Pathfinding in a Dynamically Changing Environment

Abstract

This paper is focused on a proposal of an algorithm for dynamic path planning which uses the input speed of the vehicle as a dynamic value based on which paths are searched in a graph. The proposed algorithm AV* stems from algorithm A*, works with proposed weight functions of evaluation function and, once again, uses information about nodes which have already been calculated in order to find out a faster update of a new path. This algorithm has been subjected to tests, whose results were compared with the results of algorithm A*. The proposed algorithm AV* complies with the condition of pathfinding between two points in a finite time based on the criterion of the fastest path using the maximum of the speed limit. From the point of view of dynamic optimization, all tests achieved a higher speed when the algorithm was run again at a different speed. This speed-up corresponds to approximately 25% and is linearly dependent on the amount of searched nodes.