Optimization of Path for Road Network With Modified Ant Colony Optimization (MACO)

Abstract

Optimizing routes in road networks is crucial for smooth transportation and economic progress. Different methods exist for finding the best routes, including genetic algorithms, particle swarm optimization, and simulated annealing. Ant Colony Optimization (ACO) stands out for its efficiency. In this study, we introduce a modified version called MACO, which considers accidents when determining optimal routes. Evaluating different ACO versions reveals differences in solution quality, runtime, and number of iterations. Performance metrics including maximum obtained solution, runtime, and iteration number were evaluated for each method. In Case 1, TACO, and AACO both achieved a maximum of 21 solutions from the available possible solution of 24, exhibiting run-times of 0.4359 and 0.4575 s, respectively. Meanwhile, MACO attained a maximum of 22 solutions from available possible solution 24, in a runtime of 0.5345 s and 10 iterations. In the second scenario, TACO, AACO, and MACO achieved maximum solutions of 20 with obtained solutions of 15, 16, and 17, respectively. TACO demonstrated a runtime of 0.1853 s with 26 iterations, AACO ran in 0.1749 s with 22 iterations, and MACO completed in 0.5799 s with 15 iterations. These findings highlight the varying performance of the optimization methods and suggest MACO as a promising approach for balancing solution quality and computational efficiency in road network path optimization.