Genetic-based Mobile Robot Path Planning using Vertex Heuristics

This paper presents a new genetic-based path planning method for a mobile robot in which the vertices of the obstacles, described as polygonal clusters, are encoded as genes. In comparison with other approaches that either map the robot's environment or need to consider the whole environment as the search space, the narrowing of the search to include only vertices is able to gear evolutionary progress, bringing potential benefits in terms of both calculation time and memory usage. The proposed vertex planning method is verified through experiments in a series of realistic simulated environments. Also, the efficiency and effectiveness of the proposed method are compared with the well-known evolutionary navigator/planner