基于BAS-灰色神经网络的矿区铁路沉降变形预测  

Prediction of Railway Subsidence and Deformation in Mining Areas Based on BAS-grey Neural Network

作　　者：俞栋耀 蘧振超 方荣兴 YU Dongyao;QU Zhenchao;FANG Rongxing(Zhejiang Provincial Institute of Surveying and Mapping Science and Technology,Hangzhou 311100,China;Zhejiang Xinyu Space Time Technology Co.,Ltd.,Hangzhou 311100,China)

机构地区：[1]浙江省测绘科学技术研究院,浙江杭州311100 [2]浙江信宇时空科技有限公司,浙江杭州311100

出　　处：《测绘与空间地理信息》2025年第2期221-224,共4页Geomatics & Spatial Information Technology

摘　　要：为了改进矿区铁路沉降动态预测方法,本文在灰色神经网络的基础上,针对该模型的不足,引入天牛须搜索(Beetle Antennae Search,BAS)算法对模型的权值与阈值进行优化,建立BAS-灰色神经网络模型。为了检验本文提出的BAS-神经网络模型在实际工程中的应用效果,使用某矿区铁路沉降监测数据进行实验。实验结果表明,BAS-灰色神经网络模型预测结果的均方根误差、平均绝对百分比误差均优于灰色模型、灰色神经网络,表现出了更好的预测效果与更高的预测精度,能够较好地反映出矿区地表沉降变形的发展趋势。In order to improve the dynamic prediction method of railway subsidence in mining areas,based on the grey neural network,and aiming at the shortcomings of the model,this paper introduces the beetle antenna search(BAS)algorithm to optimize the weight and threshold of the model,and establishes the model of BAS-grey neural network.In order to test the application effect of the BAS-grey neural network model proposed in this paper in practical projects,the railway subsidence monitoring data of a mining area is used for experiments.The experimental results show that the root mean square error and the average absolute percentage error of the prediction results of the BAS-gray neural network model are superior to the gray model and the gray neural network,which shows better prediction effect and higher prediction accuracy and can better reflect the development trend of surface subsidence and deformation in the mining area.

关 键 词：变形预测 天牛须搜索算法 灰色神经网络 矿山开采沉陷 

分 类 号：P25[天文地球—测绘科学与技术] TB22[天文地球—大地测量学与测量工程]