A transfer learning-based particle swarm optimization algorithm for travelling salesman problem

Abstract To solve travelling salesman problems (TSPs), most existing evolutionary algorithms search for optimal solutions from zero initial information without taking advantage of the historical solving similar problems. This paper studies a transfer learning-based particle swarm optimization (PSO) algorithm, where is used to guide find paths quickly. begin with, all cities in new and TSP are clustered into multiple city subsets, respectively, topology matching strategy based on geometric similarity proposed match each subset subset. Then, basis above-matched results, hierarchical generation feasible path (HGT) initialize improve performance PSO. Moreover, problem-specific update strategy, i.e. with adaptive crossover clustering-guided mutation, introduced enhance capability algorithm. Finally, algorithm applied 20 typical compared 12 state-of-the-art algorithms. Experimental results show that learning mechanism can accelerate efficiency PSO make achieve better paths.