基于BP神经网络的徐工挖掘机造型设计评价研究

发布时间：2018-12-16 08:49

【摘要】：随着科技的发展,工程机械行业单纯依靠技术手段占领市场已经日趋艰难,此时,实施品牌战略的方法,在社会公众中创造良好的品牌意象是一种有效的提高竞争力的手段。如何有效的评价产品的造型与企业品牌基因的一致性是本文的主要研究内容。本文以徐工集团旗下的挖掘机产品为例探讨了徐工挖掘机造型品牌性评价。人工神经网络可以模仿人脑的功能,用大量简单的处理单元构建了广泛连接的复杂网络,这种网络对于模式化知识的映射处理具有巨大的优势,可以无限逼近任意函数,具备了模糊推理能力。本文利用人工神经网络的智能特性可以更大限度的减小专家主观性评价的影响,为徐工挖掘机产品造型与品牌基因一致性的精确客观性评价提供了一个科学的方法。首先,从品牌基因的角度出发,探讨了工程机械行业的品牌基因,进而通过对徐工集团品牌基因的分析构建了徐工品牌工程机械造型设计评价指标。接下来叙述了神经网络的类型并且总结了BP神经网络的基本结构、学习算法、计算步骤,然后以造型特征为出发点分析了工程机械品牌造型特征并探讨了工程机械品牌造型特征提取的一般方法。然后,着重探讨了挖掘机造型特征的提取过程,先分析了各国著名工程机械品牌的挖掘机造型特点,根据工程机械品牌造型特征提取的一般方法研究了徐工挖掘机造型特征的具体提取方法,总结了徐工挖掘机品牌造型特征的16个要素,最终初步构建了基于BP网络模型的徐工挖掘机造型设计评价体系的输入层神经元和输出层神经元。根据BP网络的计算方法确定了神经网络模型的结构,根据模型结构、输入层、输出层数据构建了基于BP网络的徐工挖掘机造型评价模型,通过训练、测试证明了网络模型评价的实用价值。最后,通过问卷调查及神经网络模型评价发现原有XE75G型挖掘机造型上的不足,根据徐工挖掘机的品牌造型特征分析以及产品的一般设计方法,对该款挖掘机进行再设计,说明了再设计中的主要设计点,最终利用已构建的基于BP神经网络的徐工挖掘机造型评价模型对再设计后的挖掘机产品造型进行了评价。分析及优化结果表明,本文所构建的基于人工神经网络的徐工挖掘机造型设计评价系统可以较为准确的对徐工挖掘机造型进行品牌性相关的评价,并能根据评价结果对设计方案进行有针对性的修改指导。希望本文的研究能为类似工程机械产品乃至其它行业产品的造型与品牌基因一致性的评价及设计优化提供一定理论参考和方法指导。
[Abstract]:With the development of science and technology, it has become more and more difficult for the construction machinery industry to occupy the market by simply relying on technical means. At this time, the method of implementing brand strategy and creating a good brand image in the public is an effective means to improve competitiveness. How to evaluate the consistency of product modeling and corporate brand gene effectively is the main research content of this paper. This paper discusses the brand evaluation of Xugong excavator by taking the excavator products of Xugong Group as an example. The artificial neural network can imitate the function of human brain and construct a complex network with a large number of simple processing units. This kind of network has great advantages for the mapping of modular knowledge and can be approximated to any function indefinitely. It has the ability of fuzzy reasoning. This paper makes use of the intelligent characteristic of artificial neural network to reduce the influence of expert subjective evaluation to a greater extent, and provides a scientific method for accurate and objective evaluation of product modeling and brand gene consistency of Xugong excavator. Firstly, from the perspective of brand gene, the brand gene of construction machinery industry is discussed, and then the evaluation index of Xugong brand construction machinery modeling design is constructed by analyzing the brand gene of Xugong Group. Then it describes the types of neural network and summarizes the basic structure, learning algorithm, calculation steps of BP neural network. Then, the modeling features of construction machinery brand are analyzed based on the modeling features and the general methods of extracting the brand modeling features of construction machinery are discussed. Then, the extraction process of excavator modeling features is discussed emphatically, and the characteristics of excavator modeling of famous construction machinery brands in various countries are analyzed first. According to the general method of extracting the brand modeling features of construction machinery, this paper studies the specific methods of extracting the modeling features of Xugong excavators, and summarizes the 16 elements of the brand modeling features of Xugong excavators. Finally, the input layer neuron and output layer neuron of Xugong excavator modeling evaluation system based on BP network model are preliminarily constructed. According to the calculation method of BP network, the structure of neural network model is determined. According to the data of model structure, input layer and output layer, the model of model evaluation of Xugong excavator based on BP network is constructed. The test proves the practical value of network model evaluation. Finally, through questionnaire investigation and neural network model evaluation, we find that the original XE75G excavator model is insufficient. According to the analysis of the brand modeling characteristics of Xugong excavator and the general design method of the product, the excavator is redesigned. The main design points in redesign are explained. Finally, the model of Xugong excavator based on BP neural network is used to evaluate the product modeling of redesigned excavator. The results of analysis and optimization show that the evaluation system of Xugong excavator based on artificial neural network can be used to evaluate Xugong excavator modeling accurately. According to the evaluation results, the design can be modified and directed. It is hoped that the research in this paper can provide some theoretical reference and method guidance for the evaluation and design optimization of the genetic consistency of modeling and brand in similar construction machinery products and other industries.
【学位授予单位】：中国矿业大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TP183;TB472

【相似文献】