Path planning for multiagent system in a sensing field with obstacles and multiple base stations

Abstract

In large area data collection, wireless long-distance data transmission through sensor network reduces network lifetime due to large energy consumption. Therefore, this paper presents a path planning problem for a multiagent system to periodically visit a set of sensor nodes in a sensing field with obstacles while minimizing the data collection time. During data collection, all sensor nodes are visited in each period exactly once by the robots. At the end of one period the collected data is uploaded to one of the base stations. This paper proposes a new approach for constructing clusters and a visiting sequence of nodes, that each robot must visit. To design the clusters and the visiting sequence of nodes a new algorithm based on Ant Colony Optimization is developed. The path planning algorithms from one and multiple base stations are created. Simulations are conducted in various sensing fields with different numbers of agents, measurements are defined, and the results are analyzed. The created algorithms can be utilized for example, to monitor weather phenomena in an agricultural area or reconstructing the path of an intruder in a guarded area.