[1]徐根祺,李丽敏*,温宗周,等.基于宽度学习模型的泥石流灾害预报[J].山地学报,2019,(6):868-878.[doi:10.16089/j.cnki.1008-2786.000477]
　XU Genqi,LI Limin*,WEN Zongzhou,et al.Debris Flow Disaster Prediction Based on Broad Learning Model[J].Mountain Research,2019,(6):868-878.[doi:10.16089/j.cnki.1008-2786.000477]
点击复制

基于宽度学习模型的泥石流灾害预报()
分享到：

《山地学报》[ISSN:1008-2186/CN:51-1516]

卷:
期数:
2019年第6期
页码:
868-878
栏目:
山地灾害
出版日期:
2019-11-30

文章信息/Info

Title:
Debris Flow Disaster Prediction Based on Broad Learning Model
文章编号:
1008-2786-(2019)6-868-11
作者:
徐根祺; 李丽敏* ; 温宗周; 刘德阳; 程少康
西安工程大学 电子信息学院,西安 710048
Author(s):
XU Genqi;  LI Limin*;  WEN Zongzhou;  LIU Deyang;  CHENG Shaokang
College of Electronic Information, Xi'an Polytechnic University, Xi'an 710048, China
关键词:
宽度学习算法;  泥石流;  预报模型;  奇异值分解
Keywords:
broad learning algorithm;  debris flow;  forecasting model;  singular value decomposition
分类号:
P642~A
DOI:
10.16089/j.cnki.1008-2786.000477
摘要:
在泥石流灾害预报模型研究中,科学确定泥石流灾害影响因子及保证模型较高的预报准确率和快速的训练速度是关键问题,也是泥石流灾害预报预警和防灾减灾的重要基础。本研究针对目前泥石流预报模型输入数据维度较大和训练时间较长的问题,采用快速多个主成分并行提取算法(Fast multiple principle components extraction algorithm,FMPCE),选取出6个泥石流灾害影响因子,包括降雨量、山坡坡度、沟床比降、相对高差、土壤含水率和孔隙水压力。基于宽度学习(Broad learning,BL)算法,以泥石流影响因子为输入,泥石流发生概率为输出,构建了泥石流预报模型,并用矩阵随机近似奇异值分解(矩阵随机近似SVD)对模型进行了优化,将优化后宽度学习模型的预报结果与梯度下降法优化的BP神经网络预报模型(GD-BP)、基于支持向量机的预报模型(SVM)、宽度学习预报模型(BL)的结果进行对比,同时,通过输入数据集的扩展,从训练时间上对不同模型进行比较。结果表明,优化宽度学习泥石流灾害预报模型的预报准确率为93.52%,较GD-BP模型、SVM模型和BL模型的预报准确率分别高出1.60%、1.15%和0.03%; 优化宽度学习泥石流灾害预报模型的训练时间为0.9039s,较GD-BP模型、SVM模型和BL模型的训练时间分别节省了25.3867 s、17.2620 s和0.8974 s。本研究说明宽度学习算法可以用于对泥石流灾害的发生概率进行预报,同时也可为泥石流预报的实际应用提供新的思路。
Abstract:
In the study of debris flow disaster forecasting model, it is a key problem to scientifically determine the influencing factors of debris flow disasters and ensure the higher forecast accuracy and fast training speed of the model, which is also an important foundation for debris flow disaster forecasting and early warning and disaster prevention and mitigation as well.Aiming at the problems of large dimension of input data and long training time of current debris flow prediction models, this paper used a fast-multiple principle components extraction algorithm(FMPCE)to select 6 influence factors of debris flow disaster, including rainfall volume, hillside slope, ditch bed ratio, relative height difference, soil moisture content and pore water pressure. Based on the broad learning(BL)algorithm, the influence factors of debris flow were taken as inputs and debris flow occurrence probability as output to construct a debris flow forecasting model, and the model was optimized by using matrix random approximate singular value decomposition(matrix random approximation SVD). The prediction results of the optimized broad learning model were compared with the results of the BP neural network prediction model(GD-BP)optimized by the gradient descent method, the support vector machine-based prediction model(SVM), and the broad learning prediction model(BL). At the same time, through the expansion of the input data set, different models were compared from the training time. The results showed that the prediction accuracy of the optimized broad learning debris flow disaster prediction model was 93.52%, which was 1.60%, 1.15%, and 0.03% higher than the prediction accuracy of the GD-BP model, the SVM model, and the BL model; The training time of the optimized broad learning debris flow disaster prediction model was 0.9039 s, which saved 25.3867 s, 17.2620 s, and 0.8974 s compared with the training time of the GD-BP model, SVM model, and BL model, respectively, which showed that the broad learning algorithm can be used to predict the occurrence probability of debris flow disasters, and provided a new idea for the practical application of debris flow prediction as well.

