Modified A* algorithm suitable for online car-like mobile robot control

Abstract

This paper presents a theoretical analysis of different methods, which can be used for online path generation for mobile robots with nonholonomic constraints in a partially known workspace. We then introduce a new method for solving the problem. The method presented works very fast and gives the optimal path in a complex environment. The algorithm of the method presented is based on A* graph searching with the nodes placed in a discretised configuration space. This paper is focused on the analysis of different heuristic cost functions and introduces new statement of this function. The method presented is equipped with procedures for fast path replanning, very useful in partially known workspaces. The series of simulation tests and experimental results of online control for car-like robot are also presented.