Multiple traveling salesman problem (M-TSP) is a routing problem to visit n cities by m salesmen (m < n) on the condition that a salesman can only visit a city once and that the journey has to end at the city of origin. n cities are divided into m clusters and each cluster is visited by a salesman. The division of the cluster will use K-means clustering algorithm, and each cluster will look for the route using genetic ant colony system (GACS) algorithm which is a combination of genetic algorithm and ant colony system. Then three datasets are used in the implementation of the K-means clustering-genetic ant colony system (K-GACS) algorithm and K-means clustering-genetic algorithm (K-GA). The results are compared later. From the experimental results, it will be confirmed that K-GACS algorithm is better than the K-GA.
- Genetic ant colony system (GACS) algorithm
- K-means clustering algorithm
- Multiple traveling salesman problem (M-TSP)