Grid roadmap based real time path planning

The probabilistic roadmap (PRM) is forceful for path planning in static environments. Also PRM based methods may be employed for real time path planning in dynamic environments. These methods for real time path planning desire vast amount of time for preprocessing. To mitigate desired time for initialization, we propose a new method based on Grid Roadmap (GRM) which is an edge less roadmap. By suggested roadmap and utilizing a training method for robot manipulator, we attain a shape deformed model for obstacles that cancels our ambition for configuration space. Accordingly, grid roadmap construction will be on workspace. Finally, the planner searches for a collision free path in workspace with dynamic and shape changing obstacles.