A fast convergence bidirectional a star path planning algorithm

Abstract

A* and bidirectional A* have been classical pathfinding algorithms for long with many derivatives specially designed for their exclusive situations. To solve the problem of the bidirectional A* algorithm being inefficient in some situations, this paper gives a kind of optimized algorithm with a new dynamic parameter in the heuristic function and a 25% increase in efficiency. By modifying the weights, the proportion between distance and efficiency can be well balanced. This paper uses some typical problems for simulation and paints the scatter diagram of different weights to figure out the relationship between the weight and the balance. In the end, the author gives suggestions on optimizing the coefficient due to the rule, which stands for gradually promoting the weight of the dynamic parameter until the path doesnt keep optimal, for various situations and how to further modify the heuristic function for better performance.