Global path planning for mobile robot based on Artificial Bee Colony and Dijkstra’s algorithms
Rafał Szczepański, Tomasz Tarczewski
Abstract
The global path planning for mobile robot in variable robot space is not a trivial matter, since generation of optimal path is very time-consuming. A variable workspace may result in infeasible pre-optimized path, due to new obstacles or worker position. In this paper, a hybrid approach to path planning for mobile robot in variable space is proposed. It consists of off-line global path optimization algorithm and online path planning schemes. The first part requires the definition of most common situations of workers and obstacles positions, and then global optimization of the path for a mobile robot. It is made using Artificial Bee Colony algorithm. The control path points generated from off-line part are used to create a graph in on-line part. Next, the shortest path is selected using Dijkstra's algorithm. The proposed approach ensures an at least satisfactory path in variable workspace with small computational effort in real-time path planning. The examination was made in Matlab/Simulink environment and the supplementary material, incl. visualization of the considered example, is included.