Local path planning based on Bi-population Swarms optimization algorithms

Robot navigation and path planning are among the most critical challenges facing mobile robots. Modern techniques have surpassed traditional methods in reducing high complexity by utilizing probabilistic or optimal solutions, such as algorithms based on swarm intelligence. This work proposes a novel global path planning using a bipopulation grasshopper optimization algorithm (BiGOA). Superior performance results were obtained compared to the original grasshopper optimization algorithm regarding time, cost, and path length for an average of 1000 times executions in environments of varying complexities. On the other hand, the optimal path, local path planning, and dynamic obstacle avoidance are done by the Dynamic Window Approach (DWA) algorithm that works within the follow-waypoint technique in mobile vehicles in the Robot Operating System (ROS). During robot movement, it tends to find a shorter and optimal path using a modified version of the waypoint method. A state machine determines the optimal path using the improved follow-waypoint algorithm during the robot’s real-time movement, thereby controlling the path’s smoothing, shortening, and continuity. This study compares the BiGOA algorithm with an optimal path method based on arrival time over ten trials. The results show that BiGOA paths were, on average, 33 s longer. Additionally, the robot’s performance using the proposed method was tested in a dynamic and complex environment, where the results indicated a reduction in both time and effort.