Grid roadmap based ANN corridor search for collision free, path planning

The probabilistic roadmap (PRM) is a force for path planning in static environments. Altogether, a small change in obstacle position may lead to the regeneration of a new road map, invalidating colliding nodes and edges, and then causing a search for a new collision free path. These steps take a considerable amount of time for processing. In this article, a new method, based on the grid roadmap (GRM), is proposed to mitigate the desired time of path planning. By the suggested roadmap and implementation of an Artificial Neural Network (ANN) for a corridor search, a fast path planning method is achieved, which operates on static, dynamic and unseen environments.