基于RBF-GA的供应链质量成本优化研究

发布时间：2018-04-25 19:08

本文选题：供应链 + 质量成本优化　；参考：《福州大学》2014年硕士论文

【摘要】：质量管理是现代企业生产经营的关键点。在当前激烈的竞争市场经济环境中,供应链之间的角逐早已取代了企业之间的竞争。因此在供应链质量管理的背景下,具备成本优势已成为供应链赢得竞争优势的重要手段。本文在供应链环境下研究质量成本,既充分利用了供应链集成优势,又弥补了传统质量成本管理在企业互动和相互合作方面的缺陷。通过探讨供应链质量成本优化,为企业追求低成本下的高质量产品提供了新的思路,也为供应链经营管理中的质量成本优化实践提供了相关的理论参考。供应链质量成本优化包括供应链企业内部和供应链企业之间的质量成本优化。本文选择在单个企业质量成本优化的基础上探讨相邻企业间的质量成本优化方法,因而引入供应商质量成本来定义和解决两阶段供应链质量成本优化问题。供应商质量成本优化从相关定义、科目核算和优化流程等三个方面进行详细论述。在供应商质量成本涵义中,首先确定以顾客满意为标准的质量水平,同时考虑供应链质量风险传递性特征,引入下游企业质量成本影响系数。随后在质量水平与下游企业质量成本影响系数这两个变量的基础上构建了供应商质量成本核算科目表,通过采集供应商质量成本核算科目表得到质量成本优化数据。为了能详细了解质量水平、下游企业质量成本影响系数与总质量成本之间的函数关系,本文采用径向基神经网络来逼近这种关系。以供应商质量成本核算科目表数据,作为RBF神经网络的训练数据,在满足精度要求的拟合网络中通过样本数据测试网络的精度。遗传算法善于处理传统计算方法难于解决的复杂系统优化问题。在基于遗传算法的质量成本函数优化中,设定目标优化质量成本值与径向基神经网络预测输出值的绝对值差为遗传算法的个体适应度函数,经过遗传操作后,得到满足目标质量成本值的最优质量水平与下游企业质量成本影响系数。最后,通过一个实证分析来证实基于RBF-GA的供应链质量成本优化模型的有效性,总结出供应链企业可以通过调节质量水平与下游企业影响系数来确定质量成本控制效果,为企业进行供应链质量成本科目核算提供了一定的依据。经过多次仿真说明,本文提出的非线性系统寻优方法能以较快的收敛速度找到近似最优解,结果证明径向基神经网络和遗传算法来寻求供应链质量成本优化的方法是可行的。
[Abstract]:Quality management is the key point of modern enterprise production and management. In the current competitive market economic environment, the competition between supply chains has already replaced the competition between enterprises. Therefore, under the background of supply chain quality management, having the cost advantage has become an important means to win the competitive advantage of the supply chain. The study of quality cost not only makes full use of the advantages of supply chain integration, but also makes up for the shortcomings of traditional quality cost management in the interaction and cooperation of enterprises. By discussing the optimization of quality cost of supply chain, it provides a new idea for enterprises to pursue high quality products under low cost, and also optimizes the quality cost in the management of supply chain. The practice provides the relevant theoretical reference. The quality cost optimization of supply chain includes the quality cost optimization between the internal and supply chain enterprises of the supply chain. This paper selects the method of quality cost optimization between adjacent enterprises on the basis of the optimization of individual enterprise quality cost, thus introducing the quality cost of suppliers to define and solve two. The quality cost optimization of stage supply chain is discussed in detail from three aspects: related definition, subject accounting and optimization process. In the meaning of supplier quality cost, the quality level of customer satisfaction is first determined, and the quality risk of supply chain is considered, and the downstream enterprise is introduced into the supply chain. On the basis of the two variables, the quality cost and the quality cost influence coefficient of the downstream enterprise, the supplier quality cost accounting subject table is constructed, and the quality cost optimization data are obtained by collecting the supplier quality cost accounting subject table. In order to understand the quality level in detail, the quality cost of the downstream enterprise is reflected. The function relationship between the RBF and the total quality cost is used in this paper. This paper uses the RBF neural network to approach this relationship. As the training data of the RBF neural network, the accuracy of the network is tested by the sample data in the fitting network satisfying the precision requirements. It is difficult to solve the problem of complex system optimization. In the optimization of the quality cost function based on genetic algorithm, the absolute value difference between the optimal quality cost value of the target and the predicted output value of the radial basis neural network is the individual fitness function of the genetic algorithm. After the genetic operation, the optimal quality is obtained to meet the quality cost value of the target. In the end, the effectiveness of the supply chain quality cost optimization model based on RBF-GA is confirmed by an empirical analysis. It is concluded that the supply chain enterprises can determine the quality cost control effect by adjusting the quality level and the influence coefficient of the downstream enterprises, so as to carry out the supply chain quality cost for the enterprises. After many simulations, the nonlinear system optimization method proposed in this paper can find the approximate optimal solution at a faster rate of convergence. The results prove that the RBF neural network and genetic algorithm are feasible to seek the quality cost optimization of the supply chain.

【学位授予单位】：福州大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：F274;F275.3

【相似文献】