参考文献/References:

[1] 刘永垚,第宝锋,詹宇,等.基于随机森林模型的泥石流易发性评价-以汶川地震重灾区为例[J].山地学报,2018,36(5):765-773. [LIU Yongyao, DI Baofeng, ZHAN Yu, et al. Debris flows susceptibility assessment in Wenchuan earthquake areas based on random forest algorithm model[J]. Mountain Research,2018,36(5):765-773]
[2] 李璐,温宗周,张阳阳,等.基于RBF神经网络滑坡灾害发生概率预报方法[J].西安工程大学学报,2017,31(4):521-526. [LI Lu, WEN Zongzhou, ZHANG Yangyang, et al. Probability prediction method of landslide disaster based on RBF neural network [J]. Journal of Xi'an Polytechnic University,2017,31(4):521-526]
[3] 屈永平,肖进,潘义为.藏东南地区冰川泥石流形成条件初步分析-以天摩沟冰川泥石流为例[J].水利水电技术,2018,49(12):177-184. [QU Yongping, XIAO Jin, PAN Yiwei. Preliminary analysis on formation conditions of glacier debris flow in Southeast Tibet-a case of glacial debris flow in Tianmo Gully [J]. Water Resources and Hydropower Engineering,2018,49(12):177-184]
[4] 冯杭建,周爱国,唐小明,等.浙江省泥石流灾害发育分布规律及区域预报[J].地球科学,2016,41(12):2088-2099[FENG Hangjian, ZHOU Aiguo, TANG Xiaoming, et al. Development and distribution characteristics of debris flow in Zhejiang Province and Its Regional Forecast[J]. Earth Science,2016,41(12):2088-2099]
[5] 王高峰,杨强,田运涛,等.泥石流易发性评价模型的构建-以白龙江流域石门乡羊汤河段为例[J].干旱区研究,2019,36(3):761-770. [WANG Gaofeng, YANG Qing, TIAN Yuntao, et al. Establishment of assessment model for debris flow susceptibility: a case study along the Yangtang River Reach in Shimen township in the Bailong River Basin [J]. Arid Zone Research,2019,36(3):761-770]
[6] 李丽敏,程少康,温宗周,等.基于改进KPCA与混合核函数LSSVR的泥石流预测[J].信息与控制,2019,48(5):536-544[LI Limin, CHENG Shaokang, WEN Zongzhou, et al. A debris flow prediction model based on the improved KPCA and Mixed Kernel function LSSVR [J]. Information and Control,2019,48(5):536-544]
[7] 李秀珍,孔纪名,李朝凤.多分类支持向量机在泥石流危险性区划中的应用[J].水土保持通报,2010,30(5):128-133+157. [LI Xiuzhen, KONG Jiming, LI Chaofeng. Application of multi-classification support vector machine in regionalization of debris flow hazards [J]. Bulletin of Soil and Water Conservation,2010,30(5):128-133+157]
[8] 张晨,王清,张文,等.基于神经网络对泥石流危险范围的研究[J].哈尔滨工业大学学报,2010,42(10):1642-1645. [ZHANG Chen, WANG Qing, ZHANG Wen, et al. Prediction on hazardous areas of debris flow based on neural network[J]. Journal of Harbin Institute of Technology,2010,42(10):1642-1645]
[9] 曹禄来,徐林荣,陈舒阳,等.基于模糊神经网络的泥石流危险性评价[J].水文地质工程地质,2014,41(2):143-147. [CAO Lulai, XU Linrong, CHEN Shuyang, et al. Debris flow risk assessment based on fuzzy neural network[J]. Hydrogeology and Engineering Geology, 2014,41(2):143-147]
[10] 谢涛,尹前锋,高贺,等.基于激发条件和堆积体稳定性的冰川降雨型泥石流预警模型研究[J].冰川冻土,2019,41(4):884-891. [XIE Tao, YIN Qianfeng, GAO He, et al. Study on early warning model of glacial-rainfall debris flow based on excitation condition and stability of accumulation body [J]. Journal of Glaciology and Geocryology,2019,41(4):884-891]
