Pseudo-bacterial potential field based path planner for autonomous mobile robot navigation

This paper introduces the pseudo-bacterial potential field (PBPF) as a new path planning method for autonomous mobile robot navigation. The PBPF allows us to obtain an optimal and safe path, in contrast to the classical potential field approach which is not suitable for path planning because it lacks a means of obtaining the optimal proportional gains. The PBPF uses the pseudo-bacterial genetic algorithm (PBGA) and a fitness function based on the potential field concepts to construct viable paths in dynamical environments to mostly result in the optimal path being obtained. Comparative experiments of sequential and parallel implementations of the PBPF for off-line and online in structured and unstructured conditions are presented; the results are contrasted with the artificial potential field (APF) method to demonstrate how the PBPF proposal overcomes the traditional method.