Local Area Logistics Distribution Model And Its Intelligent Solution Algorithm

Logistics distribution,of which any route planning is directly related to multiple factors in the entire process of logistics transportation such as speed,cost and benefit,is both the key link of the logistics industry and one of core functions of the logistics system.It has been a crucial problem to be solved urgently how to develop a practical logistics distribution planning model,probe into its algorithm and acquire a practical logistics distribution management information system.To this end,this thesis tries to construct single-or multiobjective mathematical programming models with the performance indices of distribution route,transportation cost or vehicle allocation size,in order to formulate the problem of the logistics distribution.Therein,some constraints are considered such as time,vehicle capacity and vehicle allocation.Several hybrid fruit fly optimization algorithms are designed to solve such established programming models,after some inspirations from other intelligent optimization algorithms are incorporated into the basic fruit fly optimization algorithm.Their experimental analyses are carried out,including algorithms' performance test,application and comparison.The main works and acquired achievements are summarized as follows:A.Aiming at the problem of logistics distribution route planning which is comprehensively concerned by the logistics industry,a 0-1 planning model is established by taking a local logistics distribution route as its performance index and also considering the actual factors whether there exist a direct pathway between distribution points.A fruit fly genetic optimization algorithm with computational complexity depending on the sizes of population and logistics distribution is designed to solve the optimal solution,in which the ideas of gene exchange and maximum reservation cross are fused into the basic fruit fly optimization algorithm in order to strengthen its diversity of population.Herein,some elicit fruit flies are exploited by means of a gene inversion operation and a neighborhood detection strategy.Comparative experiments and engineering application show that thealgorithm has clear advantages over those compared methods in terms of search effect,stability,convergence speed and the rationality of the obtained distribution route scheme.B.Based on the constraint limitations of time window and cargo demand,a minimum model is designed to reflect the problem of logistics distribution cost.A hybrid fruit fly optimization algorithm is developed to seek the optimal logistics distribution scheme,after several reported strategies of two-person league selection,crossover,simulated annealing and neighborhood search are combined with the basic fruit fly optimization algorithm.Comparative experiments show that the hybrid optimization algorithm can effectively handle the problem of logistics distribution cost if the distribution size is less than 25;the search efficiency of the algorithm is high,and the obtained logistics distribution scheme is rational.C.Based on the constraints of time window and cargo demand,a bi-objective mathematical programming model is constructed to formulate the problem of logistics distribution by taking the performance indices of local logistics distribution cost and vehicles' number into account.Depending on the basic fruit fly optimization algorithm,genetic operators,pushing forward insertion heuristics rule and Arena's Principle,a bi-objective fruit fly optimization algorithm is designed to seek multiple logistics distribution schemes.Comparative experiments indicate that the algorithm can carry out satisfactory and stable solution search with high efficiency while the effect of solution search depends lowly on the number of logistics.