Settings of Algorithm Parameters in Ant Colony Algorithm

Abstract

The relevant parameters in the ant colony algorithm have great impact on algorithm performance and various parameters are closely linked, and a good parameter combination will increase the overall search capability and convergence of algorithm. At present, the parameter settings of the ant colony algorithm are determined relying on experience and experiments which have heavy workload and it is difficult to get the optimal combination of parameters. On the basis of the idea that algorithm parameters are uniformly designed, a combination of representative experimental points (parameter values) is selected for experimental design, and a satisfactory combination of algorithm parameters is found with fewer number of experiments, so that the optimum operating performance can be realized by the algorithm. The simulation experiments show that the method is feasible and effective.