Ant Colony Optimization-Based Path Planning for Autonomous Vehicle Navigation Systems

A crucial aspect of autonomous vehicles is path planning, which involves using various algorithms to determine the most efficient routes. This study investigates the application of the Ant Colony Optimization (ACO) algorithm to create a system for identifying the shortest path. The process includes collecting coordinate data from specific test locations on the Universitas Sriwijaya Indralaya campus to generate a graph-based map for analysis. The study begins with gathering route data through the Google Maps application, which includes 182 routes. In simulation tests, the routes generated by ACO differ from those suggested by Google Maps; sometimes, the ACO route is longer and sometimes shorter, due to the probabilistic nature of the ACO algorithm in finding optimal routes. The average error compared to Google Maps is 11.7 meters. Additional tests introduce obstacles on the road, requiring the system to find alternative routes. For instance, the initial route from the auditorium to the Faculty of Science and Math is 38-39-11-53-110-54-111 with a route length of 670 m. After adding a roadblock in front of the Faculty of Engineering, the route changes to 38-39-116-40-70-69-56-111, with a route length of 730 meters. These tests demonstrate that the ACO algorithm effectively identifies the best routes for autonomous vehicles, both under normal conditions and when obstacles are present.