Heterogenous Adaptive Ant Colony Optimization with 3-opt local search for the Travelling Salesman Problem

Abstract

The majority of optimization algorithms require proper parameter tuning to achieve the best performance. However, it is well-known that parameters are problem-dependent as different problems or even different instances have different optimal parameter settings. Parameter tuning through the testing of parameter combinations is a computationally expensive procedure that is infeasible on large-scale real-world problems. One method to mitigate this is to introduce adaptivity into the algorithm to discover good parameter settings during the search. Therefore, this study introduces an adaptive approach to a heterogeneous ant colony population that evolves the alpha and beta controlling parameters for ant colony optimization (ACO) to locate near-optimal solutions. This is achievable by introducing a set of rules for parameter adaptation to occur in order for the parameter values to be close to the optimal values by exploring and exploiting both the parameter and fitness landscape during the search to reflect the dynamic nature of search. In addition, the 3-opt local search heuristic is integrated into the proposed approach to further improve fitness. An empirical analysis of the proposed algorithm tested on a range of Travelling Salesman Problem (TSP) instances shows that the approach has better algorithmic performance when compared against state-of-the-art algorithms from the literature.