基于深度学习的盾构竖向姿态组合预测

李增良

doi:10.3973/j.issn.2096-4498.2021.05.008

隧道建设 >

2021 , Vol. 41 >Issue 5: 758 - 763

DOI: https://doi.org/10.3973/j.issn.2096-4498.2021.05.008

研究与探索

基于深度学习的盾构竖向姿态组合预测

李增良

展开

（中铁二十局集团有限公司，陕西西安〓710016）

李增良(1976—),男,湖南平江人,2000 年毕业于中南大学,机械制造及自动化专业,本科,高级工程师,主要从事物资设备及盾构施工技术与管理工作。E-mail: 852499699@qq.com。

网络出版日期: 2021-05-29

基金资助

中国博士后科学基金资助项目(2020M673525)

收起

Combined Prediction Model for Shield Vertical Posture Based on Deep Learning

LI Zengliang

Expand

(China Railway 20th Bureau Group Co., Ltd., Xi′an 710016, Shaanxi, China)

Online published: 2021-05-29

Fold

摘要

为解决盾构竖向姿态的精确预测问题，提出一种基于长短期记忆（long short term memory，LSTM）神经网络-支持向量回归（support vector regression，SVR）的深度学习组合预测模型。在对采集到的竖向姿态数据进行相应的数据预处理的基础上，分别构建LSTM、SVR竖向姿态预测模型，并基于最优组合赋权的方式对二者的预测结果进行赋权,以得到LSTM-SVR盾构竖向姿态组合预测模型。为验证所构建的LSTM-SVR组合深度学习预测模型的可靠性，依托昆明地铁项目，将预测结果与LSTM、SVR、BP（back propagation）模型的预测结果进行对比。结果表明：所构建的LSTM-SVR组合深度学习预测模型具有较高的预测精度。

关键词： 地铁隧道; 组合预测模型; 深度学习; 盾构竖向姿态; 长短期记忆神经网络; 支持向量回归

本文引用格式

李增良 . 基于深度学习的盾构竖向姿态组合预测[J]. 隧道建设, 2021 , 41(5) : 758 -763 . DOI: 10.3973/j.issn.2096-4498.2021.05.008

Abstract

To accurately predict the vertical posture of shields, a deeplearning combination prediction model based on long shortterm memory (LSTM) neural network and support vector regression (SVR) is proposed. The LSTM and SVR verticalposture prediction models are developed by performing the corresponding data preprocessing operations on the collected verticalposture data. Further, the prediction results of the two models are weighted by the optimal combination of weights to obtain a LSTMSVR combined prediction model. Finally, to verify the reliability of the developed LSTMSVR combined deeplearning prediction model, the prediction results are compared with those of the LSTM, SVR, and BP models based on the Kunming metro project. The results show that the LSTMSVR combined deeplearning prediction model has high prediction accuracy.

Key words： metro tunnel; combined prediction model; deep learning; shield vertical posture; long shortterm memory (LSTM) neural network; support vector regression (SVR)

Options

文章导航

〈

〉

模态框（Modal）标题

摘要

本文引用格式

Abstract