Methodological Research Of Quality Function Deployment’s Correlation Analysis Based On Rough Set Theory And Simulation Experiment

The critical step when constructing a house of quality in quality function deployment(QFD) is conducting correlation analysis between customer requirements and engineering characteristics. Therefore, scholars have tried many quantitative methods to improve the precision of this step. And currently two papers, published in Information Science journal, demonstrated that the rough set method could achieve higher precision than other methods. However, when taking linguistic theory and extent of knowledge completion into consideration, there are some potential problems that can erode the precision of rough set method applied in QFD correlation analysis.To solve such potential problems and to further improve the precision of rough set method applied in QFD correlation analysis, this dissertation systematically analyzed the. root causes of these problems and:1) proposed a new concept called linguistic variable averaging and constructed, a model to improve the precision based on this concept;2) proposed a method to assess the extent of knowledge completion based on randomization and non-parameter test.sSince to directly prove the new model’s precision improvement and the assessment method’s validity through mathematical derivation or case study is either difficult or lack of generalization validity, this dissertation adopts a research method, the simulation research method, that goes beyond the traditional pattern of QFD research. By adopting simulation research method this dissertation successfully proves the new model’s precision improvement and the assessment method’s validity.Finally, this dissertation integrates the new model and the assessment method into the process of QFD implementation and successfully applies it into a real case. This offers another effective tool and method for new product design and innovation under increasingly uncertain environment for our companies.