What is it about?
We created a smart computer method that quickly finds nearly the shortest route between many cities. It's faster than traditional approaches and useful for delivery planning.
Featured Image
Why is it important?
Our hybrid algorithm uniquely combines three search methods (simulated annealing, 2-opt, and tabu search) with a smart stopping parameter (β) that balances speed and accuracy. Unlike most TSP solvers that prioritize either quality or speed, our method achieves both: 1.44% average error while running significantly faster than ant colony, neural network, and genetic algorithms. This matters for logistics and smart manufacturing (Industry 4.0), where real-time routing decisions need near-optimal solutions quickly. Our approach proves that classical metaheuristics, when cleverly combined, can outperform more complex methods for small-to-medium problems (up to ~500 cities), offering an accessible, easy-to-implement tool for businesses needing rapid route optimization.
Perspectives
Collaborating across six institutions in Mexico made this work truly enriching. I'm particularly proud of our β parameter—a simple but powerful idea that stops the algorithm when improvements stall. This came from real-world frustration: why let a program run forever for marginal gains? I hope readers see that combining classic methods creatively often beats chasing overly complex solutions. Our algorithm won't replace state-of-the-art solvers like LKH, but it offers a practical, transparent tool for small-to-medium routing problems—perfect for students, small businesses, or anyone who needs a fast, decent route without a PhD in optimization.
Dr. Alejandro Medina Santiago
INAOE
Read the Original
This page is a summary of: A Hybrid Heuristic Algorithm for the Traveling Salesman Problem with Structured Initialization in Global–Local Search, Algorithms, April 2026, MDPI AG,
DOI: 10.3390/a19050324.
You can read the full text:
Contributors
The following have contributed to this page
