面向旅游行程规划的交互式多智能体遗传算法*

结合多智能体技术和交互式遗传算法，提出了一种面向旅游行程规划问题的交互式多智能体遗传算法。算法通过让固定在网格上的智能体展开进化和竞争行为来寻找满意行程。在算法每代中，用户只需评价选择一个当代最优智能体，就可计算得到当代所有智能体的能量，减少了评价次数，有效缓解了用户在评价过程中的疲劳问题。仿真实验验证了该算法在解决旅游行程规划问题中的可行性和有效性，并对问题规模表现出很好的可伸缩性。

Interactive multi agent genetic algorithm for travel itinerary planning

The paper proposed an interactive multi-agent genetic algorithm for the travel itinerary planning problem,which combined the multi-agent technology with the interactive genetic algorithm.The algorithm made agents fixed on a lattice evolve and compete in order to search the satisfactory itinerary.In every generation,a user only needed to evaluate and find out an agent which was the current best one,and then energies of all agents in this generation could be calculated automatically,which reduced the user's evaluations and contributes to relieve the human fatigue in the evaluation process.The simulation experiment shows that the algorithm is a feasible and effective method for the travel itinerary planning problem,and has good scalability for the problem's size.