基于关联规则的坡面土壤侵蚀评价模型与方法研究
|
摘要
土壤侵蚀不仅是地域问题,而且是全球性的环境问题,已引起广泛的研究和关注。本文以鄂西北丹江库区房县和郧西县为研究对象,结合房县和郧西县现有土壤侵蚀原始数据,研究具有代表性的鄂西北山区丹江库区径流小区的坡面土壤侵蚀状况。主要内容：
在综述国内外研究概况和研究方法的基础上,结合鄂西北山区的相关径流小区观测数据,引用关联规则中的Apriori算法,首先对坡度、植被覆盖度、土地利用类型,植被对应的措施等各种不同因子进行量化,然后分别实现在不同耕作方式、不同坡度、不同植被覆盖度因子条件下,以及综合因子影响下,他们对土壤侵蚀状况的影响,然后综合进行影响因子与坡面土壤侵蚀之间的关联规则分析,以算法的支持度与置信度为基础,实现各因子与土壤侵蚀状况的重要性程度关联,进而得到各因子与土壤侵蚀之间的重要性关系,得到研究区坡面径流小区土壤侵蚀评价指标体系。
到目前为止国内外已研究出多种土壤侵蚀评价模型,其中尤以美国的通用土壤流失方程(USLE)系列为代表,是最具有影响力的土壤侵蚀评价统计模型之一。它是基于美国10000多个坡面小区近30年的观测数据而得到的经验统计模型,直到现在仍被世界土壤保护工作及水土流失相关部门和研究者广泛采用。
鉴于基于通用土壤侵蚀方程中的各个因子,正是基于关联规则支持下坡面土壤侵蚀评价指标体系中对土壤侵蚀影响较大的因子,因此该方程在研究区坡面土壤侵蚀分析中具有特别的适用性。本文首先介绍了通用土壤侵蚀方程的原型,详细探讨了研究区通用土壤侵蚀方程各个因子的计算方法,并根据数据库资料,推导出适合房县的通用土壤侵蚀方程。根据推导的方程计算相关年份的土壤侵蚀量,并进行误差分析。试验结论表明该模型在研究区具有较好的效果和适用性。
基于数据的机器学习问题已经成为目前许多行业应用于研究的重点。支持向量机作为机器学习的一种,其研究的主要问题是从一组观测数据集的数据出发,得到一些不能通过原理分析而得到的规律,进而研究这些规律对未来的数据或者无法观测的数据进行预测和分析。本文介绍支持向量机和BP神经网络的基本思想,将二者应用到鄂西北山区土壤侵蚀预测中,选取的影响因子是研究区坡面土壤侵蚀评价指标体系中对坡面土壤侵蚀影响最大的降雨量及相关因子。对研究区坡面土壤侵蚀状况进行评价并作误差分析。结论表明基于神经网络和支持向量机技术的研究区水土流失评价,总体而言其预测结果比较理想,支持向量机极个别小区预测效果不太理想。究其原因,虽然SVM具有较高的泛化能力和计算效率,但SVM对样本的要求比较高。这种预测结果的出现,跟样本的有限性及样本的精确度、SVM参数的选取等都有很大的关系。
Soil erosion concerns are both local and completely global, and it has attracted much attention and research. This paper addresses the issue of hillslope Soil erosion of Danjiangkou Reservoir region through artificial slope runoff plots of FangXian and YunXi county. In this paper, the author First introduces the research data Sources of Danjiangkou Reservoir region and the data uploading to the SQL server database. Based on the overview of oversea and domestic research generalization and main assessment methods, The paper presents its adopted three kands of model or research methods. The first adopted method is the Aprior algorithm based on association rule aiming at qualitative analysizing the pertinence relation between the slope Soil erosion and its influence factors. The second method is to build the Universal Soil Loss Equation and apply to some county in Danjiangkou Reservoir region. Setting up the support vector machines(SVM)modle for forecasting the soil erosion modulus.
According to the current data of research area stored in SQLserver database, this paper first explains related concepts of da ta mining, and then select the Apriori algorithm for Association Rule analysizing among the hillslope soil erosion and its influence factors,such as farming methods factor, ratio of slope factor, forest vegetation coverage factor. By means of quantizing and achieving related analysizing these influence factors, we obtain the qualitative interdependence coefficients between the hillslope erosion and these factors.
Universal Soil Loss Equation model(USLE) is one of the most influential evaluateion statistical modle. And until now it is still adopted by the soil and water conservation workers and relevant department researchers in wide use. Because of the model's special fitness for researching hillslope soil erosion, the paper first introduces the prototype model of USLE, and then talks in detail about the computing method for every influence factor of the USLE model. In the end the USLE modle applicable to the Danjiangkou Reservoir region is successfully derived from the prototype model and available research data. we calculate the amount of soil erosion on hillslopesin respective year from above equations. Ultimately the error analysis has conducted. Teh experiments show that this method is satisfactory, practical and effective.
Machine Learning based on experimental data has become the focus of currently study in many industries. As two kinds of machine leardnings, back-propagation neural network(BPNN) and SVM algorithm shed light on some sets of observed data and discover some regularities which can't be obtained by principle analysis. And simultaneously these two models can use these regularities for forecasting in connection with the future or inaccessible data.
