关键词: 质量功能展开, 质量特性, 实现水平, 证据推理, 多目标优化, 协同决策, 遗传算法

Abstract: To effectively and flexibly manage imprecise information in Quality Function Deployment(QFD)and to tackle multi-object optimization problem, information provided by each QFD product developer were regarded as information source of the evidence reasoning theory, and then related reasoning algorithm was extended to fuse group information source so as to obtain a collaborative decision consensus result. The multiple constraints and multiple objective optimization models were constructed. Aims of models were to obtain maximum customer satisfaction degree, minimum cost & difficulty for quality characteristics implementation level. The improved Non-dominated Sorting Genetics Algorithm(NSGA-Ⅱ)was employed to acquire the Pareto solutions set of implementation levels of quality characteristics. And fuzzy optimal selection technique was adopted to select the best solution from the Pareto solutions set. Finally, development of large-scale deep cooling air-separating equipment in a company as the case was given to demonstrate the application and validation of the proposed method.

Key words: quality function deployment, quality characteristics, fulfillment levels, evidence reasoning, multi-object optimization, collaborative decision, genetic algorithms