Discrete Salp Swarm Algorithm for symmetric traveling salesman problem

Area of Science:

• Computational Intelligence
• Operations Research
• Metaheuristic Optimization

Background:

• The Salp Swarm Algorithm (SSA) is a nature-inspired metaheuristic known for its chain-based movement mechanism.
• Existing SSA versions primarily focus on continuous optimization problems, with limited discrete variants.
• The Traveling Salesman Problem (TSP) remains a challenging combinatorial optimization problem requiring efficient algorithmic solutions.

Purpose of the Study:

• To develop a discrete version of the Salp Swarm Algorithm (SSA) tailored for combinatorial optimization problems.
• To enhance the SSA's ability to address the Traveling Salesman Problem (TSP) by incorporating problem-specific modifications.
• To improve the algorithm's exploration and exploitation balance, mitigating the risk of premature convergence to local optima.

Main Methods:

• Modification of the 2-opt algorithm with a decreasing factor 'd' for neighborhood search control.
• Integration of Problem Aware Local Search (PALS) tailored to TSP characteristics.
• Introduction of a second leader mechanism to increase randomness and escape local optima.
• Selection and integration of the Subtour Exchange Crossover (SEC) operator within the SSA framework to create the Discrete Salp Swarm Algorithm (DSSA).

Main Results:

• The developed Discrete Salp Swarm Algorithm (DSSA) was tested on 23 benchmark TSP instances.
• Comparative simulations demonstrated DSSA's effectiveness against other advanced optimization algorithms.
• DSSA achieved satisfactory solutions for the tested TSP instances, indicating robust performance.

Conclusions:

• The proposed Discrete Salp Swarm Algorithm (DSSA) offers a viable and effective approach for solving the Traveling Salesman Problem (TSP).
• The integration of problem-specific mechanisms and enhanced exploration strategies significantly improves upon the standard SSA for discrete optimization tasks.
• DSSA demonstrates competitive performance, providing a valuable tool for complex routing and combinatorial optimization challenges.