This article first gives a brief introduction to the regression principle of BPNN and SVM model. And then the paper gets the predicting result of the amount of hillslope soil erosion and error analysis by using these models based on the experimental data of the research area. The results show that the two methods,especially the BPNN method, are very effective in connection with our research data. At the same time,there are several undesirable Predicting Outcomes. The reasons for this have to do with the limitations of existing data, experimental data precision and the parameter selection of the model.
在综述国内外研究概况和研究方法的基础上,结合鄂西北山区的相关径流小区观测数据,引用关联规则中的Apriori算法,首先对坡度、植被覆盖度、土地利用类型,植被对应的措施等各种不同因子进行量化,然后分别实现在不同耕作方式、不同坡度、不同植被覆盖度因子条件下,以及综合因子影响下,他们对土壤侵蚀状况的影响,然后综合进行影响因子与坡面土壤侵蚀之间的关联规则分析,以算法的支持度与置信度为基础,实现各因子与土壤侵蚀状况的重要性程度关联,进而得到各因子与土壤侵蚀之间的重要性关系,得到研究区坡面径流小区土壤侵蚀评价指标体系。
到目前为止国内外已研究出多种土壤侵蚀评价模型,其中尤以美国的通用土壤流失方程(USLE)系列为代表,是最具有影响力的土壤侵蚀评价统计模型之一。它是基于美国10000多个坡面小区近30年的观测数据而得到的经验统计模型,直到现在仍被世界土壤保护工作及水土流失相关部门和研究者广泛采用。
鉴于基于通用土壤侵蚀方程中的各个因子,正是基于关联规则支持下坡面土壤侵蚀评价指标体系中对土壤侵蚀影响较大的因子,因此该方程在研究区坡面土壤侵蚀分析中具有特别的适用性。本文首先介绍了通用土壤侵蚀方程的原型,详细探讨了研究区通用土壤侵蚀方程各个因子的计算方法,并根据数据库资料,推导出适合房县的通用土壤侵蚀方程。根据推导的方程计算相关年份的土壤侵蚀量,并进行误差分析。试验结论表明该模型在研究区具有较好的效果和适用性。
基于数据的机器学习问题已经成为目前许多行业应用于研究的重点。支持向量机作为机器学习的一种,其研究的主要问题是从一组观测数据集的数据出发,得到一些不能通过原理分析而得到的规律,进而研究这些规律对未来的数据或者无法观测的数据进行预测和分析。本文介绍支持向量机和BP神经网络的基本思想,将二者应用到鄂西北山区土壤侵蚀预测中,选取的影响因子是研究区坡面土壤侵蚀评价指标体系中对坡面土壤侵蚀影响最大的降雨量及相关因子。对研究区坡面土壤侵蚀状况进行评价并作误差分析。结论表明基于神经网络和支持向量机技术的研究区水土流失评价,总体而言其预测结果比较理想,支持向量机极个别小区预测效果不太理想。究其原因,虽然SVM具有较高的泛化能力和计算效率,但SVM对样本的要求比较高。这种预测结果的出现,跟样本的有限性及样本的精确度、SVM参数的选取等都有很大的关系。
Soil erosion concerns are both local and completely global, and it has attracted much attention and research. This paper addresses the issue of hillslope Soil erosion of Danjiangkou Reservoir region through artificial slope runoff plots of FangXian and YunXi county. In this paper, the author First introduces the research data Sources of Danjiangkou Reservoir region and the data uploading to the SQL server database. Based on the overview of oversea and domestic research generalization and main assessment methods, The paper presents its adopted three kands of model or research methods. The first adopted method is the Aprior algorithm based on association rule aiming at qualitative analysizing the pertinence relation between the slope Soil erosion and its influence factors. The second method is to build the Universal Soil Loss Equation and apply to some county in Danjiangkou Reservoir region. Setting up the support vector machines(SVM)modle for forecasting the soil erosion modulus.
According to the current data of research area stored in SQLserver database, this paper first explains related concepts of da ta mining, and then select the Apriori algorithm for Association Rule analysizing among the hillslope soil erosion and its influence factors,such as farming methods factor, ratio of slope factor, forest vegetation coverage factor. By means of quantizing and achieving related analysizing these influence factors, we obtain the qualitative interdependence coefficients between the hillslope erosion and these factors.
Universal Soil Loss Equation model(USLE) is one of the most influential evaluateion statistical modle. And until now it is still adopted by the soil and water conservation workers and relevant department researchers in wide use. Because of the model's special fitness for researching hillslope soil erosion, the paper first introduces the prototype model of USLE, and then talks in detail about the computing method for every influence factor of the USLE model. In the end the USLE modle applicable to the Danjiangkou Reservoir region is successfully derived from the prototype model and available research data. we calculate the amount of soil erosion on hillslopesin respective year from above equations. Ultimately the error analysis has conducted. Teh experiments show that this method is satisfactory, practical and effective.
Machine Learning based on experimental data has become the focus of currently study in many industries. As two kinds of machine leardnings, back-propagation neural network(BPNN) and SVM algorithm shed light on some sets of observed data and discover some regularities which can't be obtained by principle analysis. And simultaneously these two models can use these regularities for forecasting in connection with the future or inaccessible data.
This article first gives a brief introduction to the regression principle of BPNN and SVM model. And then the paper gets the predicting result of the amount of hillslope soil erosion and error analysis by using these models based on the experimental data of the research area. The results show that the two methods,especially the BPNN method, are very effective in connection with our research data. At the same time,there are several undesirable Predicting Outcomes. The reasons for this have to do with the limitations of existing data, experimental data precision and the parameter selection of the model.
