基于3種算法的海南州泥石流易發(fā)性預(yù)測及致災(zāi)因子研究
關(guān)鍵詞:泥石流;致災(zāi)因子;易發(fā)性預(yù)測;RF算法;GBDT算法;XGBoost算法;海南藏族自治州中圖分類號:P642.22 文獻(xiàn)標(biāo)志碼:A doi:10.3969/j.issn.1000-1379.2025.07.003引用格式:胡少偉,郭要輝,葉宇霄,等.基于3種算法的海南州泥石流易發(fā)性預(yù)測及致災(zāi)因子研究[J].人民黃河,2025,47(7):13-19.
Study on Debris Flow Susceptibility Prediction and Disaster-Causing Factors in Hainan Based on Three Algorithms
HU Shaowei 1,2 ,GUO Yaohui2,YE Yuxiao1, LIAO Yi 1, ZHANG Zhiwei2,,LI Liang2 (1.College of Water Conservancy and Transportation, Zhengzhou University,Zhengzhou 45Oo01,China; 2.School of Civil Engineering, Chongqing University,Chongqing 4Oo045,China)
Abstract:Inodertosrentedisasterausingfactors tatiducedebrisfows,evaluatetepefoanceof tetreemachilaingal gorithmsofRF,GBDndXGBot,preditigtesuseptiblityfebrisos,ndprovdeferenesfgeolocalsastert anddisasterpreventionandmitigationintheareaspronetodebrisflows,akingHainanTibetanAutonomousPrefecture,QinghaiProvinceas thestudyaraandasedonstorcaldebrisflowisasterdataasedoteitiallseleced7ifuencngfactorsandteeasoe tioncoeficientfebrsfodssters,edsstercausigfctorstatiducedebrsereseedTemaingdisastecausigc torswereclasifidnd8ombatiosofdisaster-ausingfctos ereset.Tethe macheangalgiofRF,GBD,andXGoost wereusedtopredictthesusceptibilityofebrisflow.Thpredictionfectwasevaluatedyindicatorssuchasaccuracyateprecisiorate, recallrate,soendOAUC.esulssotata)distaefroeiverhael,levatio,oilrodibilityopgacs tureindexaliallaoias ture,curvature,lointofvatiooevationdsptareisastecaingfactosatiduedebrisflowiestudy area.Amogtlevatiostaefroiveaeloliltpogapcoedalfallal vegetationndexaretheaindisaster-causingfactos.b)WhentetreealgoitsofRF,GBDandXGBoostareusedtopredictthlikeliodfebrisflow,thepredictionfctbasedontedisaster-causingfactrcombiationC7(tiscombiationdosnotcosiderslopedirection)isthebst,andthepredictionefectsbasedontedisaster-causigfactorcombnatiosC5(thiscombnationdoesnotcosiderurvature,coeficientofvarationofelevationandlopedirection)andC8(thisombiationcosidersalldisaster-ausingfactors)arealso goodc)Whenpredictingthesuseptibilityofebrisflowsbasedonthdsaster-causingfactorcombinationC7,theankingofteadvantages and disadvantages of the three algorithms is XGBoost, GBDT and RF.
Keywords:debrisflow;disaster-causingfactors;susceptibilityprediction;RFalgori;GBDalgorit;XGBostalgoritHaan betan Autonomous Prefecture
0 引言
泥石流是由強(qiáng)降水或洪水等誘發(fā)的一種地質(zhì)災(zāi)害,其挾帶大量泥沙、石塊和巨礫等固體顆粒沿山谷或坡地快速流動(dòng),具有突發(fā)性強(qiáng)、破壞力大、災(zāi)害嚴(yán)重、波及范圍廣等特點(diǎn)[1-4]。(剩余8071字)
- 耦合地理探測器和機(jī)器學(xué)習(xí)模型的...
- 基于熵指數(shù)法的黃河龍羊峽一積石...
- 基于3種算法的海南州泥石流易發(fā)...
- 地震作用下混凝土面板堆石壩三維...
- 地震-滑坡涌浪疊加下高拱壩動(dòng)力...
- 基于邏輯回歸模型的黃河上游段滑...
- 黃河上游干流庫壩群段多災(zāi)種大數(shù)...
- 黃河上游干流庫壩群段重大自然災(zāi)...
- 基于SMOTE策略的數(shù)據(jù)不完整...
- 基于先驗(yàn)知識的甘肅省積石山震后...
- 基于耦合數(shù)值模型的降雨誘發(fā)滑坡...
- 基于SBAS-InSAR的黃河...
- 基于典型小概率法-LSTM的大...
- 黃河上游多災(zāi)種多源預(yù)警數(shù)據(jù)庫管...
- 面向?yàn)?zāi)害損失評估的黃河流域自然...
- 基于Word2Vec模型的泥石...
- 基于多源信息融合的黃河上游多災(zāi)...
- 水庫漫壩的潰決過程與數(shù)學(xué)表達(dá)...
- 基于集成學(xué)習(xí)和考慮滑坡負(fù)樣本的...
- 基于復(fù)雜網(wǎng)絡(luò)與數(shù)值模擬的暴雨-...
- 基于多維數(shù)據(jù)轉(zhuǎn)換的混凝土壩異常...
- 基于深度自編碼器的混凝土壩變形...
- 基于深度學(xué)習(xí)模型的混凝土壩變形...
- 考慮多點(diǎn)監(jiān)測數(shù)據(jù)的混凝土壩智能...
- 基于CORDEX數(shù)據(jù)的黃河上游...
- 基于后向散射特性的SAR遙感水...