Digital-Twin-Enhanced Quality Prediction for the Composite Materials
Yu-Cheng Wang, Fei Tao, Ying Zuo, Meng Zhang, Qinglin Qi
Abstract
Composite materials are widely used in many fields due to their excellent properties. Quality defects in composite materials can lead to lower quality components, creating potential risk of accidents. Experimental and simulation methods are commonly used to predict the quality of composite materials. However, it is difficult to predict the quality of composite materials accurately due to the uncertain curing environment and incomplete feature space. To address this problem, a digital twin (DT) visual model of a composite material is first constructed. Then, a static autoclave DT virtual model is coupled with a variable composite material DT virtual model to construct a model of the curing process. Features are added to the proposed model by generating simulated data to enhance the quality prediction. An extreme learning machine (ELM) for quality prediction is trained with the generated data. Finally, the effectiveness of the proposed method is verified through result analysis.