[11] 郭瑞,彭杨宏,王其杰,等.基于灰色关联法的泥石流灾害易发性影响因素分析-以甘肃舟曲瓜咋沟泥石流为例[J].地质论评,2018,64(6):1481-1488[GUO Rui,PENG Yanghong, WANG Qijie, et al. analisis on hazard evaluation of debris flow based on grey relational theory-a case study on the debris flow in Guaza Gully, Zhouqu, Gansu [J]. Geological Review,2018,64(6):1481-1488]
[12] 刘双,余斌,马二龙,等.山西省平定县寨坪沟泥石流灾害特征及预警[J].泥沙研究,2018,43(6):61-66. [LIU Shuang, YU Bin, MA Erlong, et al. Study on early warning of debris flow disaster in Zhaiping Gully in Shanxi Province [J]. Journal of Sediment Research,2018,43(6):61-66]
[13] 吴赛儿,陈剑,ZHOU Wendy,等.基于逻辑回归模型的泥石流易发性评价与检验:以金沙江上游奔子栏-昌波河段为例[J].现代地质,2018,32(3):611-622. [WU Saier, CHEN Jian, WENDY Zhou, et al. Debris-flow susceptibility assessment and validation based on logistic regression model: an example from the Benzilan-Changbo segment of the upper Jinshajiang River [J].Geoscience,2018,32(3):611-622]
[14] 黄启乐,陈伟,傅旭东.斜坡单元支持下区域泥石流危险性AHP-RBF评价模型[J].浙江大学学报(工学版),2018,52(9):1667-1675.[HUANG Qile, CHEN Wei, FU Xudong. AHP-RBF assessment model of regional debris flow hazard supported by unit slope[J].Journal of Zhejiang University(Engineering Science),2018,52(9):1667-1675]
[15] CHEN C L P, LIU Zhulin. Broad learning system: an effective and efficient incremental learning system without the need for deep architecture. IEEE Transactions on Neural Networks and Learning Systems, 2018, 29(1): 10-24.
[16] 孔令智,高迎彬,李红增,等.一种快速的多个主成分并行提取算法[J].自动化学报,2017,43(5):835-842.[KONG Lingzhi, GAO Yingbin, LI Hongzeng, et al. A fast parallel extraction algorithm for multiple principal components[J].Acta Automatica Sinica,2017,43(5):835-842]
[17] HALKO N, MARTINSSON P G. TROPP J A. Finding structure with randomness: stochastic algorithms for constructing approximate matrix decompositions [J]. Journal SIAM Review, 2011, 53:217-288.
[18] 彭贵芬,刘盈曦.基于模糊信息概率区间数的突发地质灾害降水因子评价研究[J].灾害学,2015,30(1):1-4. [CHEN Guifen, LIU Yingbiao. Evaluation of precipitation factors of sudden geological disasters based on probability interval number of fuzzy information [J]. Journal of Catastrophology,2015, 30(1):1-4]
[19] 赵岩.流域发育程度遥感反演及其在泥石流预报中的应用[D].中国科学院大学,2012:15-30. [ZHAO Yan. Remote sensing inversion of watershed development degree and its application in debris flow forecasting [D]. University of Chinese Academy of Sciences, 2012:15-30]
[20] 方成杰,钱德玲,徐士彬,等.基于云模型的泥石流易发性评价[J].合肥工业大学学报(自然科学版),2017,40(12):1659-1665. [FANG Chengjie, QIAN Deling, XU Shibin, et al. Evaluation of debris flow susceptibility based on cloud model[J]. Journal of Hefei University of Technology(Natural Science), 2017, 40(12):1659-1665]

