Enhanced A* Algorithm for the Time Efficient Navigation of Unmanned Vehicle by Reducing the Uncertainty in Path Length Optimization

摘要

Optimal path planning is considered as a crucial problem in mobile robotics. The conventional A* algorithm is primarily focused on the heuristic values of nodes. However, during the experimental analysis, it observed that the occupied node due to obstacles, effect the selection of shortest and time efficient path in conventional A*. Therefore, a modified bidirectional path planning technique-Time Optimized A* (TOA*) algorithm is proposed that always choose the time efficient and shortest path with lesser number of operations to reach the destination from start position. The proposed TOA* algorithm is trial in various simulated tests and the execution time is reduce by 33.75% (Conventional A*), 71.47% (Breadth First Search) and 67.66% (Jump point search) is observed likewise, a reduction in number of operations by 66.14% (conventional A*), 91.35% (Breadth First Search) and 89.80% (Jump point search) is observed. The proposed TOA* is also tested on a mobile robot in real-world experiments and a significant reduction of 31.33% in number of closed nodes and 14.08% reduction in sharp turns has achieved. The reduction of sharp turns with lesser number of operations resulted in reduction in total time by 5.63% and the acceleration of mobile robot is increased by 7.01% proportionate to conventional A*.