The degree of satisfying the customer requirements was determined by the diversity of product function and form, in order to improve the satisfaction, the mapping relation between customer requirements and technical characteristics should be analyzed to construct configuration optimization model. This paper integrated triangular fuzzy number, BP network and GA to propose a method of constructing optimization model of product technical characteristics configuration corresponding to customer requirements, the model was established based on customer transaction data history. This method acquired the information of customer requirements and preference and that of purchased products’ function and form characteristic by analyzing and extracting customer transaction information, then applied BP network to express the mapping relations between customer requirements and product technical characteristics. The optimization model took the cost and time as constraint with the customer requirements as target, finally used genetic algorithm with MATLAB to solve. The electric bicycle was taken to illustrate the build process.