针对QFD系统内在的模糊性和不精确性，基于来自于在线评论的大数据，本项目主要围绕动态QFD规划进行模糊建模与优化研究。首先，基于在线评论，从顾客和顾客需求两个维度，应用模糊双聚类分析进行模糊建模，对关键顾客需求进行挖掘与分析；第二，考虑到不同顾客需求之间对产品的工程特性的相关作用，由此导致“演化均衡漂移”现象，进行演化博弈建模，设计算法，以确定顾客需求最终重要度；第三，研究基于改进遗传算法的广义模糊时间序列模型，并基于所建模型对动态顾客需求进行分析与预测；第四，为实现顾客满意度最大化，以各工程特性为博弈方，建立纳什议价函数，进而构建模糊合作博弈模型，以确定工程特性目标值，并以某产品设计为例，对经典合作博弈模型和模糊合作博弈模型进行对比分析。本项目将帮助设计者基于在线评论的大数据，不断逼近真实的顾客需求，对动态顾客需求进行预测，实现产品的各工程特性之间的最优补偿，提高新产品的设计质量。

Owning to the inherent fuzziness and imprecision in QFD modeling, fuzzy modeling and optimization for dynamic QFD planning will be investigated in this project based on the big data from online reviews. Firstly, considering the two dimensions from customer and customer requirements (CRs), fuzzy bi-clustering analysis is mainly employed to mine and analyze the key CRs based on the online review. Secondly, different CRs will influence the related engineering characteristics (ECs) of the new product, which will lead to the phenomenon of “evolutionary equilibrium drift”, so that this project will develop the evolutionary game model and design algorithm to determine the final importance weights of CRs. Thirdly, based on the modified genetic algorithm, a generalized fuzzy time series model is developed to forecast and analyze the dynamic CRs. Finally, to achieve the goal of maximizing the overall customer satisfaction, the players are viewed as ECs to formulate the bargaining function and to develop the fuzzy cooperative games model to set target levels of ECs in QFD. Moreover, the classical cooperative games model and fuzzy cooperative games model will be compared by a numerical example of the product design. The purpose of this project is to improve the quality of new product development by helping designers approach the real CRs based on the big data from online reviews, forecasting dynamic CRs, and facilitating the good performance in one EC to compensate for poor performance in other ECs optimally.