Digital Model of Deflection of a Cable-Stayed Bridge Driven by Deep Learning and Big Data Optimized via PCA-LGBM

Based on big data, we can build a regression model between temperature field and temperature-induced deflection to provide control group representing the service performance of bridges, which has positive effect full life cycle maintenance bridges. However, spatial information cable-stayed bridge is difficult describe. To establish with high precision, improved PCA-LGBM (principal component analysis light gradient boosting machine) algorithm proposed extract main features that reflect as accurately efficiently possible. Then, in this article, searched for suitable digital tool modeling regressive relationship variables bridge. The multiple linear relatively low precision. precision backpropagation neural network (BPNN) been improved, but it still unsatisfactory. nested long short-term memory (NLSTM) improves nonlinear expression ability more precise than BPNN models classical LSTM. architecture NLSTM optimized avoid waste computational costs. four extracted via PCA-LGBM, double hidden layers 256 units each layer much higher other models. For bridge, mean absolute error only 4.76 mm, square 18.57 mm2. value thus provides potential early detection anomalies. This article reference processes data extraction