Study on settlement prediction model of deep foundation pit in sand and pebble strata based on grey theory and BP neural network

In the process of urban underground space development, it is necessary to predict the ground settlement around the deep foundation pit to protect the surrounding environment. At present, there are few studies on the subsidence of sand and pebble strata. Furthermore, in some engineering sites with multifactor coupling effects, the theoretical calculation is large and the accuracy is low. Consequently, this study introduces the measured data analysis method and focuses on the comparative analysis of settlement prediction models based on grey theory and the back propagation (BP) neural network theory. The surface subsidence data around the deep foundation pit at the metro station in Chengdu, China (which is mainly composed of sand and pebble strata) are used to formulate a prediction and propose an optimized grey prediction model of land subsidence and a one-dimensional double-hidden-layer BP neural network subsidence prediction model. The results show that the prediction results of the two models are accurate and the final prediction results are of certain engineering reference value.