A Rapidly-Exploring Random Tree Algorithm with Reduced Random Map Size

Abstract

Mobile robots have been widely used in automated factory applications such as raw material delivery and product storage transportation. Path planning algorithms have been proposed to generate a feasible global approach. The path result must be free from obstacle regions and shortest. Previously we proposed an Improved Rapidly-Exploring Random Tree (Improved RRT*) algorithm. The algorithm consists of the pre-processing step for feasible mapping, primary processing with RRT* for path generating, and post-processing with Bacterial Mutation and Node Deletion operators. This paper aims to improve further the capability of the Improved RRT* algorithm by reducing the overall computation time. The proposed method reduces the complexity of the random map after each iteration by deleting the used nodes. In this way, the computational time could be reduced.