市政技术2025，Vol.43Issue(7)：230-237,300,9.DOI:10.19922/j.1009-7767.2025.07.230
基于SSA-LSTM的老旧房屋变形监测数据预测
Prediction of Deformation Monitoring Data of Old Buildings Based on SSA-LSTM
孙振林 1周鹏 2杨越东 1董奇 3王子鑫 3杨晓辉3
作者信息
- 1. 北京市市政工程研究院,北京 100037
- 2. 中国计量科学研究院,北京 100029
- 3. 北京市建设工程质量第三检测所有限责任公司,北京 100037
- 折叠
摘要
Abstract
On the basis of the deformation monitoring project of old buildings in Beijing,a deformation prediction model of old buildings based on the sparrow search algorithm and the long short-term memory network(SSA-LSTM)was proposed.This model,combined with the advantages of SSA and LSTM,can effectively learn the pattern of long-term historical deformation data and predict deformation parameters such as structural displacement,founda-tion settlement,wall crack width and strain more accurately.In order to evaluate the prediction accuracy of the model,the BP neural network model was established to compare with traditional LSTM model.The results showed that the SSA-LSTM model has a higher prediction accuracy.The MA PE and RMSE significantly decreased compared with the traditional models and the R2 was greater than 0.99,and the SSA-LSTM model obtained a better fitting effect.Therefore,the intelligent early warning system for deformation monitoring of old buildings based on the SSA-LSTM model can predict various deformation parameters more accurately to achieve early warning.This research provides effective technical support for the safety monitoring and early warning of old buildings with great practical value.关键词
麻雀搜索算法/长短期记忆网络/老旧房屋监测/物联网/变形预测模型Key words
sparrow search algorithm/long short-term memory network/old building monitoring/internet of things/deformation prediction model分类
建筑与水利引用本文复制引用
孙振林,周鹏,杨越东,董奇,王子鑫,杨晓辉..基于SSA-LSTM的老旧房屋变形监测数据预测[J].市政技术,2025,43(7):230-237,300,9.
