With the growing reliability of modern Ad Hoc Networks, it is encouraging to analyze potential involvement of autonomous Ad Hoc agents in critical situations where human involvement could be perilous. One such critical scenario is the Search and Rescue effort in the event of a disaster where timely discovery and help deployment is of utmost importance. This paper demonstrates the applicability of a bio-inspired technique, namely Ant Algorithms (AA), in optimizing the search time for a near optimal path to a trapped victim, followed by the application of Dijkstra's algorithm in the rescue phase. The inherent exploratory nature of AA is put to use for a faster mapping and coverage of the unknown search space. Four different AA are implemented, with different effects of the pheromone in play. An inverted AA, with repulsive pheromones, was found to be the best fit for this particular application. After considerable exploration, upon discovery of the victim, the autonomous agents further facilitate the rescue process by forming a relay network, using the already deployed resources. Hence, the paper discusses a detailed decision making model of the swarm, segmented into two primary phases, responsible for the search and rescue respectively. Different aspects of the performance of the agent swarm are analyzed, as a function of the spatial dimensions, the complexity of the search space, the deployed search group size, and the signal permeability of the obstacles in the area.
翻译:随着现代特设网络的可靠性不断提高,令人鼓舞的是分析自治特设人员在人类参与可能十分危险的危急局势中的潜在参与,这种关键情景之一是在灾害发生时进行搜索和救援工作,及时发现和帮助部署至关重要,本文件表明生物启发技术,即Ant Algorithms(AA)在优化搜索时间以找到被困受害者近乎最佳的道路方面最适用,然后在救援阶段应用Dijkstra的算法。AAA的内在探索性质被用来更快地绘制地图和覆盖未知的搜索空间。实施了四种不同的AAA,其作用不同,即地球的及时发现和帮助部署最为重要。在发现受害者后,自主代理进一步便利了救援进程,利用已经部署的资源组成了一个中继网络。因此,文件讨论了一个详细的决定模型,将其分成两个主要的搜索空间阶段,分别负责搜索空间的搜索和搜索空间的方位,分别负责搜索方位的搜索和搜索方位的功能。