Pattern-Based Genetic Algorithm Approach to Coverage Path Planning for Mobile Robots

Sensor-based mobile robot coverage path planning (MRCPP) problem is a challenging problem in robotic management. We here develop a genetic algorithm (GA) for MRCPP problems. The area subject to coverage is modeled with disks representing the range of sensing devices. Then the problem is defined as finding a path which runs through the center of each disk at least once with minimal cost of full coverage. The proposed GA utilizes prioritized neighborhood-disk information to generate practical and high-quality paths for the mobile robot. Prioritized movement patterns are designed to generate efficient rectilinear coverage paths with no narrow-angle turn; they enable GA to find optimal or near-optimal solutions. The results of GA are compared with a well-known approach called backtracking spiral algorithm (BSA). Experiments are also carried out using P3-DX mobile robots in the laboratory environment.