Indian Journal of Science and Technology
DOI: 10.17485/ijst/2016/v9i45/102431
Year: 2016, Volume: 9, Issue: 45, Pages: 1-10
Original Article
Velappa Ganapathy1* Priyanka Sudhakara1 Titus Tang Jia Jie2 and S. Parasuraman3
1School of Computing, SRM University, Kancheepuram - 603203, Tamil Nadu, India; [email protected], [email protected] 2Computer Software, Monash University, Melbourne, Australia; [email protected] 3Monash University, Lagoon Selatan, Bandar Sunway, Subang Jaya - 47500, Selangor, Malaysia; [email protected]
*Author for correspondence
Velappa Ganapathy School of Computing, SRM University, Kancheepuram - 603203, Tamil Nadu, India; [email protected]
Objectives: This paper presents an amended Ant Colony Optimization (ACO) algorithm for a mobile robot navigation to find the most optimal path. Methods: A modified design and development of an improved Ant Colony Optimization algorithm based upon a prior research work done is proposed in this paper. The algorithm put forth is enhanced by simplifying the equations already proposed and enlarging the area of the simulation framework, extending the task capabilities of the robot, as well as testing the algorithm in real time on an autonomous mobile robot. Findings: The proposed algorithm has to calculate optimal trajectory for the mobile robot to traverse to perform the following tasks: target-searching, boundaryfollowing and obstacle avoidance. The total length of the path traversed determines the efficiency of path traced. This proposed method also enhances the utility of the ACO algorithm by designing and creating a feasible ACO graphical user interface. Further we carried out the research on the working of the ACO algorithm by performing systematic testing, simulations and real-time implementation. Improvements: Future work could involve the implementation of a positioning system that allows the robot to determine its actual real world position and then provide feedback to the ACO algorithm so that adjustments could be made. The basic ACO algorithm could be modified to model ants to move in eight directions. All simulations and real time implementations could be done in pre-known environments with static well defined obstacles. By including dynamic obstacle avoidance capabilities, the range of real life applications in which the algorithm could be implemented on would be greatly expanded.
Keywords: ACO Algorithm, Known Environment, Mobile Robot, Machine Learning, Navigation Planning, Static Obstacles
Subscribe now for latest articles and news.