基于多目标遗传算法的灾后可靠路径选择

Reliable path selection after disaster based on multi-objective genetic algorithm

为了尽快运出灾民、运进物资,灾后需要选择最短的运输路径;考虑到灾害对道路的损毁,安全的路径是道路选择中的重要因素,基于灾后道路的损毁状况,采用3种方法定义道路可靠性,以最小化路径长度、最大化路径可靠性为目标建立双目标优化模型,采用多目标遗传算法求解.分别求得3种情形下的解,开展2个目标的Pareto分析;分析交叉概率和变异概率对结果的影响:交叉概率增大,覆盖范围波动式变化,理想积波动式下降;变异概率增大,覆盖范围波动式变化,理想积缓慢增长后逐渐下降.采用遗传算法能够有效地解决最短和最可靠路径的搜索问题.

The shortest path is needed to transport victims and materials as soon as possible.Safe path becomes the critical factor in road selection considering the road damage in the disaster.Three means were adopted to define the road reliability based on the damage condition of roads after a disaster,and a bi-objective optimization model was established.The model tries to minimize the length of road and maximize the roads＇ reliability.The priority-based multi-objective genetic algorithm was used to handle the problem.The solutions were respectively got under three situations.Pareto analysis was conducted.The impacts of crossover rate and mutation rate on the results were analyzed.With the crossover ratio increases,the coverage fluctuates and the fluctuated ideal volume decreases.With the mutation ratio increases,the coverage fluctuates and the ideal volume grows slowly and then gradually declines.The genetic algorithm can effectively solve the shortest and most reliable path problem.