An improved genetic algorithm for combinatorial optimization

摘要

By analyzing the deficiency of traditional genetic algorithm (GA for short) in solving the Traveling Salesman Problem (TSP for short) which is one representative problem of the combination optimization, we improved the algorithm structure of traditional genetic algorithm. By improving the population variation by adjusting fitness values and proposing heuristic crossover operation, 2-opt local searching and self-adapting genetic parameter, the algorithm achieved a balance between the quality and efficiency. According to the analysis and tests, the improved generic algorithm could get better result than the traditional genetic algorithm. This showed that the method had better feasibility and practicability.