引文
[1]Greenland D L. Soil science and sustainable land management. In:Syers J K and Rimmer D L (eds.). Soil science and sustainable land management in the tropics. CAB International,1994,1-15
[2]Yoder Daniel, et al.The future of RUSLE:Inside new Revised Universal Soil Loess Equation.Soil and Water Conservation.1995,50(5):484-489
[3]Simon Haykin. Neural networks, a comprehensive foundation, second edition, Prentice Hall,1998.161-175,400-438.
[4]Hertz J, et al. Introduction to Theory of Neural Compution. Sant Fee Complexity Science Series,1991,156
[5]Rumelhart D. E, Hinton G. E and Williams R. J. Leaning Internal Representations by ErrorPropagation. Parallel Distributed Processing:Explorationsin the Microstructures of Cognition. MIT Press, MA.1986.318-362
[6]Hecht-Nielsen R. Theory of the backpropagation neural network. In IEEE Int. Joint Conf. on Neural Networks,1989,1:593-605
[7]Vapnik V N. Statistical learning theory{M}.New York,1998.
[8]Cortes C,Vapnik V N.Support vector networks[J]. Machine Learning.1995,20: 273-297
[9]Zhang X. Using class-center vectors to build support vector machines [C].In:Proceedings of NNSP'99,1999
[10]吴以教,张胜利,孟庆枚.黄河流域严重水土流失区小流域治理规划方法探讨[J].1982,2
[11]詹小国,谭德宝,朱永清,黄奇.基于RS和GIS的三峡库区水土流失快速动态监测研究[J].长江科学院院报2003,1
[12]Watrous R L. Learning algorithms for connectionist networks:applied gradient method of nonlinear optimization. In:Proc IEEE 1st Int Conf Neural Networks. San Diego, CA,1987,619-627
[13]Voge T, wangis I,et al. Accelerating and convergence of the back propagation method. Biol Cybermeties,1998,59:257-261
[14]蒋平.石瓮小流域土壤侵蚀遥感定量和制图的研究[J].国土与自然资源研究,1994,2
[15]刘建善.天水水土流失测验的初步分析[J].科学通报,1953,(12)：5965.
[16]史更申.鲁南山区水土流失加剧_演变及防治[J].中国水土保持1990,5
[17]郑粉莉,王占礼,杨勤科.我国水蚀预报模型研究的现状、挑战与任务[J].中国水土保持科学,2005,3(1)：7-14
[18]张汉雄.黄土高原重点水土流失区农业资源合理利用的动态模拟[J].中国水土保持,1991,9
[19]卜兆宏；孙金庄；周伏建.水土流失的定量遥感及其应用研究进展[J].农村生态环境,1995,4
[20]马国斌.基于GIS的黄土高原小流域土壤侵蚀定量评价研究[J].南京师范大学地理科学学院2003,5硕士论文
[21]万晔,段昌群,王玉朝,王学林,王建萍.基于3S技术的小流域水土流失过程数值模拟与定量研究[J].水科学进展,2004,5
[22]王小丹,钟祥浩,范建容.西藏水土流失敏感性评价及其空间分异规律[J].地理学报,2004,2
[23]Hoseyamada K, The effects of rainfall and soil properties on farmland, conservation, Irrig. Eng。and Rural Planning Jpn. SOc. Irrig. and Reclamation Eng.,1986.9:5-14
[24]Bajracharya R M. Lal R. Seasonal soil loss and erodibillty variation on a Miamian sih loam soil. Soil Sei. SOe. Am. J.,1992,56:1560-1565
[25]Rejman J, Turski R, Palnszek J. Spatial and temporal variations in erodibility of loess soil. Soil and Tillage Research,1998,46:61-68
[26]Wall G J, Dickinson W T. RudraR P,et al. Seasonal soil erodibility variation in Southwestern Ontario. Can. J. Soil Sei,1988.68:417-424
[27]卢彦,刘勇,崔丽红.基于GM(1,1)模型对齐齐哈尔市水土流失治理面积的预测[J].黑龙江水利科技2006,6
[28]孙佳,紫色土土壤侵蚀评价指标体系研究,西南大学硕士学位论文,2007.5
[29]Zingg A W.degree and length of land slope as it affects soil loss in runoff. Agricultural engineering,1940,(21):59-64
[30]邱扬,傅伯杰.异质景观中水土流失的空间变异与尺度变异[J].生态学报,2004,2
[31]鞠文学,李天宏,倪晋仁.隶属度判别法在水土流失快速评估中的应用[J].地理科学进展2003,22(4)
[32]United States Department of Agriculture. EPIC-Erosion/productivity Impact Calculator 1. Model Documentation. Technical Bulletin Numher 1768. Washington D. C., USDA-ARS,1990
[33]魏兴萍,赵纯勇,杨华.基于GIS的小流域土壤侵蚀评价研究[J].重庆师范大学学报(自然科学版),2005,22(4)
[34]赖奕卡.坡面土壤侵蚀影响因子研究进展,亚热带水土保持,2008.20(1)：12-16
[35]龙栋材,李斌兵.基于支持向量机的小流域水蚀预报模型研究[J].中国水土保持科学,2006,4(6)
[36]李如忠.基于模糊理论的流域土壤侵蚀量计算[J].环境科学与技术,2007,30(3)
[37]Wischmeier W H, Smith D D. Predicting Rainfall Erosion Losses from Cropland East of the Rocky Mountains. Agricultural Handbook No.282[M]. Washington D C:USDA,1965
[38]刘宝元,谢云,张科利.土壤侵蚀预报模型[M].北京：中国科学技术出版社,2001
[39]江忠善,郑粉莉.坡面水蚀预报模型研究[J].水土保持学报,2004,18(1)：66-69.
