Hybrid Genetic Algorithm Approach for Mobile Robot Path Planning

The proposed hybrid approach integrates the trajectory planning with the path planning for the mobile robot navigation in the indoor simulated environment. GA (Genetic Algorithm) makes the robot to choose the optimal path selection to reach the gol with global path planning. GJK distance algorithm supports the collision avoidance of convex shaped dynamic obstacles during navigation. The GJK algorithm finds the obstacles position in the environment and incorporates the value in the fitness function during every step movement of the robot and adds the internal mechanism of obstacles values into the Genetic Algorithm which makes the effective optimal path selection in the environment. To ensure statistical significance of the hybrid approach, paired sample„t‟ test were performed using SPSS tool. Hybrid GA takes less time over GA in various tests of different shapes of obstacles in the environment. The overall various shaped obstacles the average time taken to reach the goal by Hybrid GA (164.33 seconds) is lesser than the GA (671.33seconds) in reaching the goal. Integration of the distance based algorithm with heuristic approach makes the best optimal path selection with the effective obstacle prediction and avoidance in the environment.