Ant colony algorithm based on magnetic neighborhood and filtering recommendation

To deal with the problems that the ant colony algorithm has slow convergence speed and easy falling into the local optimum when solving TSP, an ant colony algorithm based on magnetic neighborhood and filtering recommendation (MRACS) is proposed to solve these problems. First, a dynamic magnetic neighborhood strategy is adopted to balance the convergence speed and the solution accuracy by magnetic attraction. It attracts ants to enlarge the exploration of a better neighborhood, thus improving the accuracy of the result. Second, a cross-excitation strategy based on filtering recommendation is applied to increase the diversity of the algorithm by dynamic weakening or enhancing local pheromones in the neighborhoods. It aids the algorithm get rid of the local optimum. Through the simulation experiments and the rank-sum test analysis, it is observed that the MRACS can effectively balance the convergence speed and the accuracy of the solution.