相似文献/References:

[1]谢湘平,苏鹏程,王小军,等.工程弃渣泥石流易发性评估方法[J].山地学报,2016,(02):216.[doi:10.16089/j.cnki.1008-2786.000121]
　XIE Xiangping,SU Pengcheng,WANG Xiaojun,et al.Occurrence Probability of Channel Waste-slag Debris Flow Assessment[J].Mountain Research,2016,(6):216.[doi:10.16089/j.cnki.1008-2786.000121]
[2]陈 剑,黎 艳,许 冲.金沙江干热河谷区泥石流易发性评价模型及应用[J].山地学报,2016,(04):460.[doi:10.16089/j.cnki.1008-2786.000151]
　CHEN Jian,LI Yan,XU Chong.Susceptibility Assessment Model of Debris Flows in the Dry-hot Valley of the Jinsha River and Its Application[J].Mountain Research,2016,(6):460.[doi:10.16089/j.cnki.1008-2786.000151]
[3]邹 强,唐建喜,李淑松,等.基于水文响应单元的泥石流灾害易发性分区方法[J].山地学报,2017,(04):496.[doi:10.16089/j.cnki.1008-2786.000247]
　ZOU Qiang,*,TANG jianxi,et al.Susceptibility Assessment Method of Debris Flows Based on Hydrological Response Unit[J].Mountain Research,2017,(6):496.[doi:10.16089/j.cnki.1008-2786.000247]
[4]廖丽萍,朱颖彦*,杨志全,等.震区砾石土泥石流起动临界状态与力学性状[J].山地学报,2017,(04):506.[doi:10.16089/j.cnki.1008-2786.000248]
　LIAO Liping,,et al.The Mechanical Property of Gravel Soil in Seismic Area and Its Critical State in Initiating Debris Flow[J].Mountain Research,2017,(6):506.[doi:10.16089/j.cnki.1008-2786.000248]
[5]方迎潮,王道杰*,何松膛,等.云南东川蒋家沟泥石流2003-2014年冲淤演变特征[J].山地学报,2018,(06):907.[doi:10.16089/j.cnki.1008-2786.000386]
　FANG Yingchao,WANG Daojie*,HE Songtang,et al.Characteristics of Debris Flow Erosion and Deposition at Jiangjia Gully, Dongchuan, Yunnan Province, China for 2003-2014[J].Mountain Research,2018,(6):907.[doi:10.16089/j.cnki.1008-2786.000386]
[6]陈宁生*,佘德彬.基于弃渣综合利用的矿山泥石流灾害防治新模式--以冕宁盐井沟泸沽铁矿为例[J].山地学报,2019,(01):78.[doi:10.16089/j.cnki.1008-2786.000401]
　CHEN Ningsheng,SHE Debin.A New Approach to Debris Flow Disaster Control Based on Comprehensive Utilization of Waste Slag—A Case Study of Lugu Iron Mine at the Yanjing Valley of Mianning County, Sichuan, China[J].Mountain Research,2019,(6):78.[doi:10.16089/j.cnki.1008-2786.000401]
[7]王凤娘,贺 拿,陈 容,等.九寨沟县西番沟泥石流调查[J].山地学报,2019,(04):622.[doi:10.16089/j.cnki.1008-2786.000453]
　WANG Fengniang,HE Na,CHEN Rong,et al.Investigation of Debris Flow in Xifangou, Jiuzhaigou County, China[J].Mountain Research,2019,(6):622.[doi:10.16089/j.cnki.1008-2786.000453]
[8]谢湘平,王小军,闫春岭.漂木灾害研究现状及研究展望[J].山地学报,2020,(4):552.[doi:10.16089/j.cnki.1008-2786.000533]
　XIE Xiangping,WANG Xiaojun,YAN Chunling.A Review of the Research on Woody Debris Related Disaster and Its Prospect[J].Mountain Research,2020,(6):552.[doi:10.16089/j.cnki.1008-2786.000533]
[9]姚鸿坤,邹 强*,蒋 虎,等.泥石流承灾体群体效应特征反演——以7·12四川平武县黑水沟泥石流为例[J].山地学报,2024,(1):47.[doi:10.16089/j.cnki.1008-2786.000803]
　YAO Hongkun,ZOU Qiang*,JIANG Hu,et al.Collective Effects of Debris Flow Striking on Structure Clusters Downstream: A Case Study of the Heishui Gully Debris Flow on July 12th in Pingwu County, Sichuan Province, China[J].Mountain Research,2024,(6):47.[doi:10.16089/j.cnki.1008-2786.000803]

备注/Memo

备注/Memo:
收稿日期(Received date):2019-06-13; 改回日期(Accepted date):2020-01-02
基金项目(Foundation item):陕西省自然科学基础研究计划资助项目(2019JQ-206)。[Natural Science Basic Research Plan in Shaanxi Province of China(2019JQ-206)]
作者简介(Biography):徐根祺(1984-),男,陕西西安人,硕士研究生,工程师,主要研究方向:人工智能算法。 [XU Genqi(1984-), male, born in Xi'an, Shanxi province, M.Sc. candidate, engineer, research on artificial intelligence algorithms AI] E-mail: 2017041028@stu.xpu.edu.cn
更新日期/Last Update: 2019-11-30