基于RS和SVM的风电总装企业物料分类研究与应用

发布时间：2018-06-01 17:38

本文选题：物料分类 + 风电总装企业　；参考：《重庆大学》2011年硕士论文

【摘要】：物料管理是在解决第二次世界大战中航空工业出现的难题时提出的概念,它是从整个公司的角度来解决物料问题,包括:协调不同供应商之间的协作,使不同物料之间的配合性和性能表现符合设计要求;提供不同供应商之间以及供应商与公司各部门之间交流的平台;控制物料流动率等。物料管理可以有效解决库存控制的矛盾等问题,缓解物料成本的压力,提高企业生产管理效率。物料分类作为物料管理的基础,其管理水平的高低在很大程度上决定了风电总装企业生产管理水平以及生产效率的高低,同时,物料分类管理还涉及到供应商的分类选择、采购、库存控制、物料管理等多个方面。因此,高效科学的物料分类管理对于风电组装企业具有重大的现实意义。本文拟对风电总装企业物料分类模型及其算法进行研究,提出基于粗糙集(RS)和支持向量机(SVM)的物料分类模型,期望解决样本数据较少和指标过多的问题。论文的主要研究内容安排如下: 首先,论文对国内外物料管理、物料分类的研究现状进行综述,阐述了本文的主要研究内容及总体框架,提出了本文的创新点。其次,研究了常用物料分类方法、粗糙集以及支持向量机的相关基础理论。再次,构建了风电总装企业物料属性评估指标体系,提出了运用粗糙集(RS)进行属性约减和支持向量机(SVM)设计分类器进行物料分类的技术思路,即应用粗糙集方法约简冗余属性,简化训练集和测试集,降低SVM分类器设计的复杂程度,提高SVM分类器的分类精度。上述的技术方案既充分利用了SVM较好的抑噪能力,也弥补了粗糙集对噪声敏感的缺点,增强了分类器的容错及抗干扰能力。在此基础上,论文建立了基于RS和SVM的物料分类模型,并对模型求解过程进行了详细说明。最后,基于上述理论研究,将建立的模型应用于某企业的实际工程中,对分类结果进行对比分析,结果表明基于RS和SVM的物料分类模型的分类精度可以达到较高水平,该模型具有一定工程应用价值。
[Abstract]:Material management is the concept of solving the problem of the aviation industry in the Second World War , which is to solve the material problem from the angle of the whole company . It includes : coordinating the cooperation among different suppliers , making the cooperation and performance of different materials meet the design requirements ; providing the platform for the communication between different suppliers and between the supplier and the company ; controlling the material flow rate , etc . The material management can effectively solve the contradiction of inventory control and the like , relieve the pressure of the material cost , and improve the production management efficiency of the enterprise .

As the basis of material management , the level of management of materials determines the level of production management and production efficiency of wind power assembly enterprises .

The main research contents of the thesis are as follows :

Firstly , this paper summarizes the research status of material management and material classification at home and abroad , expounds the main research contents and overall framework of this paper , and puts forward the innovation points of this paper .

Secondly , the general material classification method , rough set and related basic theory of support vector machine are studied .

In this paper , an index system for evaluating the material properties of a wind power assembly enterprise is constructed , and the technical idea of using the rough set ( RS ) to carry out the classification of materials by using the attribute reduction and support vector machine ( SVM ) design classifier is put forward , namely , the rough set method is applied to reduce the complexity of the design of the SVM classifier and improve the classification precision of the SVM classifier .

Finally , based on the above - mentioned theoretical research , the model is applied to the actual project of an enterprise , and the classification result is compared and analyzed . The results show that the classification accuracy of the material classification model based on RS and SVM can reach a higher level , and the model has certain engineering application value .
【学位授予单位】：重庆大学
【学位级别】：硕士
【学位授予年份】：2011
【分类号】：TH186;TP18

【相似文献】