[40]Williams J R,Neitsch S L,Arnold J G. Soil and Water Assessment Tool User's Manual [M]. Texas Blackland Research Center:Texas Agricultural Experiment Station
[41]Morgan R P C. The European Soil Erosion Model (EUROPEM):a dynamic approach for predicting sediment transport from field and small catchment[J]. Earth surface and Landform,1998,23:527-544
[42]De Roo A P J. LISEM:a single event physical-based hydrologic and soil ersion model for drainage basins Ⅰ:Theory, input and output[J]. Processes,1996,10(8):1 119-1126
[43]江忠善.黄土地区天然降雨雨滴特征研究[J].中国水土保持,1983(3)：32-36.
[44]De Jong S M,Paracchini M L,Bertolo F,et al. Regional assessment of soil erosion using the distributed mode SEMMED and remotely sensed data[J]. Catena,1999,37:291-308
[45]史培军,刘宝元,张科利.土壤侵蚀过程与模型研究[J].资源科学,1999,21(5)：9-18
[46]卢光发.十堰市旅游地质资源及其保护[J].湖北地矿,2000,14(1)：40-44
[47]陈华,郭生练,柴晓玲.汉江丹江口以上流域降水特征及变化趋势分析[J].人民长江,2005,36(11)：29-31
[48]张利华,梁俊,蒋金龙,陈于.基于RS-GIS的湖北丹江库区土壤水力侵蚀定量分析[J].生态环境,2006,15(6)：1319-1323)
[49]赵光耀,赵兴华,王答相.陕西省丹江口水库水源区生态保护与重建存在的问题及对策[J].中国水土保持SWCC,2003,11 (2):11-12
[50]宋轩,杜丽平.基于GIS得淮河流域伏牛山区土壤侵蚀动态变化研究[J].中国水土保持,2008(2)
[51]NOEL D U, JAMES A L. The dynamics of soil erosion in USagriculture [J]. The Science of the Total Environment,1998,218:45-58
[52]杨存建,刘纪远,张增祥等.GIS支持下不同坡度的土壤侵蚀特征分析[J].水土保持学报,2002,16(6)：46-49
[53]赵文武,傅伯杰等.多尺度土壤侵蚀评价指数的技术与方法[J].地理科学进展,2007(2)
[54]陈三维,谢莉等.广东大宝山矿区土壤侵蚀强度判别方法[J].中国水土保持,2007(11)
[55]黄艳艳,曾志远,陈少辉.MODIS数据在土壤侵蚀动态监测中的应用[J].水土保持通报,2007,27(5)
[56]何丽君,董蕊,袁克杰.常见关联规则算法分析与比较.大连民族学院学报.2005.7.5：39-42
[57]李勉,李占斌,刘普灵.中国土壤侵蚀定量研究进展[J].水土保持研究,2002,9(3)：243-248
[58]张金英,葛京凤,于江海.基于RS和GIS的河北太行山土壤侵蚀强度研究[J].安徽农业科学.2006,34(24)：6391-6395
[59]姚水萍,任佶.浙江省土壤侵蚀等级划分模糊综合评判模型的初步探讨.水土保持通报,2006,26(6)：32-34
[60]汪东川,卢玉东.国外土壤侵蚀模型发展概述[J].中国水土保持科学,2004,2(2)：35-40
[61]龙栋材,李斌兵.基于支持向量机的小流域水蚀预报模型研究[J].中国水土保持科学,2006,4(6)：48-51
[62]邵峰晶,于忠清.数据挖掘原理与算法[J].中国水利水电出版社,2003.8
[63]刘淑珍,刘海军,钟祥浩等,西藏高原土壤侵蚀评价体系及监测方法,山地学报,2009.27(1)：76-81
[64]陈华,郭生练,柴晓玲等.汉江丹江口以上流域降水特征及变化趋势分析[J].人民长江,2005,36(11)：29-31
[65]张俊飚.中国土壤侵蚀影响因素及其危害分析[J].云南环境科学,2001,20(2)：4-7
[66]李雄飞,李军.数据挖掘与知识发现[M].北京：高等教育出版社,2003,5
[67][意]Paolo Giudici著.袁方,王煜,王丽娟等译.实用数据挖掘[M].北京：电子工业出版社,2004,6
[68]韩家炜等.Web挖掘研究.计算机研究与发展,2001,04
[69]水利部水土保持司.土壤侵蚀分类分级标准(SL 190-96).中华人民共和国水利部,1997
[70]USDA(United States Department of Agriculture)703手册, http://naldr. nal.usda. gov/Exe/ZyNET.exe?Client=National+Agricultural+Library+Digital+Repository&Init=0&ZyAc tion=ZyActionL%7C4&User=ANONYMOUS&Cleartext=&ZyActionlD=0&ImageQuality=r10 5g16%2Fr105g16%2Fx150y150g16%2Fi600&ZyEntry=0&Query=&Index=AH%7CAH2%7C AIB%7CBIC%7CBooks%7CCirc%7CERS%7CFVMNR%7CHGB%7CJAR%7CMP%7CNFC %7CPI%7CROS%7CRural%7CTB%7CTB1%7CTB2%7CUSDA Div Bulletin%7CWPC%7C YOA1%7CYOA2&File=&SortMethod=h%7C-&MaximumDocuments=10&FuzzyDegree=-1& MaximumPages=1&Time=&EndTime=&ConceptsFile=&SeekPage=&DefSeekPage
[71]杨艳生.区域性土壤流失预测方程得初步研究[J].土壤学报,1990,27(1)
[72]杨艳生,梁音,刘佳桂.长江三峡区土壤坡面流失及重力侵蚀[J].水土保持学报,1991
[73]刘宝元,谢云,张科利.土壤侵蚀预报模型[M].北京：中国科学技术出版社,2001
[74]李辉,陈晓玲.不同空间尺度下的土壤侵蚀模型研究进展[J].华中师范大学学报,2006,40(4)
[75]Z.H. Shi, Soil conservation planning at the small watershed level using RUSLE with GIS:a case study in the Three Gorge Area of China, CATENA,2003
