneity,that is,members of a group are composed of different types of models,instead of being composed of homogeneous systems with the same dynamic model.In practical applications,a large number of occasions require heterogeneous systems,such as sea-air coordination,land-air coordination,etc.Therefore,the consensus algorithm of heterogeneous systems has attracted the attention of a large number of scholars.Due to the differences in dynamics and state spaces among group members,the controller design of heterogeneous distributed consensus algorithms will become very complicated.This paper mainly studies the design of consensus algorithms and the expansion of related algorithms under several types of heterogeneous models.The main research contents are as follows:1.Aiming at the heterogeneous model of unmanned vehicles in different environments,such as the environmental differences of unmanned aerial vehicles(UAV),unmanned ships(USV),and submarine vehicles(UUV)in water environment.Considering the differences in environmental parameters and the dimensions of different motion spaces,the consistency control laws of three heterogeneous models are designed.Based on the high-order and second-order models,the design process of this type of consistency algorithm is deduced.Kronecker product and matrix transformation prove the stability and give some sufficient conditions for parameter selection.2.Consider the optimization problem and obstacle avoidance problem in the forming process at the same time.Aiming at the obstacle avoidance model problem in actual waters,different regions such as circular and mountain threat models are introduced to construct different potential function expressions.After that,the heterogeneous multi-agent system is transformed into a closed-loop error system,the formation control,obstacle avoidance optimization function and index function are defined,and the inverse optimal control algorithm is used to derive the consistent non-quadratic obstacle avoidance optimal control strategy.3.Finally,this article discusses the concept of consistent arrival control in heterogeneous regions.During the execution of the task,the unmanned vehicle needs to reach a designated area instead of a set point.When the error requirements are not particularly high,the final target position of the multi-agent system is limited to one area of the desired point,so the area arrival control has practical significance for heterogeneous formations.Compared with most current research work,the main innovations of this article are as follows:(1)Discuss the heterogeneous consistency problem of different degrees of freedom in two-dimensional and three-dimensional space for the second-order-Lagrangian system model,and prove its stability.(2)For the high-order-second-order integral system model,a consistent algorithm under different motion dimensions is proposed,and the parameter design range is derived;for the underwater obstacle avoidance model problem,the cone-shaped peak threat model is introduced.And use the inverse optimal non-quadratic function optimal control algorithm to ensure the optimal index.(3)Two kinds of unmanned vehicles' cross-medium area arrival control problems are studied,and the effectiveness of the heterogeneous consensus algorithm in the area arrival control is verified.