不确定用户信息驱动的产品平台创新设计研究

发布时间：2018-10-30 18:47

【摘要】：在大批量定制的生产方式下，企业成功的关键在于能否抓住不确定的用户需求信息。也就是说，用户在企业的经营决策中扮演着非常重要的角色。随着科技进步和产品竞争的加剧，企业对由不确定用户驱动的、以产品平台为基础的产品设计理论及其实践产生了迫切的需求。本论文针对大批量定制环境下，由不确定用户驱动的产品平台设计进行了研究，主要包含以下几部分内容：（1）获取不确定用户需求信息后，对其进行分析和处理。根据用户需求信息具有模糊语义、缺失信息、冗余信息的特点，首先采用模糊集对模糊语义信息做量化处理，，再利用模糊集中的隶属度将其离散化；然后，采用粗糙集对需求信息做粒化处理以补全缺失信息，再利用其属性约简消除需求信息中的冗余信息。这将为后续映射产品设计参数做准备。（2）使用TRIZ理论拓展用户需求。构建产品特性模型，并以注射机产品为例，建立产品实例库。使用NN聚类分析中的广义欧氏距离，比较需求信息与产品实例库中的产品信息，在产品实例库中选取较接近的多个实例产品。建立了基于F-AHP的评价模型对其进行综合评价，以选出用户最满意的产品。（3）对应于产品设计参数为连续数值，建立SVR（回归）映射模型；为离散数值，建立SVC（分类）映射模型。使用TRIZ理论解决在产品映射过程中产生的冲突，并根据冲突消解方案调整产品设计参数。然后，使用LSSVC聚类方法对产品进行模块化设计，并通过产品设计参数和模块之间的关系，对相似实例产品的模块进行修改、调整或变型设计。（4）以注射机产品为例，通过对不确定用户提出订单的处理，展开了基于-NN注射机产品的配置设计过程，并基于F-AHP对其配置方案进行了评价。在用户不满意的情况下，展开了基于粗糙集+支持向量机映射模型，注射机产品的定制设计过程。通过注射机产品实例，验证了本论文面向不确定用户需求信息，以产品平台为基础的，产品配置设计和定制设计方法研究的正确性和实用性。
[Abstract]:In mass customization production mode, the key to enterprise success lies in whether or not to grasp the uncertain user demand information. In other words, users play a very important role in business decisions. With the development of science and technology and the aggravation of product competition, there is an urgent need for the theory and practice of product design based on product platform driven by uncertain users. In this paper, the design of product platform driven by uncertain users in mass customization environment is studied. The main contents are as follows: (1) after obtaining the information of uncertain user requirements, it is analyzed and processed. According to the characteristics of fuzzy semantics, missing information and redundant information, the fuzzy set is used to quantify the fuzzy semantic information, and then the membership degree of the fuzzy set is used to discretize the fuzzy semantic information. Then, the requirement information is granulated by rough set to complement the missing information, and then the redundant information is eliminated by attribute reduction. This will prepare for subsequent mapping of product design parameters. (2) using TRIZ theory to expand the user's demand. The product characteristic model was constructed, and the product case library was established by taking the injection machine product as an example. Using the generalized Euclidean distance in NN clustering analysis, the requirement information is compared with the product information in the product case library. A comprehensive evaluation model based on F-AHP is established to select the most satisfied products. (3) the SVR (regression) mapping model and the SVC (classification) mapping model are established for discrete values corresponding to the continuous values of product design parameters. The TRIZ theory is used to solve the conflicts in the process of product mapping, and the product design parameters are adjusted according to the conflict resolution scheme. Then, the modular design of the product is carried out by using LSSVC clustering method, and through the relationship between the product design parameters and the module, the module of the product with similar examples is modified, adjusted or modified. (4) taking the injection machine product as an example, the configuration design process based on-NN injection machine product is carried out, and the configuration scheme is evaluated based on F-AHP. Based on the rough set support vector machine mapping model, the customized design process of injection machine product is carried out in the case of user dissatisfaction. Through the product example of injection machine, the correctness and practicability of product configuration design and customization design method based on product platform are verified in this paper.
【学位授予单位】：河北工业大学
【学位级别】：博士
【学位授予年份】：2014
【分类号】：TB472

【参考文献】