Real Time Evaluation of Grid Based Path Planning Algorithms: A comparative study

摘要

Path Planning algorithms are indispensable in Autonomous Navigators. Though rover may not have prior information about the environment, path planning algorithms must be robust enough to navigate it to the target. A~* and D~* are grid based Path Planning algorithms which enable to reach the target in shortest path. Cost computations are fixed in A~* algorithm while re-computations are done in D~* algorithm, whenever new obstacles are identified. A modified version of D~* algorithm is presented in this paper where cost re-computation is done based on certain criterion. By reducing the number of times cost re-computation occurs, Modified D~* aims to produce shortest path similar to D~* algorithm with less memory requirements and execution time. This paper also brings out the implementation of A~*, D~* and Modified D~* algorithms with NXT LEGO Mindstorms kit and their evaluation in real time scenario. Comparison of the above algorithms is elicited.