Hybrid Genetic Algorithms for Large Scale Optimization Problem

In the fields, such as engineering system and multiple tasks cooperation, many real-world problems can be modeled by colored traveling salesmen problem (CTSP). Since CTSP has been proved belong to the NP-hard, the intelligent algorithms, such as genetic algorithm (GA), have been used for solving small or median scale cases where the number of cities is less than 1000. This paper uses three improved hybrid genetic algorithms, including GA with greedy algorithm (GAG), hill-climbing GA (HCGA), and simulated annealing GA (SAGA), to solve large scale CTSP, where three algorithms greedy algorithm, hill-climbing and simulated annealing are used to improve GA. The experiments show that hybrid genetic algorithms demonstrate an improvement over GA in term of solution quality.