Prediction of Fabric Drape Based on BP Neural Network Paper
- Resource Type
- Conference
- Authors
- Xia, Shuhui; Fan, Xiujuan
- Source
- 2020 IEEE 7th International Conference on Industrial Engineering and Applications (ICIEA) Industrial Engineering and Applications (ICIEA), 2020 IEEE 7th International Conference on. :206-210 Apr, 2020
- Subject
- Engineering Profession
General Topics for Engineers
Fabrics
Biological neural networks
Neurons
Training
Transfer functions
Fitting
lightweight fabrics
drape coefficient prediction
BP neural network
- Language
This paper studies the design of fabric drape prediction model based on BP neural network, and predicts the static and dynamic drape coefficients according to the number of warp and weft yarns, density of warp and weft yarns, weight per square meter and thickness of fabric. The effect of the number of intermediate layer neurons on the accuracy of neural network was studied. The influence of activation function on neural network prediction and initial value on local optimization. The simulation results show that the prediction results of fabric drape by BP neural network are good and meet the expected requirements.