Enhanced Ant Colony Optimization Algorithm for TSP Problem with Improved Convergence

In traditional ant colony optimization algorithms, the inherent randomness often leads to premature convergence into local optima, thereby compromising both solution accuracy and computational efficiency. To mitigate this limitation, our enhanced algorithm introduces a critical parameter—the pheromone evaporation factor—to strengthen global search capabilities and improve algorithmic stability. The pheromone evaporation factor controls the rate at which pheromone trails diminish during the ant path exploration process. In implementation, this is typically regulated through an evaporation coefficient (ρ) ranging between 0 and 1. A smaller ρ value results in longer pheromone persistence, encouraging ants to explore broader solution spaces and reducing the probability of locally trapped solutions. This mechanism directly enhances both search precision and convergence efficiency. From a coding perspective, the evaporation process can be implemented through matrix operations where pheromone levels τ are updated iteratively using the formula: τ = (1-ρ)*τ + Δτ. The Δτ component represents pheromone deposits from elite paths, while the (1-ρ) term ensures gradual decay of historical trails. Algorithm optimization can be achieved by dynamically adjusting ρ values through techniques like adaptive parameter tuning or simulated annealing mechanisms, ultimately achieving superior search performance and solution accuracy for Traveling Salesman Problem instances.