Global path planning using improved ant colony optimization algorithm through bilateral cooperative exploration

We proposed the Heterogeneous Ant Colony Optimization (HACO) algorithm to solve the global path planning problem for autonomous mobile robot in the previous paper. The HACO algorithm was modified and optimized to solve the global path planning problem unlike the conventional ACO algorithm which was proposed to solve the Traveling Salesman Problem (TSP) or Quadratic Assignment Problem (QAP). However, there is a common shortcoming in the ACO algorithms for global path planning, including HACO algorithm. Ants carry out the exploration task relatively well around the starting point. On the other hand, they are hindered in their work as they approached the goal point, because they are attracted by the intensity of heuristic value and the accumulated pheromone while the ACO algorithm works. As a result, they have a strong tendency not to explore and most of them follow the path that found in the beginning of the search. This could cause the local optimal solutions. Thus, we propose a way to solve this problem in this paper. It is the Bilateral Cooperative Exploration (BCE) method. The BCE is the idea that performs the search task again by changing the goal point into the starting point and vice versa. The simulation shows the effectiveness of the proposed method.