城市系统工程

为探究厦门某机场工程中软土地基大面积堆载造成的地表沉降问题，文章基于厦门某机场工程自动化监测项目，辅以人工监测进行对比，分析偏差的成因，同时利用神经网络 (back propagation, BP) 和长短期记忆网络 (long short-term memory, LSTM) 模型分别对典型区域表层沉降监测点的累计沉降量进行预测分析及精度对比，发现 LSTM 网络预测模型精度更高，整体预测效果优于 BP 神经网络模型，预测效果也更符合实际情况，能为计算工后沉降、评判处理效果、核实工程量等提供一定的参考依据。

In order to explore the surface settlement caused by large-scale stacking of soft soil foundation in an airport project in Xiamen, based on an airport engineering automatic monitoring project in Xiamen, supplemented by manual monitoring for comparison, this paper analyzes the causes of deviation. At the same time, the Back Propagation (BP) and Long Short-Term Memory (LSTM) models are used to predict and analyze the cumulative settlement of surface settlement monitoring points in typical areas and compare the accuracy. It is found that the LSTM network prediction model has higher accuracy, the overall prediction eff ect is better than that of BP neural network model, and the prediction eff ect is more in line with the actual situation, which can provide some reference for calculating post-construction settlement, judging treatment eff ect and verifying engineering quantity.