[1]姜永杰,张书豪,吴光,等.基于灰色系统及BP神经网络算法的边坡变形预测精确度影响分析[J].高速铁路技术,2020,11(05):13-19.[doi:10.12098/j.issn.1674-8247.2020.05.003]
　JIANG Yongjie,ZHANG Shuhao,WU Guang,et al.Influence Analysis of Slope Deformation Prediction Accuracy Based on Grey System and BP Neural Network Algorithm[J].HIGH SPEED RAILWAY TECHNOLOGY,2020,11(05):13-19.[doi:10.12098/j.issn.1674-8247.2020.05.003]
点击复制

基于灰色系统及BP神经网络算法的边坡变形预测精确度影响分析()
分享到：

《高速铁路技术》[ISSN:1674-8247/CN:51-1730/U]

卷:
11卷
期数:
2020年05期
页码:
13-19
栏目:
出版日期:
2020-10-28

文章信息/Info

Title:
Influence Analysis of Slope Deformation Prediction Accuracy Based on Grey System and BP Neural Network Algorithm
文章编号:
1674—8247(2020)05—0013-07
作者:
姜永杰;  张书豪;  吴光;  张广泽;  柴春阳
1. 西南交通大学, 成都 611756;
2. 中铁二院工程集团有限责任公司, 成都 610031
Author(s):
JIANG Yongjie;  ZHANG Shuhao;  WU Guang;  ZHANG Guangze;  CHAI Chunyang
1. Southwest Jiaotong University, Chengdu 611756, China;
2. China Railway Eryuan Engineering Group Co., Ltd., Chengdu 610031, China
关键词:
边坡变形|预测精确度|算法耦合|数据预处理|位移时序图
Keywords:
slope deformation|accuracy of prediction|algorithm coupling|data preprocessing|displacement sequence chart
分类号:
TU43
DOI:
10.12098/j.issn.1674-8247.2020.05.003
文献标志码:
A
摘要:
在边坡变形数字化分析中，数据算法的选取已成为影响预测结果的重要因素之一。本文旨在排除工程地质条件等其他现场因素的影响，研究分析工具中常见的窗口平滑多项式、小波分析，对灰色系统GM（1，1）及BP神经网络在预测边坡变形时的精确度影响。结果表明：（1）当边坡变形稳定，边坡位移时序图呈直线抬升趋势时，此时基于窗口平移多项式的BP神经网络能够获得较为精确的预测值；（2）当边坡变化异常，位移时序图呈阶跃式曲线时，此时基于小波分析的BP神经网络能够获得较为精确的预测值。
Abstract:
In the digital analysis of slope deformation, the selection of data algorithm has become one of the important factors affecting prediction results. The purpose of the paper is to research the influence of window smoothing polynomials and wavelet analysis, which are common in analysis tools, on the accuracy of grey system GM(1,1) and BP neural network in prediction of slope deformation by eliminating the influence of engineering geological conditions and other field factors. The results show that:(1) When the slope deformation is stable and the slope displacement sequence chart shows a straight upward trend, the BP neural network based on window smoothing polynomials can obtain more accurate predicted values; (2) When the slope changes abnormally and the displacement sequence chart shows a step curve, the BP neural network based on wavelet analysis can obtain more accurate predicted values.

备注/Memo

备注/Memo:
作者简介:姜永杰(1994-),男,硕士。基金项目:四川省科技计划资助（2019YFG0460）引文格式:姜永杰, 张书豪, 吴光, 等. 基于灰色系统及BP神经网络算法的边坡变形预测精确度影响分析[J]. 高速铁路技术,2020,11(5):13-19.JIANG Yongjie, ZHANG Shuhao, WU Guang, et al. Influence Analysis of Slope Deformation Prediction Accuracy Based on Grey System and BP Neural Network Algorithm[J]. High Speed Railway Technology, 2020, 11(5):13-19.
更新日期/Last Update: 2020-10-28