An analysis method for change propagation based on product feature network

Abstract

At present, how to evaluate the impact of product design changes in product development process has been a problem to be solved. The research on product design changes has gradually shifted from the component level to the product feature level. Therefore, the components are divided into feature level, and a product feature network is built to analyse the propagation of changes. Because it is not unique to divide features of the component, three-dimensional (3D) entities are used to transit components into 2D features. The strong and weak ties are defined according to the relationship between the divided features, the product feature network is constructed by the divided features and defined ties. Three discrete states of feature nodes are defined based on the change propagation model and the meaning of change propagation index (CPI) in the feature network is proposed. By analysing two types of product feature change propagation and calculating its change probability, the impact of change propagation in the network can be evaluated. Finally, the feasibility of the proposed model is verified by constructing a featured network of a pumping unit and analysing its change propagation.