Swarm Robotics and Rapidly Exploring Random Graph Algorithms Applied to Environment Exploration and Path Planning

Abstract

We propose an efficient scheme based on a swarm robotics approach for exploring unknown environments. The initial goal is to trace a map which is later used to find optimal paths. The algorithm minimizes distance and danger. The proposed scheme consists in three phases: exploration, mapping and path optimization. A cellular automata approach is used for the simulation of the fist two phases. For the exploration phase, a stigmergy approach is applied in order to allow for swarm communication in a implicit way. For the path planning phase a hybrid method is proposed. First an adapted Rapidly-exploring Random Graph algorithm is used and then a scalarized multiobjective technique is applied to find the shortest path.