A New Cost Function Heuristic Applied to A* Based Path Planning in Static and Dynamic Environments

Abstract

An important task for mobile robots is autonomous navigation, where the robot travels between a starting point and a target point without the need for human intervention. This task can be described as a planning path problem, whose purpose is to locate sequential segments of state transitions (Cells) from an initial to a final goal. This paper investigates a family of trajectory generation algorithms (A*), which are commonly used in path planning for static environments, stressing their main properties. Then, it is presented a new cost function heuristic that is used to optimize the results presented in the original approaches. The comparison of all algorithms is carried out via a set of experiments, which show that the new heuristic reduces the computational cost of the search, the amount of expanded cells and mainly the time required to locate targets. These experiments also carried out in both static and dynamic environments.