[76]Zhi-Hua. Shi Research on the SCS-CN initial abstraction ratio using rainfall-runoff event analysis in the Three Gorges Area, China, CATENA,2008
[77]B. J. Fu, Assessment of Soil Erosion at Largewatershed Scale Using RUSLE and GIS:a Case Study in the Loess Plateau of China, land degradation & development Land Degrad. Develop.16:73-85 (2005)
[78]Wischmeier W. H. and Smith D.D., Rainfall energy and its relationship to soil loss. Transactions of the American Geophysical Union,1958,39:285-291
[79]蔡崇法,丁树文.应用USLE模型与地理信息系统IDRISI预测小流域土壤侵蚀量 得研究[J].水土保持学报,2000,14(2)
[80]王万忠,焦菊英.中国土壤侵蚀因子定量评价研究[J].水土保持通报,1996,16(5)
[81]章文波,谢云,刘元宝.利用日雨量计算降雨侵蚀力的方法研究.地理科学,2002,22(06)：705-711
[82]M.J.柯克比和R.P.C.摩根编著,王礼先等译,土壤侵蚀,北京：水利电力出版社,1987
[83]RomkensM.J.M., Prasad S.N. and Poesen J.W.A., Soil erodibility and properties. In: Transactions of theⅧCongress of the International Society of Soil Science, Hamburg,1986,492-504
[84]Declercq F. and Poesen J., Evaluation of twomodels to calculate the soil erodibility factor k. Pedologie,1992, XLII (2):158-169
[85]卜兆宏、李全英.土壤可蚀性(K)值图编制方法的初步研究[J].遥感技术与应用,1994,9(4)：22-27
[86]Wischmeier W.H. and Smith D.D., Predicting rainfall erosion losses-A guide to conservation planning with the Universal Soil Loss Equation (USLE). Agriculture Handbook No.537, United States Department of Agriculture, Springfield, USA. 1978
[87]罗志军,张俊.土壤侵蚀模型的研究现状与展望[J].安徽农业科学,2007,35(27)：8547-8576
[88]江忠善,郑粉莉,武敏.中国坡面水蚀预报模型研究[J].泥沙研究,2005(4)
[89]谢云,刘宝元,章文波.侵蚀性降雨标准研究[J].土壤侵蚀与水土保持学报[J].2000,14(4)：6-11
[90]王万忠,焦菊英.中国土壤侵蚀因子定量评价研究[J].水土保持通报,1999,16(5)
[91]白鹏、张喜斌,张斌,李彦,谢文俊,刘君华.支持向量机理论及工程应用实例[J].西安电子科技大学出版社,2008,8
[92]龙栋材,李斌兵.基于支持向量机的小流域水蚀预报模型研究[J].中国水土保
持科学,2006,4(6)
[93][英]Nello Cristianini, John Shawe-Taylor著,李国正,王猛,曾华军译,支持向量机器导论(an introduction to support vector machines and other kernel-based learning methods)电子工业出版社,北京,2004,3
[94]Vapnik V N. An overview of statistical learning theory[J].IEEE Trans on Neural Networks.1999,10(5):988-999
[95]杜熊禹.用于数据挖掘的支持向量机算法研究[D].电子科技大学,2007
[96]http://map.sogou.com/#city=%u6E56%u5317&lq=%u4E39%u6C5F%u5E93%u533 A&page=1&hb=1,1&c=12384500,3856500,8
[97]中华人民共和国行业标准SL277-2002.北京：中国水利水电出版社,2002：
[98]张科利,刘宝元,蔡永明.土壤侵蚀预报研究中得标准小区问题论证[J].地理研究,2000,19(3)
[99]黄煜煜,李勇志,刘予时.房县地区地形因子对水土流失的影响[J].湖北水利水电职业技术学院学报,2006
[100]张洪江.土壤侵蚀原理[M],北京：中国林业出版社,1999,10
[101]WischmeierW H, et al. A soil erodibility nomorgraph farm land and construction sites[J]. Journal of Soil andWater Conservation,1971,26:189-193
[102]杨子生.滇东北山箧坡耕地土壤流失方程研究[J].水土保持通报,1999,19(1)
[103]Jacobs R. Increased rates of convergence through learning rate adaptation. Neural Networks,1988,1:295-307
[104]Chauvin Y and Rumelhart D. E. Backpropagation:Theory,Architectures and Applications. Hillsdale, NJ:LawrenceErlbaum Associates,1995
[105]王春香,付云中,王永章,路华,基于BP神经网络的张力控制系统,中国机械工程,2001.12(8)：873-875
[106]D.C. Flanagan, J.C. Ascough Ⅱ, A.D. Nicks, M.A. Nearing and J.M. Laflen Overview of the WEPP erosion prediction model,1995
[107]王万忠,焦菊英.中国的土壤侵蚀因子定量评价研究[J].水土保持通报,1996,
16(5)：1-20
[108]李甚悟,李君莲.降雨及土壤湿度对水土流失的影响[J].土壤学报,1992,29(1)：94-103
