到货时间不确定的多商品订购批量优化模型及算法研究
发布时间：2019-03-12 21:06
【摘要】：随着全球竞争的日益加剧,以客户为中心的供应链管理模式逐步取代了以生产和产品为中心的传统管理模式。采购管理作为供应链管理中的重要组成部分,对整条供应链起到了价值创造以及增长的作用。企业对采购活动的有效管理不仅可以降低运营成本,增强市场竞争力,而且可以保证其在快速多变的市场中处于领先地位。本文首先对供应链管理理论中库存管理和采购批量理论以及不确定理论进行了综述。总结了某营销商所处供应链的各个环节的特点及其自身的特点。分析了由于供应链的不确定性导致商品到货时间的不确定性,以及由到货时间不确定引起的到货量不确定,并将该问题转化为不确定的到货率处理。通过历史数据和专家经验值估计,将不确定的到货率拟合成随机模糊变量,提出了在下游客户订单需求确定而订单到货时间不确定的情况下,如何分配该营销商在全年多个订购周期同时订购多品种商品的批量,使企业得到最小化运营成本的优化模型。本文综合考虑订单的最低订购量限制、库存容量和流动资金的约束、必须满足客户需求等限制条件,基于不确定理论中的随机模糊理论,建立了到货时间不确定的多商品订购批量随机模糊优化模型；基于不确定规划中的期望值规划,将含有随机模糊变量的优化模型转化为随机模糊期望值模型进行求解。其次,针对所建立的模型,设计了模型求解的遗传算法。在算法设计中,针对问题的特点,设计了以商品各周期订购量为基因段的整数染色体编码方式；采用一种启发式方法实现种群的初始化；根据染色体编码的特点,设计了基于基因段的均匀交叉和基因段内的单点交叉算子,以及基于基因段的启发式变异和基因段内的互换变异算子；针对选择算子,本文混合采用轮盘赌的选择策略和精英选择策略；针对不可行染色体,设计了修复策略。最后算法采用C语言编程实现,结合企业实际部分数据,进行了模型的仿真实验。算法参数实验包括遗传算法的相关参数的仿真实验,如交叉变异算子对遗传算法性能的影响分析、遗传算法性能与收敛性的实验分析、不同选择策略的对比分析；模型参数实验包括对库存容量、流动资金、调货价格、不同折扣优惠力度以及到货时间的确定与否对运营成本的影响进行了详细分析,最后分析调货启动成本和调货可变成本对调货总成本的影响。仿真实验结果验证了本文所建立的模型和算法的有效性及可行性。
[Abstract]:With the increasing competition in the world, the customer-centered supply chain management model gradually replaces the traditional production and product-centered management model. Procurement management, as an important part of supply chain management, plays an important role in value creation and growth of the whole supply chain. The effective management of purchasing activities can not only reduce the operating cost, enhance the market competitiveness, but also ensure that it is in a leading position in the fast changing market. In this paper, the inventory management, procurement batch theory and uncertainty theory in the theory of supply chain management are reviewed. This paper summarizes the characteristics of each link of the supply chain where a certain marketer is located and its own characteristics. In this paper, the uncertainty of arrival time and the uncertainty of arrival quantity caused by the uncertainty of supply chain are analyzed, and the problem is transformed into uncertain arrival rate. Through the historical data and expert experience value estimation, the uncertain arrival rate is combined with random fuzzy variables, and it is proposed that when the downstream customer order demand is determined and the order arrival time is uncertain, How to allocate the distributor to order the batch of multi-product at the same time in multiple ordering cycles in the whole year, so that the enterprise can get an optimal model to minimize the operating cost. Based on the stochastic fuzzy theory of uncertainty theory, this paper considers the minimum order quantity limit of order, the restriction of inventory capacity and liquidity, and so on, which must satisfy the demand of customers, and is based on the stochastic fuzzy theory of uncertainty theory. A stochastic fuzzy optimization model of multi-product ordering batch with uncertain arrival time is established. Based on the expected value programming in uncertain programming, the optimization model with random fuzzy variables is transformed into a stochastic fuzzy expected value model to solve the problem. Secondly, according to the established model, the genetic algorithm for solving the model is designed. In the algorithm design, according to the characteristics of the problem, the integer chromosome coding method based on the order quantity of each period of the commodity is designed, and a heuristic method is used to realize the initialization of the population. According to the characteristics of chromosome coding, uniform crossover based on gene segment and single point crossover operator in gene segment are designed, heuristic mutation operator based on gene segment and exchange mutation operator within gene segment are designed. For the selection operator, the roulette selection strategy and elite selection strategy are used in this paper, and the repair strategy is designed for the infeasible chromosome. Finally, the algorithm is programmed in C language and combined with the actual data of the enterprise, the simulation experiment of the model is carried out. The parameters of the algorithm include the simulation experiments of the related parameters of the genetic algorithm, such as the influence of crossover mutation operator on the performance of the genetic algorithm, the experimental analysis of the performance and convergence of the genetic algorithm, and the comparative analysis of different selection strategies. The model parameter experiment includes a detailed analysis of the impact of inventory capacity, liquidity, transfer price, different discount rates and arrival time on operating costs. Finally, the influence of start-up cost and variable cost of transfer on the total cost of transfer is analyzed. The simulation results verify the validity and feasibility of the proposed model and algorithm.
