A multi-group ant colony system algorithm for TSP

As a new class of global searching algorithms, ant colony system algorithm could solve TSP (traveling salesman problem). These algorithms includes ACS, MAX-MIN ant system, et al. This work presents a new method named multi-group ant colony system algorithm. This algorithm avoids some defects of ACS and MAX-MIN ant system. These defects make algorithm not iterate when it has arrived at the stagnating state of the iteration or local optimum point. But for multi-group ant colony system, it creates new groups of ants to iterate when meeting those states. In this paper, a simple convergence proof is presented. At the end of this paper, the experimental result is presented to show the effectiveness of this method.