基于BP神经网络的通信设备供应商评价与选择

发布时间：2018-10-09 21:23

【摘要】：在供应链管理日趋重要的今天,越来越多的企业开始意识到与优秀的供应商建立长久合作的关系对整个供应链的运作具有相当重要的意义。当前虽然国内已有部分学者针对通信设备行业供应商评价进行研究,但是评价指标较为陈旧,针对某企业进行系统的研究较少,且评价方法主观成分较多,不足以为我国通信设备行业供应商评价选择提供足够的理论依据。鉴于以上问题,本文以M公司为研究对象,对其供应商评价选择问题进行了详细的研究。首先,本文对国内外文献进行了深入的研究,阐述了供应商评价选择的重要性,对国内外供应商评价与选择的研究现状进行了详细的说明,并对BP神经网络进行了必要的解释。其次,本文对通信设备行业的特点和供应商评价现状进行了分析,并对M公司当前供应商评价选择存在的问题进行研究。在此基础上本文通过阅读大量文献、对相关人员发放调查问卷和因子分析等过程重新构建了 M公司供应商评价指标体系,并对各个指标进行详细说明。最后,本文根据已经建立的通信设备企业的供应商评价指标体系,运用BP神经网络算法建立供应商评价模型。通过MATLAB对该供应商评价模型进行了实现,利用收集到的大量供应商数据,对模型进行训练,选择合适的各项参数,再对测试数据进行误差分析,确定模型的拟合性和预测能力良好。运用该模型对目标供应商进行仿真应用,得出评价结果,通过仿真结果对M公司的供应商选择提供参考依据。
[Abstract]:With the increasing importance of supply chain management, more and more enterprises begin to realize that it is very important for the operation of the whole supply chain to establish a long-term cooperative relationship with excellent suppliers. At present, some domestic scholars have studied the supplier evaluation of communication equipment industry, but the evaluation index is relatively old, the systematic research for a certain enterprise is less, and the subjective component of the evaluation method is more. It is not enough to provide sufficient theoretical basis for supplier evaluation and selection of communication equipment industry in China. In view of the above problems, this paper studies M Company's supplier evaluation and selection in detail. Firstly, this paper makes a deep research on domestic and foreign literature, expounds the importance of supplier evaluation and selection, explains the research status of supplier evaluation and selection at home and abroad in detail, and gives a necessary explanation to BP neural network. Secondly, this paper analyzes the characteristics of communication equipment industry and the status quo of supplier evaluation, and studies the problems existing in M Company's current supplier evaluation and selection. On the basis of this, this paper reconstructs the supplier evaluation index system of M company by reading a lot of literature, and explains each index in detail. Finally, according to the established supplier evaluation index system of communication equipment enterprises, the BP neural network algorithm is used to establish the supplier evaluation model. The supplier evaluation model is realized by MATLAB. The model is trained by a large number of supplier data collected, and the appropriate parameters are selected, and then the error analysis of the test data is carried out. The fitting ability and prediction ability of the model are good. The model is used to simulate the target suppliers, and the evaluation results are obtained. The simulation results provide a reference for the supplier selection of M Company.
【学位授予单位】：昆明理工大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：F274;F426.6

【参考文献】