【作者】 马福晶；

【导师】 赵庆祯；

【作者基本信息】 山东师范大学 ， 管理科学与工程， 2006， 硕士

【摘要】 近年来,随着农业生产水平的提高,加入WTO后发达国家农产品进入我国市场,我国农业发展已经进入节约要素投入和改善要素配置效率阶段,迫切需要我们实现农业现代化。农业现代化需要在农业领域中引入供应链思想和供应链管理,以增强农产品市场竞争力,增加农民收入,加快农业现代化发展。农业供应链已经成为我国经济发展的重要产业和新的经济增长点,其优化研究也开始在我国迅速兴起。如何优化农业供应链,提高农产品的物流效率,增加利润成为一个重要的研究课题。本文在对农业供应链及其优化管理方面作系统分析的基础上,针对供应链中所需考虑的两个核心问题——选址和配送问题从模型和求解方法上进行了探讨研究。本文首先介绍了农业供应链和物流中心的概念,分析了在现代物流观点下,农业供应链管理的发展对物流中心功能设计及选址的影响,探讨了选址和配送决策应掌握的基本原则,并对物流中心选址和配送方法中的数学模型和方法进行了系统分析。其次,在原有小生境遗传算法的基础上,为提高算法中种群的多样性和增强最优解的局部搜索能力,引入了模拟退火机制,提出一种改进的小生境遗传模拟退火算法(SNGA)。对于选址问题,从农产品易腐蚀的特性出发,提出了更具实用性的综合考虑运输成本、固定投资成本、管理成本和腐蚀农产品的损失费用的数学模型,并用前面提出的改进的小生境遗传模拟退火算法对实例进行了求解分析,结果证明该改进算法的有效性,使得算法的可应用性得到验证。在配送问题中,探讨了旅行商和车辆调度问题的数学模型和方法。由于旅行商(TSP)是一个典型的NP难题(Karp,1972),人们每提出一种新的寻优方法,总是力图用TSP来检验。本文以48个城市的TSP为例,用原有的简单遗传算法、模拟退火算法和本文提出的改进的小生境遗传模拟退火算法进行了matlab的编程实验仿真,分析比较了该改进算法相对前面两种算法的求解优劣性,得到该改进算法的局部搜索能力增强和种群的多样性增加的结果;并在求解带时间窗的车辆调度问题中,通过与简单遗传算法和2-opt遗传算法的结果比较分析,进一步验证了该改进算法的求解优劣性。最后,本文在进行总结的同时还提出了今后进一步研究和改进的方向。更多还原

【Abstract】 Recent years,with the process that we became a member of WTO and thedevelopment of the agricultural production in our country,agriculturalproducts in development countrys approach the market of ourcountry.Agricultural development in our country has step into a new stage whichwe need to husband and improve the setting of the constituents we have putin the agriculture ,so our country cries for the the realization of themodernization agriculture.Whereas this process requires leading new supplychain ideas and its management into the areas of the agriculture. It canreinforce the competitive of the agricultural market,augment farmer’s income,speed up modernization of our agricultural.Supply chain in agriculture hasbecoming one of the major industries in our economic development,which leadsto our new point of economic growth,and its optimization study springs uprapidly.How to optimize the supply chain in agriculture,and raise thelogisticsefficiency of the agriculture products, increase profits asthe most importantresearch subject.Elaborating on the knowledge of the supply chain in agriculture and itsoptimization,this paper attempts to take location and delivery as the two coreproblem for further research from the models and approaches.After that,on account of keeping the varity of the population,meanwhile,improving the abilities of the local research, we rise a modified niche geneticsimulated annealing algorithm(SNGA) based on simple niche genetic algorithm.The notion of supply chain in agriculture and logistics centers areintroduced first in this paper,then from view of modern logistics,we analysethe effect of developing supply chain management in agricultre on logisticscenters function design and location,after that wediscuss the fundamentalprinciples of location and delivery decision,and still further generalizemodels and approachs of the logistics centers location and delivery.On logistics centers location problem,concerning the decaying characterof agricultural products and the cost of transportation,fixed investment ,management ,demaged agriculture products, we set up a model with the virtueof being practical,then the availability of model is illustrated with apractical exampleby genetic algorithms, which overcomes some shortcomings and incarnatesthe more objective process.On delivery problem,we introduce the models of the TSP(Traveling SalesmanProblem),VRP(Vehicle routing problem).On account of NP-hardproblems,whenpeople rise a new optimized algorithms,they alwayse inspect it with the TSP.Theexample about TSP of 48 cities shows thatkeeping more varities of thepopulation and ability of local search comparing SGA (Simple Genetic Algorithm)and SA (Simulated Anneling Algorithm),and still further after solving theVRP with time windows by the modified algorithm and comparing with SGA ahd2-opt SGA,the performance of the ability to sovling problems has been greatlyimproved.At the end of this paper,some aspects which can be improved by furtherstudy are advanced.更多还原