[109]Y. k Bissonnais, B. Renaux, H. Delouche. Interactions between soil properties and moisture content in crust formation, runoff and interill erosion from tillde loess soils[J]. CATENA,1995,25:33-46
[110]Bernard B, Eric R. Aggregate stability as all indicator of soil susceptibility to runoff and erosion:validation at several levels [J]. Catena,2002,47:133-149
[111]Gyssels G, Poesen J. The importance of plant root characteristics in controlling concentrated flow erosion rates[J]. Earth Surface Processes and Landforms,2003, 28:371-384
[112]Hofzullah Aksoy, M. Levent Kawas. A review of hillslope and watershed scale erosion and sediment transport models [J]. Catena,2005,64:247-271
[113]Misra, R. K, Rose, C. W. Application and sensitivity analysis of process based erosion model-GUEST[J]. European Journal Soil Science,1996,10:593-604
[114]Casermeim M. A., Molina J. A., de la Cruz Caravaea M T. Influence of scrubs on runoff and sediment loss in soils of Medeterranean climate climate[J]. Catena, 2004,57:91-107
[2]Yoder Daniel, et al.The future of RUSLE:Inside new Revised Universal Soil Loess Equation.Soil and Water Conservation.1995,50(5):484-489
[3]Simon Haykin. Neural networks, a comprehensive foundation, second edition, Prentice Hall,1998.161-175,400-438.
[4]Hertz J, et al. Introduction to Theory of Neural Compution. Sant Fee Complexity Science Series,1991,156
[5]Rumelhart D. E, Hinton G. E and Williams R. J. Leaning Internal Representations by ErrorPropagation. Parallel Distributed Processing:Explorationsin the Microstructures of Cognition. MIT Press, MA.1986.318-362
[6]Hecht-Nielsen R. Theory of the backpropagation neural network. In IEEE Int. Joint Conf. on Neural Networks,1989,1:593-605
[7]Vapnik V N. Statistical learning theory{M}.New York,1998.
[8]Cortes C,Vapnik V N.Support vector networks[J]. Machine Learning.1995,20: 273-297
[9]Zhang X. Using class-center vectors to build support vector machines [C].In:Proceedings of NNSP'99,1999
[10]吴以教,张胜利,孟庆枚.黄河流域严重水土流失区小流域治理规划方法探讨[J].1982,2
[11]詹小国,谭德宝,朱永清,黄奇.基于RS和GIS的三峡库区水土流失快速动态监测研究[J].长江科学院院报2003,1
[12]Watrous R L. Learning algorithms for connectionist networks:applied gradient method of nonlinear optimization. In:Proc IEEE 1st Int Conf Neural Networks. San Diego, CA,1987,619-627
[13]Voge T, wangis I,et al. Accelerating and convergence of the back propagation method. Biol Cybermeties,1998,59:257-261
[14]蒋平.石瓮小流域土壤侵蚀遥感定量和制图的研究[J].国土与自然资源研究,1994,2
[15]刘建善.天水水土流失测验的初步分析[J].科学通报,1953,(12)：5965.
[16]史更申.鲁南山区水土流失加剧_演变及防治[J].中国水土保持1990,5
[17]郑粉莉,王占礼,杨勤科.我国水蚀预报模型研究的现状、挑战与任务[J].中国水土保持科学,2005,3(1)：7-14
[18]张汉雄.黄土高原重点水土流失区农业资源合理利用的动态模拟[J].中国水土保持,1991,9
[19]卜兆宏；孙金庄；周伏建.水土流失的定量遥感及其应用研究进展[J].农村生态环境,1995,4
[20]马国斌.基于GIS的黄土高原小流域土壤侵蚀定量评价研究[J].南京师范大学地理科学学院2003,5硕士论文
[21]万晔,段昌群,王玉朝,王学林,王建萍.基于3S技术的小流域水土流失过程数值模拟与定量研究[J].水科学进展,2004,5
[22]王小丹,钟祥浩,范建容.西藏水土流失敏感性评价及其空间分异规律[J].地理学报,2004,2
[23]Hoseyamada K, The effects of rainfall and soil properties on farmland, conservation, Irrig. Eng。and Rural Planning Jpn. SOc. Irrig. and Reclamation Eng.,1986.9:5-14
[24]Bajracharya R M. Lal R. Seasonal soil loss and erodibillty variation on a Miamian sih loam soil. Soil Sei. SOe. Am. J.,1992,56:1560-1565
