基于物流配送路线规划的改进型果蝇优化算法

Logistics distribution route planning based improved fruit fly optimization algorithm

针对物流业普遍关注的物流配送路线规划问题,以配送路程为性能指标并结合配送点之间是否可直接通行的实际因素,获得0-1规划模型。在基本果蝇优化算法中,引入基因互换和最大保留交叉操作增强果蝇种群的多样性,借助基因逆位操作和邻域探测策略搜索优质果蝇,获得计算复杂度依赖于种群规模和配送点数的改进型果蝇优化算法。比较性的数值实验和工程应用结果表明,该算法在搜索效果与稳定性、收敛速度及获得的配送路线方案的合理性方面具有明显优势。

Aiming at the actual factor that whether any two distribution points can arrive to each other,one 0-1 programming model with the objective function of distribution route was modeled to formulate the hot problem of logistics distribution route planning. An improved fruit fly optimization algorithm was developed to seek the optimal route planning scheme,in which the modules of gene exchange and maximum saving crossover were used to strength the diversity of fruit fly population and also those of gene inversion and neighbor exploitation were cited to seek elitist fruit flies. Its computational complexity depended mainly on the sizes of fruit flies and distribution points. Numerically comparative experimental results show that the proposed algorithm clearly performs well over the compared approaches in terms of performance effect,search stability,convergence speed,and the rationality of the acquired logistics distribution scheme.