A heterogeneous guided ant colony algorithm based on space explosion and long–short memory

Abstract

Ant colony optimization (ACO) has a good performance in solving discrete problems, but it inevitably has some disadvantages. Although it has good stability, it has some shortcomings in the convergence speed and solution accuracy when dealing with a large amount of data. Therefore, to solve the above problems, we proposed a heterogeneous guided ant colony algorithm based on space explosion and long–short memory. First, the difference between the current optimal path and the globally optimal path is used to search for the solution, which could accelerate the convergence of the algorithm. Second, the space explosion strategy is used to recombine the solution in different directions to avoid the algorithm falling into the local optimum. In this paper, 37 TSP data sets are selected to carry out simulation experiments. From the results and the rank-sum test, it could be concluded that the improved ant colony algorithm improved significantly in terms of convergence speed and solution accuracy.