面向顾客需求的在线匹配定制方法

发布时间：2018-06-24 19:05

本文选题：定制配置 + 匹配　；参考：《重庆大学学报》2017年03期

【摘要】：在线定制的产品涉及到设计、生产等复杂环节,所以定制品的方案设计及其可行性等信息的快速反馈是实施在线定制的难点。提出基于CBR(case-based reasoning)的匹配定制方法,减少定制时间避免不必要的重复设计。首先运用层次分析法和质量功能展开确定需求工程设计权重;结合模糊数学修改高斯函数,构造出一种具有接近客观现实、方便处理、区分性强等优点的匹配度计算方法;利用Beta分布构建机器学习方法,得到随市场偏好变动而调整的匹配度阀值。其次针对顾客对不同产品的价格敏感度不同,引入调节因子修正最终匹配值对比匹配阀值从而获得定制产品的设计方案及可行性并反馈。最后以电冰箱定制为例,证明该方法有效且易于实施。
[Abstract]:The product of online customization involves complex links such as design and production, so the fast feedback of the information such as the scheme design and feasibility of the fixed product is the difficulty to implement online customization. A method of matching customization based on CBR (case-based reasoning) is proposed to reduce the customization time and avoid unnecessary repeated design. Firstly, the weight of demand engineering design is determined by AHP and quality function expansion, and the Gao Si function is modified by fuzzy mathematics to construct a matching calculation method which has the advantages of close to objective reality, convenient processing and strong differentiability. By using Beta distribution to construct the machine learning method, the matching threshold adjusted with the change of market preference is obtained. Secondly, according to the different price sensitivity of the customer to different products, the adjustment factor is introduced to revise the final matching value to compare the matching threshold so as to obtain the design scheme and feasibility of the customized product and feedback. Finally, taking the refrigerator customization as an example, it is proved that the method is effective and easy to implement.
【作者单位】：重庆大学机械传动国家重点实验室;重庆大学现代物流重庆市重点实验室;重庆大学机械工程学院;
【基金】：国家科技支撑计划资助项目(2015BAF05B03)~~
【分类号】：TP181;F273;F274

【相似文献】