Enhanced Mutation Operator Improved from the Original Ant Colony Algorithm's Mutation Mechanism

Main Text: The Ant Colony Optimization (ACO) algorithm serves as a classical optimization method for solving the Traveling Salesman Problem (TSP). Its core principle mimics the pheromone-based foraging behavior of ants to identify optimal paths. In traditional ACO implementations, introducing a mutation operator helps escape local optima. However, for TSP instances with 100 cities, standard mutation operators often face challenges such as insufficient convergence speed or excessive disturbance.

The enhanced mutation operator introduces two key improvements: 1. Adaptive Mutation Probability: Dynamically adjusts mutation intensity based on iteration progress. In early stages, higher mutation rates allow broader path exploration across the solution space. As iterations advance, the probability gradually decreases to focus on local refinement. Implementation typically involves a decay function like: mutation_prob = initial_prob * exp(-iteration/total_iterations). 2. Elite Path Protection: Preserves the best path from each generation by exempting it from random mutations, preventing disruption of high-quality solutions. Non-elite paths undergo diversity-enhancing operations such as swap mutations (exchanging two random city positions) or inversion mutations (reversing a path segment). This is often coded using conditional checks before applying mutation functions.

These enhancements significantly improve algorithm performance for 100-city TSP - effectively preventing premature convergence while maintaining efficiency by avoiding excessive randomization. Experimental results demonstrate that the refined mutation operator enables more stable approximation of global optima during mid-to-late iteration phases.