MR-APG: An Improved Model for Swarm Intelligence Movement Coordination

Swarm intelligence based on the bionics have become an emerging research field, and researchers have proposed some kinds of intelligent models, such as ant colony, free group wolves, fish flock, insects group, bacteria and bacteria immune etc. The behavior of common features is: the individual behavior and function is simple, but a rally for the swarm after a complex cluster behavior ability. In mobile swarm group individuals within a certain space can keep a certain distance to avoid collision in order to group coordinated movement. In this paper, an attempt has been made to optimize the continuous and collaborative of the swarm, when the swarm nodes in the process of moving and turning. This paper puts forward the Multi-Restrained Artificial Pigeon Group (MR-APG) model based on multiple constraint factors, and introduce the concepts of quantitative coordination of nodes to optimize the nodes “over-dispersion” or “over-assembly” phenomenon. The experimental results show that, the higer α1 value will lead to average movement synchronization ratio μsync of MR-APG model will be lower. The results also show that, comparing with the traditional A/R model (μsync = 73%) and 3D-VFA model (μsync = 84%), the proposed MR-APG method can make μsync equal to 95%, moreover, its only takes half time of A/R model to make μsync equal to 80%.