A Hybrid System Based on Neural Network and Immune Co-Evolutionary Algorithm for Garment Pattern Design Optimization

Abstract

The purpose of this study is to develop a system to utilize the successful experiences and help the beginners of garment pattern design (GPD) by optimization methods. A hybrid algorithm (NN-ICEA) based on Neural Network (NN) and immune co-evolutionary algorithm (ICEA) to predict the fit of the garments and search optimal sizes. ICEA takes NN as fitness function and procedures including clonal proliferation, hyper-mutation and co-evolution search the optimal size values. Then, a series of experiments with a dataset of 450 pieces of garments are conducted to demonstrate the prediction and optimization capabilities of NN-ICEA. In the comparative studies, NN-ICEA is compared with NN-GA to show the value of immune inspired operators. Four types of GPD methods are summarized and compared. Moreover, the hybrid system for general features of garment is discussed. The fit prediction based on NN can achieve the high accuracy with the error rate less than 0.2. The size optimization based on ICEA works well when number of the missing sizes is less than 1/3 of the total size number. The research is a feasible and effective attempt aiming at a valuable problem and provides key algorithms for fit prediction and size optimization. The algorithms can be incorporated into garment CAD system.