BP神经网络及其在供应商选择评价中的应用.doc

PAGE PAGE 15 BP神经网络及其在供应商选择评价中的应用 钱芝网1、2 （1.上海理工大学，上海 200093；2.上海医疗器械高等专科学校，上海 200093） 摘要：供应商选择评价方法是否得当，将直接决定了企业所选供应商的优劣，并将最终影响到整个供应链的竞争能力。论文首先基于二十一世纪采购环境的变化及其对采购供应提出的新要求，结合前人的研究成果，构建了一套由质量、成本、服务、柔性、可靠性、管理水平、研究与发展能力这7个一级指标和交货柔性、协调能力、信息化水平等27个二级指标组成的供应商选择评价指标体系，并对这些指标的内涵进行了分析。考虑到供应商选择指标之间呈现出复杂的非线性关系，而诸多的评价方法中，唯有人工神经网络具有独特的非线性适应性信息处理能力，据此论文以此为工具，运用前向神经网络中的BP神经网络设计了一套算法，并基于该算法建立了供应商选择评价模型，最后进行了实例分析。结果表明，运用这一评价模型来选择应商，能较好地避免其他方法人为计取权值和相关系数过程中的主观影响和不确定性，选择的结果比较客观、科学。 关键词：BP神经网络；供应商选择；输入向量；网络收敛 On the BP Neural Network and its Application in Supplier Selection and Evaluation Qian Zhiwang1、2 （1.University of Shanghai for Science Technology，Shanghai 200093；2.Shanghai Medical Instrumentation College，Shanghai 200093） Abstract: An appropriate method of supplier selection and evaluation will directly determine the quality of the suppliers selected by the enterprise, and ultimately affect the competition ability of the whole supply chain. Firstly, based on the changes to purchasing environment and the new requirements proposed for purchasing in the 21st century, with the achievements of previous researches, this paper presents a system of supplier selection of evaluation that includes 7 standards such as quality, cost, service, flexibility and reliability, management level, research and development, and 27 performance indicators such as delivery flexibility, coordination, Informationization, etc., and analyses the connotation of these indicators. Considering the complex nonlinear relations among the supplier selection indicators, the artificial neural network is the only method that enables nonlinear adaptive information processing in all methods. Thus, this paper designs a set of algorithms by using the BP neural network in the feed forward ones, and builds up an evaluation model followed by the analyzed examples. The findings show that this evaluation model to select supplier can be relatively objective and scientific, avoiding some subjective influence and uncert