GO-GASA: Grid Optimization Genetic A* Algorithm for Mobile Robots Path Planning Utilizing Grid Optimization

Abstract

Path planning, as a central challenge in mobile robotics, directly impacts the operational efficiency and safety of robotic systems. To enhance navigation performance in complex environments, this paper proposes a two-stage optimization framework: First, a Grid-Optimized Genetic A* Algorithm (GO-GASA) is developed by integrating the global search capability of genetic algorithms with an enhanced A* algorithm for optimizing the generation of initial populations. This integration not only improves the initialization performance of genetic algorithms but also enhances their overall optimization capacity. Second, a Grid-Optimized Redundant Path Removal Algorithm (GO-RPRA) is introduced to perform secondary optimization on the paths generated by GO-GASA, further improving path smoothness and efficiency. Simulation results demonstrate that in initialization experiments, the individual diversity of GO-GASA is 7.8 times higher than that of traditional genetic algorithms (GA). Under identical testing conditions, compared to six other popular metaheuristic algorithms, GO-GASA reduces the number of turns by 53.85%, the max turning angle by 50%, and the total distance by 25.24%. Overall, these improvements significantly enhance the quality of path planning and strengthen the capability of mobile robots to handle complex tasks.