【学位授予单位】：东北大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TP18;F274
本文编号：2439153
[Abstract]:With the increasing competition in the world, the customer-centered supply chain management model gradually replaces the traditional production and product-centered management model. Procurement management, as an important part of supply chain management, plays an important role in value creation and growth of the whole supply chain. The effective management of purchasing activities can not only reduce the operating cost, enhance the market competitiveness, but also ensure that it is in a leading position in the fast changing market. In this paper, the inventory management, procurement batch theory and uncertainty theory in the theory of supply chain management are reviewed. This paper summarizes the characteristics of each link of the supply chain where a certain marketer is located and its own characteristics. In this paper, the uncertainty of arrival time and the uncertainty of arrival quantity caused by the uncertainty of supply chain are analyzed, and the problem is transformed into uncertain arrival rate. Through the historical data and expert experience value estimation, the uncertain arrival rate is combined with random fuzzy variables, and it is proposed that when the downstream customer order demand is determined and the order arrival time is uncertain, How to allocate the distributor to order the batch of multi-product at the same time in multiple ordering cycles in the whole year, so that the enterprise can get an optimal model to minimize the operating cost. Based on the stochastic fuzzy theory of uncertainty theory, this paper considers the minimum order quantity limit of order, the restriction of inventory capacity and liquidity, and so on, which must satisfy the demand of customers, and is based on the stochastic fuzzy theory of uncertainty theory. A stochastic fuzzy optimization model of multi-product ordering batch with uncertain arrival time is established. Based on the expected value programming in uncertain programming, the optimization model with random fuzzy variables is transformed into a stochastic fuzzy expected value model to solve the problem. Secondly, according to the established model, the genetic algorithm for solving the model is designed. In the algorithm design, according to the characteristics of the problem, the integer chromosome coding method based on the order quantity of each period of the commodity is designed, and a heuristic method is used to realize the initialization of the population. According to the characteristics of chromosome coding, uniform crossover based on gene segment and single point crossover operator in gene segment are designed, heuristic mutation operator based on gene segment and exchange mutation operator within gene segment are designed. For the selection operator, the roulette selection strategy and elite selection strategy are used in this paper, and the repair strategy is designed for the infeasible chromosome. Finally, the algorithm is programmed in C language and combined with the actual data of the enterprise, the simulation experiment of the model is carried out. The parameters of the algorithm include the simulation experiments of the related parameters of the genetic algorithm, such as the influence of crossover mutation operator on the performance of the genetic algorithm, the experimental analysis of the performance and convergence of the genetic algorithm, and the comparative analysis of different selection strategies. The model parameter experiment includes a detailed analysis of the impact of inventory capacity, liquidity, transfer price, different discount rates and arrival time on operating costs. Finally, the influence of start-up cost and variable cost of transfer on the total cost of transfer is analyzed. The simulation results verify the validity and feasibility of the proposed model and algorithm.
【学位授予单位】：东北大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TP18;F274
【参考文献】
相关期刊论文 前1条
1 马士华;王福寿;;时间价格敏感型需求下的供应链决策模式研究[J];中国管理科学;2006年03期
，本文编号：2439153
本文链接：https://www.wllwen.com/guanlilunwen/gongyinglianguanli/2439153.html
