Perceptual evaluation for Zhangpu paper-cut patterns by using improved GWO-BP neural network

Abstract

Abstract In order to understand consumers’ perceptual cognition of Zhangpu paper-cut patterns and grasp the innovative application direction. The four design elements of paper-cut patterns were extracted by morphological analysis, and representative perceptual vocabulary were selected using Kansei engineering theory and factor analysis, then the design elements and perceptual evaluation scores of representative words are used as the input and output data of the GWO-BP neural network, respectively, to establish an intelligent model that can predict consumers’ perceptual cognition of paper-cut patterns. To verify the superiority of the model, the predicted result of BP and FA-BP are compared with GWO-BP neural network. The results show that although the convergence speed of the GWO-BP model is slightly lower than that of the FA-BP model, its prediction accuracy is significantly better than other algorithms. Designers can use the model to quickly redesign the paper-cut pattern to better meet the aesthetic needs of modern consumers.