[25]Rejman J, Turski R, Palnszek J. Spatial and temporal variations in erodibility of loess soil. Soil and Tillage Research,1998,46:61-68
[26]Wall G J, Dickinson W T. RudraR P,et al. Seasonal soil erodibility variation in Southwestern Ontario. Can. J. Soil Sei,1988.68:417-424
[27]卢彦,刘勇,崔丽红.基于GM(1,1)模型对齐齐哈尔市水土流失治理面积的预测[J].黑龙江水利科技2006,6
[28]孙佳,紫色土土壤侵蚀评价指标体系研究,西南大学硕士学位论文,2007.5
[29]Zingg A W.degree and length of land slope as it affects soil loss in runoff. Agricultural engineering,1940,(21):59-64
[30]邱扬,傅伯杰.异质景观中水土流失的空间变异与尺度变异[J].生态学报,2004,2
[31]鞠文学,李天宏,倪晋仁.隶属度判别法在水土流失快速评估中的应用[J].地理科学进展2003,22(4)
[32]United States Department of Agriculture. EPIC-Erosion/productivity Impact Calculator 1. Model Documentation. Technical Bulletin Numher 1768. Washington D. C., USDA-ARS,1990
[33]魏兴萍,赵纯勇,杨华.基于GIS的小流域土壤侵蚀评价研究[J].重庆师范大学学报(自然科学版),2005,22(4)
[34]赖奕卡.坡面土壤侵蚀影响因子研究进展,亚热带水土保持,2008.20(1)：12-16
[35]龙栋材,李斌兵.基于支持向量机的小流域水蚀预报模型研究[J].中国水土保持科学,2006,4(6)
[36]李如忠.基于模糊理论的流域土壤侵蚀量计算[J].环境科学与技术,2007,30(3)
[37]Wischmeier W H, Smith D D. Predicting Rainfall Erosion Losses from Cropland East of the Rocky Mountains. Agricultural Handbook No.282[M]. Washington D C:USDA,1965
[38]刘宝元,谢云,张科利.土壤侵蚀预报模型[M].北京：中国科学技术出版社,2001
[39]江忠善,郑粉莉.坡面水蚀预报模型研究[J].水土保持学报,2004,18(1)：66-69.
[40]Williams J R,Neitsch S L,Arnold J G. Soil and Water Assessment Tool User's Manual [M]. Texas Blackland Research Center:Texas Agricultural Experiment Station
[41]Morgan R P C. The European Soil Erosion Model (EUROPEM):a dynamic approach for predicting sediment transport from field and small catchment[J]. Earth surface and Landform,1998,23:527-544
[42]De Roo A P J. LISEM:a single event physical-based hydrologic and soil ersion model for drainage basins Ⅰ:Theory, input and output[J]. Processes,1996,10(8):1 119-1126
[43]江忠善.黄土地区天然降雨雨滴特征研究[J].中国水土保持,1983(3)：32-36.
[44]De Jong S M,Paracchini M L,Bertolo F,et al. Regional assessment of soil erosion using the distributed mode SEMMED and remotely sensed data[J]. Catena,1999,37:291-308
[45]史培军,刘宝元,张科利.土壤侵蚀过程与模型研究[J].资源科学,1999,21(5)：9-18
[46]卢光发.十堰市旅游地质资源及其保护[J].湖北地矿,2000,14(1)：40-44
[47]陈华,郭生练,柴晓玲.汉江丹江口以上流域降水特征及变化趋势分析[J].人民长江,2005,36(11)：29-31
[48]张利华,梁俊,蒋金龙,陈于.基于RS-GIS的湖北丹江库区土壤水力侵蚀定量分析[J].生态环境,2006,15(6)：1319-1323)
[49]赵光耀,赵兴华,王答相.陕西省丹江口水库水源区生态保护与重建存在的问题及对策[J].中国水土保持SWCC,2003,11 (2):11-12
[50]宋轩,杜丽平.基于GIS得淮河流域伏牛山区土壤侵蚀动态变化研究[J].中国水土保持,2008(2)
[51]NOEL D U, JAMES A L. The dynamics of soil erosion in USagriculture [J]. The Science of the Total Environment,1998,218:45-58
[52]杨存建,刘纪远,张增祥等.GIS支持下不同坡度的土壤侵蚀特征分析[J].水土保持学报,2002,16(6)：46-49
[53]赵文武,傅伯杰等.多尺度土壤侵蚀评价指数的技术与方法[J].地理科学进展,2007(2)
[54]陈三维,谢莉等.广东大宝山矿区土壤侵蚀强度判别方法[J].中国水土保持,2007(11)
[55]黄艳艳,曾志远,陈少辉.MODIS数据在土壤侵蚀动态监测中的应用[J].水土保持通报,2007,27(5)
[56]何丽君,董蕊,袁克杰.常见关联规则算法分析与比较.大连民族学院学报.2005.7.5：39-42
[57]李勉,李占斌,刘普灵.中国土壤侵蚀定量研究进展[J].水土保持研究,2002,9(3)：243-248
[58]张金英,葛京凤,于江海.基于RS和GIS的河北太行山土壤侵蚀强度研究[J].安徽农业科学.2006,34(24)：6391-6395
[59]姚水萍,任佶.浙江省土壤侵蚀等级划分模糊综合评判模型的初步探讨.水土保持通报,2006,26(6)：32-34
[60]汪东川,卢玉东.国外土壤侵蚀模型发展概述[J].中国水土保持科学,2004,2(2)：35-40
[61]龙栋材,李斌兵.基于支持向量机的小流域水蚀预报模型研究[J].中国水土保持科学,2006,4(6)：48-51
