To ensure the construction safety of large foundation pit engineering (FPE) and optimize the data collection method, this paper applies sensors and automatic collection items to automatic data collection based on traditional manual measurement. Data processing and analysis are carried out through the large FPE detection system construction. In the early warning module, the improved Grey Model (GM) (1,1), namely the improved method G, is used to predict the settlement of buildings around the monitoring project, and its reliability is analyzed. The improved method G has a good test result. Compared with method G and the Autoregressive Integrated Moving Average model (ARIMA), the improved Method G has a better prediction effect and smaller prediction error. When compared with method G, the mean absolute error (MAE) of the method used in the paper is smaller, which is −0.01. The root mean square value of absolute error is smaller, which is 0.0.54 less than method G. Compared with the ARIMA model, the average RE of the method used in this paper is 3.97%, which is 0.05% less than that. This method can be used to collect and predict data on large foundation pit (FP) construction sites, which is conducive to ensuring construction safety.
Read full abstract