Research on Consensus Problem of Swarming Behavior of Scalable Intelligent System

Aiming at consensus problem of system comprised of vast intelligent individuals, we study its swarming behavior based on three kinds of initial distribution: 1) initial distribution that all individuals can find each other; 2) initial distribution that an mutual observable chain exists in any two individuals; 3) initial distribution that some individuals can not find each other directly or indirectly, and then an individual motion equation is proposed to improve the consensus of swarming behavior, which is decided by information of some characteristic neighbor individuals and virtual leader individuals specified in advance in the intelligent system. Scalable swarming behavior can be achieved without global information of all individuals based on the individual motion equation.