锻压技术

基于AHP和线性神经网络的模具标准评价

英文标题：Evaluation of mould standard based on the AHP and linear neural network

作者：郭颖颖廖宏谊

单位：桂林电子科技大学

分类号：TG76

出版年,卷(期):页码：2014,39(11):150-155

摘要：

在模具标准制修订工作中，是否采用国际标准或国外先进标准、预测标准对国情的适用性是重要环节。通过对我国现行模具标准的大量分析，确定标准评价因素，构建标准评价体系，运用层次分析法（AHP）计算各因素的重要程度值，并做量化分析和评分，以获得线性神经网络模型样本，进而对样本进行训练、验证，最终获得模具标准评价模型。该模型充分吸收了专家的知识和经验，降低了评价的人为因素。结果表明，基于AHP和线性神经网络的模具标准评价方法计算的最大相对误差为1.6%，该评价方法正确可行。

Whether to adopt international standards or advanced foreign standards and predicting the applicability of standard to the national conditions are important links in the mould standard system revision work. Through a large number of analyses of our current mould standard, the standard evaluation factors and the standard evaluation system were determined, and the importance of the various factors were calculated by using the analytic hierarchy process. And then by doing quantitative analysis and evaluation, the samples of the linear neural network model were obtained, and after training and validation samples, mould standard evaluation model was gained ultimately. By this, not only the knowledge and experience of the experts can be absorbed completely by the model, but also the evaluation of human factors has gone down. As the experiment shows, the evaluation result of maximum relative error is 1.6%, the method of the mould standard evaluation based on the AHP and linear neural network is correct and feasible.

基金项目：

国家标准化管理委员会“装备制造业重点领域标准体系研究计划”项目

郭颖颖（1988-），女，硕士研究生

参考文献：

［1］李荣发. 优先数和优先数系简介[J]. 洪都科技, 1979,（4）: 15-32.Li R F. Priority number and priority number system introduction [J]. Journal of Hongdu Science and Technology, 1979, （4）:15-32.
［2］李春田.标准化概论[M]. 北京：中国人民大学出版社, 1987.Li C T. Introduction to Standardization[M]. Beijing: Renmin University of China Publishing House, 1987.
［3］金颖. 技术标准《高速工具钢锻件技术条件》的编制及解读[J].锻压技术，2013,38(3):179-180.Jin Y. Preparation and amendment of technical standard of technical requirements for forgings of high-speed tool steel [J]. Forging & Stamping Technology, 2013, 38(3):179-180.
［4］秦季冬. 基于AHP和BP神经网络的广西大中型工业企业技术创新能力评价研究[D]. 南宁：广西大学，2010. Qin J D. Based on AHP and BP Neural Network of Large and Medium-sized Industrial Enterprises in Guangxi Technological Innovation Ability Evaluation Research [D]. Nangjing: Guangxi University, 2010.
［5］李成. 神经网络系统理论[M]. 西安: 西安电子科技大学出版社, 1990.Li C. Neural Network System Theory[M]. Xian: Xian University of Electronic Science and Technology Press, 1990.
［6］区健芬, 陈贤宣. 基于 AHP-BP 的虚拟经营风险评价[J]. 现代商业, 2011, (9): 136-137. Qu J F, Chen X X. Virtual operation risk assessment based on AHP and BP [J]. Journal of Modern Business, 2011,(9): 136-137.
［7］陈明. MATLAB 神经网络原理与实例精解[M]. 北京: 清华大学出版社, 2013. Chen M. MATLAB Neural Network Theory and Example Extract Solution [M]. Beijing: Tsinghua University Press, 2013.
［8］高梁. 基于层次分析法的绩效评估权重设计[D]. 乌鲁木齐:新疆大学, 2007.Gao L. Performance Evaluation Based on Analytic Hierarchy Process (ahp) Weight Design [D]. Urumuqi: Xinjiang University, 2007.
［9］洪志国, 李焱. 层次分析法中高阶平均随机一致性指标 (RI) 的计算[J]. 计算机工程与应用, 2002, 38(12): 45-47. Hong Z G, Li Y. Analytic hierarchy process (AHP) and middle mean random consistency index (RI) calculations [J]. Computer Engineering and Application, 2002, 38 (12) : 45-47.
［10］钟义山. 正态分布[J]. 陕西林业科技, 1988, (2): 31-33.Zhong Y S. Normal distribution [J]. Journal of Shanxi Forestry Science and Technology, 1988, (2):31-33.
［11］丛爽. 面向 MATLAB 工具箱的神经网络理论与应用[M]. 合肥:中国科学技术大学出版社, 1998.Cong S. Facing the Toolbox of MATLAB Neural Network Theory and Application [M]. Hefei: University of Science and Technology of China Press, 1998.

服务与反馈：

【文章下载】【加入收藏】