Generation of Optimal Coverage Paths for Mobile Robots Using Hybrid Genetic Algorithm

Abstract

This paper presents new optimal offline approaches to solve the coverage path planning problem. A novel hybrid genetic algorithm (HGA), which uses, the turn-away starting point and backtracking spiral algorithms for performing local search, is proposed for grid-based environmental representations. The HGA algorithm is validated using the following three different fitness functions: the number of cell visits, traveling time, and a new energy fitness function based on experimentally acquired energy values of fundamental motions. Computational results show that compared to conventional methods, HGA improves paths up to 38.4%; moreover, HGAs have a consistent fitness for different starting positions in an environment. Furthermore, experimental results prove the validity of the fitness function.