Dynamic Modeling, Stability Analysis And Simulation On Collective Behavior Of Intelligent Swarm Systems

Swarming is the ubiquitous phenomenon in nature. This phenomenon is thehighly abstracted of the orderly collective behavior, which is emerging from socialbiological population and man-made population in the real word. Flock is made up bythe interaction individuals with enormous number. The intrinsic character of flock isthat the intelligent individual has little or no intelligence at all. This kind ofintelligence it does is spontaneously emerging from bottom to top via the interactionsamong individuals. It shows the internal reason of biologic evolution, the mechanismof distributed coordination and cooperation among intelligent agents, and so on.Intelligent swarm system is essentially a kind of complex adaptive system.Collective intelligent control study the motion behavior’s soft controlling ofintelligent swarm system. Meantime, the multi-agent system with a lot of intelligentagents via the mechanism of distributional coordination and cooperation can realizethe control on complicated system. The control problem of multi-agent system linkswith the swarming motion control problem of intelligent swarm systems in natureinherent. The objective of the control problem of multi-agent’s distributionalcoordination and cooperation is multi-agent coincident with the consensus, which iswell coincide with the research objective of collective intelligent control system. Inrecent years, the studying emphasis in consensus problem is studying the concreteproperties of the consensus mechanism’s intelligent swarm engineering applications.Some relative problems are analyzed here respectively, such as synchronization,formation problem, swarming and flocking, rendezvous problem, and so on.Aiming at the motion behavior’s soft controlling of intelligent swarm systems, themapping relation is built up to reflect the relations between the emergency mechanismof social biological population’s collective behavior and consensus problem in thispaper. The group’s motion law is be drawn from the former by taking the mappingrelations as bond, and use it as a tool to be used in dynamic modeling and soft controlon coordination behavior of intelligent swarm systems. And then unifies the mainlinks’ require of the consensus mechanism’s applications, the soft control strategydesign and simulation are studied. In which, the main links refer to such motionbehavior as collision-avoidance, obstacle-avoidance, formation configuration,trajectory tracking, coverage search, and so on. The main links are extended frommigration or flocking behavior. Being focused on this topic, the main work in thispaper is as following:(1) Drawing lessons from the perception and interaction capabilities of socialbiological, the basic frame of the motion behavior analysis of swarming system isestablished, the mapping relation is given to reflect the relations between the dynamiccharacteristic of collective behavior and consensus problem. It is revealed that there is an internal relation between coordination control of intelligent swarm systems andconsensus application mechanism.(2) The collective behavior of biological population which implicates thepotential interior operational principle. For the convenience of the analysis on motionlaw of swarm systems, three kinds of the dynamic models of different types onintelligent swarm systems are constructed respectively based on the two differentframeworks which are Lagrangian framework and Eulerian framework, and which aresubordinate to the spatial approach is adopted while modeling in this paper.Meanwhile, the detailed analysis and description of these three kinds of models hasbeen done respectively which is aiming at the difference of the perceptionenvironment’s condition and considering under different situation such as finiteperception, global perception, isotropy, anisotropy, stochastic disturbance, withtime-delay, without time-delay, and so on. And then the collective behavior issimulated and tested with the two kinds of Lagrangian models which are the rangelimit perceived swarms model and exponential type stochastic swarms model basedon the analyses of the control rules for the two kinds of Lagrangian models underdifferent situation in this paper. The correctness and validity of the two kinds ofLagrangian models has been verified by simulation.(3) Theoretically, it has been proved that the aggregation and stability on the threekinds of swarm systems models which are the range limit perceived swarms modeland the exponential type stochastic swarms model and the cooperative diffusionswarms model from control system stability analysis, and in addition to this, thesimulation and testing on the soft control performance of the two kinds of swarmsmodels which are the range limit perceived swarms model and exponential typestochastic swarms model is done which is aiming at the consensus applicationsproblems during the soft control on coordination behavior of intelligent swarmsystems such as obstacle-avoidance, formation configuration, noise suppression,trajectory tracking, coverage search, and so on. The simulation result confirms the twokinds of intelligent swarm systems models which all posses better stability.(4) Aiming at the particular requirement of coordination and cooperationproblems on the brine extraction and monitoring process with multiple biomimeticrobotic fishes system as well as some basic problems in salt field, the basic frameworkof the control architecture of multiple bionic robot fishes system is constructed. Someproblems bearing relationship with it have been discussed, such as control schemedesign, research objective, technology route, and so on. The feasibility demonstrationis done yet simultaneously on the track navigation strategy and coordinative operationmechanisms which have been established in this paper.