Car-Like Robot Path Planning Based on Voronoi and Q-Learning Algorithms

This paper discusses a differential path planning issue for the mobile robot depending on Voronoi diagram (VD) and Q-learning algorithms (QL). The issues with re-arranging paths in a dynamic environment with obstructions are treated as an issue of looking for the best route between the start and target stage. Since the car-like robot is differentially system and its mathematically involved with the inward state of impediment because of that some modification will be embedded like the orientation method. This is a unique instance of a solitary vehicle, which just goes ahead at a consistent speed and can just turn left also, right. Voronoi diagram presents the world encompassing brilliant specialists for robots, computer games, and military issues, so improving the dependability, of arranging the environment using it will help and decrease Q learning calculations by re-updating the Q-table, according to the new state, also decreasing both existence intricacy, relies upon earlier information that came from the environment. The work arrangement was tested for a 2D environment. The result of the proposed work showed better performance in time, speed, and length of the path also it can be utilized as an alternate style of guides. A comparison with other related works is performed and the result of these comparisons showed that our work provides a good trajectory with performance.