PSO-GA with Dynamic Coefficients for Solving the Traveling Salesman Problem

The Traveling Salesman Problem (TSP) is a classic combinatorial optimization problem with wideranging applications in various real-world fields. To solve the TSP, this study introduces the neighborhood measure and integrates a Genetic Algorithm (GA) to enhance the Particle Swarm Optimization (PSO) framework, proposing a new PSO variant (PSO-GA-DC) with dynamic coefficients and additional local search operations. Using the neighborhood measure, three new dynamic coefficients are defined, replacing the traditional fixed and random coefficients, thus reducing redundancy in coefficient settings. After each iteration, the neighborhood measure is used to optimize and rearrange local segments, accelerating convergence. Additionally, GA's crossover and mutation mechanisms are incorporated to perform local searches on each candidate solution, avoiding path crossings while exploring new path possibilities. Through empirical testing on TSP instances and comparisons with other algorithms reported in the literature, the proposed hybrid model (PSO-GA-DC) achieves multiple best results in terms of both the best value (Best) and average value (AVG), providing shorter paths while demonstrating excellent accuracy and stability.