摘要
高速铁路隧道工程监测点沉降曲线具有小沉降、大波动的特点,沉降变形数据中存在大量随机噪声,将影响沉降变形分析的准确性。将Kalman滤波应用于高铁隧道沉降变形数据预处理,对沉降变形数据进行去噪,再利用小波神经网络对去噪后的沉降变形数据进行预测分析,从而提高单一小波神经网络的预测精度。通过工程实例分析表明,结合Kalman滤波的小波神经网络预测精度优于单一小波神经网络,具有更好的应用价值。
The settlement curve of high-speed railway tunnel engineering monitoring points has the characteristics of small settlement and large fluctuation.There is a lot of random noise in the settlement deformation data,which affects the accuracy of settlement deformation analysis.In this paper,we introduced Kalman filter to preprocess the settlement deformation data of high-speed railway tunnel and denoise the settlement deformation data.Then,we used wavelet neural network to predict and analyze the denoised settlement deformation data,aiming to improve the prediction accuracy of single wavelet neural network.The analysis of engineering examples shows that the prediction accuracy of wavelet neural network combined with Kalman filter is improved compared with that of single wavelet neural network,which has better application value.
作者
陈冠宇
胡小伍
洪雪倩
党沙沙
周吕
CHEN Guanyu;HU Xiaowu;HONG Xueqian;DANG Shasha;ZHOU Lyu(Zhejiang Province Land Survey and Planning Co.,Ltd.,Hangzhou Branch,Hangzhou 310030,China;Zhejiang Academy of Surveying and Mapping,Hangzhou 310030,China;College of Geomatics and Geoinformation,Guilin University of Technology,Guilin 541004,China)
出处
《地理空间信息》
2023年第1期101-103,112,共4页
Geospatial Information
基金
广西科技计划资助项目(桂科AD19110107)。
关键词
高铁隧道
KALMAN滤波
小波神经网络
变形分析
high-speed railway tunnel
Kalman filter
wavelet neural network
deformation analysis
作者简介
第一作者:陈冠宇(1987—),硕士研究生,工程师,主要从事精密工程测量、地理信息、自然资源等方面的工作;通信作者:胡小伍。
