PDF全文阅读

刘海龙

 

（南昌市建设工程综合监督事务中心 ，南昌 330038)

摘   要 ：为准确掌握深基坑水平位移变化规律，基于某工程实测现场监测数据，先利用统计方法开展其变形特征分析，再进一步利用动态权重粒子群算法、 自适应模糊神经网络和极限学习机构建变形预测模型，对基 坑水平位移预测进行研究。结果表明：基坑水平位移变化范围为 23 . 84 ~ 31 . 96  mm ,  平均值为 29. 03  mm ,  且最 大水平位移主要位于 0. 74H 处 ；基坑预测结果的相对误差范围为 2. 02%~ 2. 17% ,  具有较高预测精度，预测模型 合理，预测结果显示基坑水平位移仍会进一步增加，但增加速率趋于减小，变形总体趋于稳定。通过基坑变形 预测，掌握了基坑水平位移变化规律，为其安全施工提供了技术支持。

关键词：深基坑 ；水平位移 ；数据分析 ；变形预测 ；极限学习机

中图分类号：TU433            

文献标志码：A           

文章编号： 1005- 8249   (2025)  01- 0065- 05 

DOI:10. 19860/j.cnki.issn1005 - 8249.2025 .01 .013

 

LIU Hailong

(  Nanchang  Construction Engineering Comprehensive Supervision Center ,  Nanchang 330038 ,  China)

Abstract: To accurately grasp the variation law of horizontal displacement in deep foundation pits, based on on-site monitoring data, statistical methods are first used to analyze their deformation characteristics. Then, dynamic weight particle swarm optimization algorithm, adaptive fuzzy neural network, and limit learning mechanism are further used to build deformation prediction models to achieve research on horizontal displacement prediction of foundation pits. The case analysis shows that the horizontal displacement range of the foundation pit is 23.84~31.96mm, with an average value of 29.03mm, and the maximum horizontal displacement is mainly located at 0.74H. Meanwhile, through deformation prediction, the relative error range of the predicted results is 2.02%~2.17%, which has high prediction accuracy and verifies the rationality of the prediction model. The subsequent prediction results have high credibility, and the prediction results show that the horizontal displacement of the foundation pit will continue to increase, but the rate of increase tends to decrease, and the deformation overall tends to stabilize. Through research, the horizontal displacement variation law of the foundation pit has been effectively mastered, providing technical support for its safe construction.

Key words: deep foundation pit; horizontal displacement; data analysis; deformation prediction; extreme learning machine

作者简介：刘海龙  (1980—) ,   男，硕士研究生，工程师， 研究方向： 工程质量安全监督管理。

收稿日期：2023 - 02 - 24