Obstacle Avoidance Path Planning Based on Target Heuristic and Repair Genetic Algorithms

Using genetic algorithm (GA) to optimize the mobile robot path planning has the disadvantages of low initial population generation efficiency and low initial population quality, especially under large size and complex environment model in grids. In order to overcome this problem, a novel methodology contains a target heuristic operator and a reparation operator is proposed in this paper. In this study, the maps consist of the obstacles areas and the feasible areas are decomposed by grids. These two operators are integrated into the GA and applied to acquire the collision-free shortest path in a static two-dimension environment. The experimental data show that the methodology can decrease significantly the random search time for generating the initial population and improve the quality of the initial population generation. Results suggested that the proposed requires a shorter amount of time and possesses a better global searching performance, compared with the conventional methods.