Modified adaptive ant colony optimization algorithm and its application for solving path planning of mobile robot

• Published in: Expert systems with applications Vol. 215; p. 119410
• Main Authors: Wu, Lei, Huang, Xiaodong, Cui, Junguo, Liu, Chao, Xiao, Wensheng
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 01.04.2023
• Subjects: Ant colony optimization algorithm; Heuristic function; Path planning; Pheromone updating; State transition probability; Heuristic function; Path planning; Pheromone updating; Ant colony optimization algorithm; State transition probability
• ISSN: 0957-4174; 1873-6793
• DOI: 10.1016/j.eswa.2022.119410

•A novel variant of ACO called MAACO is proposed to solve mobile robot path planning.•The proposed MAACO includes four novel mechanisms.•Parameter optimization of MAACO is operated to find suitable parameter combination.•The MAACO is compared with several existing approaches based on several instances.•Experimental results show the superiority of MAACO. As the key point for auto-navigation of mobile robot, path planning is a research hotspot in the field of robot. Generally, the ant colony optimization algorithm (ACO) is one of the commonly used approaches aiming to solve the problem of path planning of mobile robot. Nevertheless, the traditional ACO has the shortcomings such as slow convergence speed, inefficiency and easily fall into local optimal values. Thus, a novel variant of ACO is proposed in this study. In detail, a new heuristic mechanism with orientation information is firstly introduced to add direction guidance during the iteration process, further to advance the convergence speed of algorithm. Secondly, an improved heuristic function is presented to enhance the purposiveness and reduce the number of turn times of planned path. Then, an improved state transition probability rule is introduced to improve the search efficiency significantly and increase the swarm diversity. Moreover, a new method for unevenly distributing initial pheromone concentration is proposed to avoid blind searching. After integrating the four improvements, the new variation of ACO called modified adaptive ant colony optimization algorithm (MAACO) is formed. Subsequently, parameter optimization of MAACO is carried out. For verifying the effectiveness of the proposed MAACO, a series of experiments are conducted based on five static space environment modes and one dynamic environment mode. Comparing with 13 existing approaches for solving the problem of path planning of mobile robot, including several variants of ACO and two commonly used algorithms (A* algorithm and Dijkstra algorithm), the experimental results demonstrate the merits of MAACO in terms of decreasing the path length, reducing the number of turn times, and promoting the convergence speed. In detail, in all the static simulation experiments, the proposed MAACO generates the shortest path length with a standard deviation of zero, and achieves the least number of turn times within the smallest convergence generation. In terms of the five experiments, the average number of reducing turn times is two with a generally reduction ratio of 22.2% compared with the best existing results. The obtained results of MAACO prove its practicality and high-efficiency for path planning.