实时双向群组搜索及其在蚁群觅食动画中的应用

Real Time Bidirectional Crowd Search and the Application in Ant Colony Foraging Animation

基于双向搜索和群组协作的研究,提出一种新颖的搜索算法--实时双向群组搜索(real-time bidirectional crowd search,简称RBCS).基于这个搜索算法所提出的蚁群觅食模型包含了有限状态机和一系列表示蚂蚁和环境交互的规则,具有在复杂动态环境下找到食物和巢穴之间最短路径的能力.2D/3D实验结果表明,算法的搜索能力具有可信性,将其扩展到人群在固定点之间的来回往复运动也获得了满意的效果;和传统基于信息素的蚁群觅食模型的仿真实验对比表明了算法的优越性.

This paper proposed a real-time bidirectional crowd search（RBCS） approach based on real time bidirectional search and crowd collaboration.The collaboration mechanism is inspired by the foraging behaviors of an ant colony.We present an ant colony foraging model based on RBCS and it consists of a finite state machine（FSM） and a serial of rules defining the interactions between ants and environment.The model has the ability of adapting to dynamically changing environment to find the shortest route.Experimental results demonstrate the power of our approach：ant colony behaves reasonably in different environments.Extension to human group moving between fixed positions time and again acquires approving result.And comparison with traditional ant colony foraging model based on using pheromone shows the superiority of this method.