项目名称: 基于QFD和数据挖掘的卷烟产品叶组配方优化关键技术研究
项目编号: No.61273204
项目类型: 面上项目
立项/批准年度: 2013
项目学科: 自动化技术、计算机技术
项目作者: 唐加福
作者单位: 东北大学
项目金额: 81万元
中文摘要: 本项研究结合卷烟产品叶组配方的特征和领域知识,在前期研究成果的基础上,研究如何采用系统工程的思想、结合QFD技术、优化技术、数据挖掘技术,充分利用积累的知识、经验和数据,解决卷烟产品叶组配方优化设计中的一些关键科学和技术问题。建立卷烟叶组配方优化设计的QFD过程模型;建立顾客需求(烟民吸食感官指标或烟气指标)与卷烟产品(叶组配方)的理化指标间的关系模型;建立配方单元烟气指标与单料烟烟气指标之间关系的数据模型(预测评估模型);建立配方单元理化指标与单料烟理化指标之间的数据模型(预测评估模型);建立配方单元的粒度和质量特征的描述和表达方法;建立基于数据和模式的产品规划(烟叶物理化学指标值)质量屋、配方结构质量屋和配方单元(叶组配方)质量屋的优化模型和求解算法。不仅从理论上解决卷烟产品叶组配方优化中的关键技术问题,而且为我国烟草企业构筑中式卷烟核心技术,开展卷烟配方辅助设计提供方法和技术支持。
中文关键词: 质量功能展开;数据挖掘;优化方法;质量设计;卷烟叶组配方
英文摘要: This project aims to address key optimization theory and methods for the cigarette formula design problem. Based on the research findings from our previous projects and the ideas of system engineering, quality function deployment (QFD), optimization technology and data mining will be applied for optimal design of cigarette formula with consideration of the characteristics and domain knowledge of cigarette products and by utilizing the knowledge, experience and history data on cigarette formula design. The purpose of the project is to: develop a QFD process model for cigarette formula design optimization; establish the relation model for customer requirements (sensory quality or smoke indices) and physical & chemical components of cigarette product (cigarette formula); develop a data-driven relation model (predictive evaluation model) for smoke indices between unblended cigarette and formula unit; present a formula and expression method for granularity and quality characteristics of cigarette formula unit; develop optimization models based on data and schemes and solving algorithms for product planning HoQ (tobacco physical & chemical indices), formula structure HoQ and formula unit (cigarette formula) HoQ. The research work of this project will solve the crucial technical problems in cigarette formula design opt
英文关键词: Quality function deployment;Data mining;Optimization methods;Quality design;Cigarette formula