摘要
提出一种基于长短期神经网络的深度学习预测模型,依托现场数据对土仓压力进行预测。结果表明,在5个可控因素的基础上,增加与土仓压力具有相关关系的不可控因素作为输入,评价指标平均绝对误差、均方误差分别降低了0.901%、0.021%,校正后的决定系数提高了16%,为土仓压力的精准预测和设定提供了借鉴。
A deep learning prediction model is proposed based on long-term and short-term neural networks to predict chamber earth pressure based on field data.Research results show that on the basis of 5 controllable factors,adding uncontrollable factors related to chamber earth pressure as input,the average absolute error and mean square error of evaluation index have been reduced by 0.901%and 0.021%respectively.The corrected coefficient of determination is increased by 16%,which provides a reference for the accurate prediction and setting of the chamber earth pressure.
作者
成晓元
凌静秀
黄继辉
吴勉
CHENG Xiaoyuan;LING Jingxiu;HUANG Jihui;WU Mian(School of Mechanical and Automotive Engineering, Fujian University of Technology, Fuzhou 350118, China;Fujian Institute of Technology Innovation for CNC Equipment Industry, Fujian University of Technology, Fuzhou 350118, China;CSCEC Strait Construction and Development Co.,Ltd., Fuzhou 350000, China)
出处
《福建工程学院学报》
CAS
2022年第1期13-18,共6页
Journal of Fujian University of Technology
基金
福建省自然科学基金资助项目(2020J01871)
福建工程学院校科研启动基金(GY-Z160048)。