[62]邵峰晶,于忠清.数据挖掘原理与算法[J].中国水利水电出版社,2003.8
[63]刘淑珍,刘海军,钟祥浩等,西藏高原土壤侵蚀评价体系及监测方法,山地学报,2009.27(1)：76-81
[64]陈华,郭生练,柴晓玲等.汉江丹江口以上流域降水特征及变化趋势分析[J].人民长江,2005,36(11)：29-31
[65]张俊飚.中国土壤侵蚀影响因素及其危害分析[J].云南环境科学,2001,20(2)：4-7
[66]李雄飞,李军.数据挖掘与知识发现[M].北京：高等教育出版社,2003,5
[67][意]Paolo Giudici著.袁方,王煜,王丽娟等译.实用数据挖掘[M].北京：电子工业出版社,2004,6
[68]韩家炜等.Web挖掘研究.计算机研究与发展,2001,04
[69]水利部水土保持司.土壤侵蚀分类分级标准(SL 190-96).中华人民共和国水利部,1997
[70]USDA(United States Department of Agriculture)703手册, http://naldr. nal.usda. gov/Exe/ZyNET.exe?Client=National+Agricultural+Library+Digital+Repository&Init=0&ZyAc tion=ZyActionL%7C4&User=ANONYMOUS&Cleartext=&ZyActionlD=0&ImageQuality=r10 5g16%2Fr105g16%2Fx150y150g16%2Fi600&ZyEntry=0&Query=&Index=AH%7CAH2%7C AIB%7CBIC%7CBooks%7CCirc%7CERS%7CFVMNR%7CHGB%7CJAR%7CMP%7CNFC %7CPI%7CROS%7CRural%7CTB%7CTB1%7CTB2%7CUSDA Div Bulletin%7CWPC%7C YOA1%7CYOA2&File=&SortMethod=h%7C-&MaximumDocuments=10&FuzzyDegree=-1& MaximumPages=1&Time=&EndTime=&ConceptsFile=&SeekPage=&DefSeekPage
[71]杨艳生.区域性土壤流失预测方程得初步研究[J].土壤学报,1990,27(1)
[72]杨艳生,梁音,刘佳桂.长江三峡区土壤坡面流失及重力侵蚀[J].水土保持学报,1991
[73]刘宝元,谢云,张科利.土壤侵蚀预报模型[M].北京：中国科学技术出版社,2001
[74]李辉,陈晓玲.不同空间尺度下的土壤侵蚀模型研究进展[J].华中师范大学学报,2006,40(4)
[75]Z.H. Shi, Soil conservation planning at the small watershed level using RUSLE with GIS:a case study in the Three Gorge Area of China, CATENA,2003
[76]Zhi-Hua. Shi Research on the SCS-CN initial abstraction ratio using rainfall-runoff event analysis in the Three Gorges Area, China, CATENA,2008
[77]B. J. Fu, Assessment of Soil Erosion at Largewatershed Scale Using RUSLE and GIS:a Case Study in the Loess Plateau of China, land degradation & development Land Degrad. Develop.16:73-85 (2005)
[78]Wischmeier W. H. and Smith D.D., Rainfall energy and its relationship to soil loss. Transactions of the American Geophysical Union,1958,39:285-291
[79]蔡崇法,丁树文.应用USLE模型与地理信息系统IDRISI预测小流域土壤侵蚀量 得研究[J].水土保持学报,2000,14(2)
[80]王万忠,焦菊英.中国土壤侵蚀因子定量评价研究[J].水土保持通报,1996,16(5)
[81]章文波,谢云,刘元宝.利用日雨量计算降雨侵蚀力的方法研究.地理科学,2002,22(06)：705-711
[82]M.J.柯克比和R.P.C.摩根编著,王礼先等译,土壤侵蚀,北京：水利电力出版社,1987
[83]RomkensM.J.M., Prasad S.N. and Poesen J.W.A., Soil erodibility and properties. In: Transactions of theⅧCongress of the International Society of Soil Science, Hamburg,1986,492-504
[84]Declercq F. and Poesen J., Evaluation of twomodels to calculate the soil erodibility factor k. Pedologie,1992, XLII (2):158-169
[85]卜兆宏、李全英.土壤可蚀性(K)值图编制方法的初步研究[J].遥感技术与应用,1994,9(4)：22-27
[86]Wischmeier W.H. and Smith D.D., Predicting rainfall erosion losses-A guide to conservation planning with the Universal Soil Loss Equation (USLE). Agriculture Handbook No.537, United States Department of Agriculture, Springfield, USA. 1978
[87]罗志军,张俊.土壤侵蚀模型的研究现状与展望[J].安徽农业科学,2007,35(27)：8547-8576
[88]江忠善,郑粉莉,武敏.中国坡面水蚀预报模型研究[J].泥沙研究,2005(4)
[89]谢云,刘宝元,章文波.侵蚀性降雨标准研究[J].土壤侵蚀与水土保持学报[J].2000,14(4)：6-11
[90]王万忠,焦菊英.中国土壤侵蚀因子定量评价研究[J].水土保持通报,1999,16(5)
[91]白鹏、张喜斌,张斌,李彦,谢文俊,刘君华.支持向量机理论及工程应用实例[J].西安电子科技大学出版社,2008,8
[92]龙栋材,李斌兵.基于支持向量机的小流域水蚀预报模型研究[J].中国水土保
持科学,2006,4(6)
[93][英]Nello Cristianini, John Shawe-Taylor著,李国正,王猛,曾华军译,支持向量机器导论(an introduction to support vector machines and other kernel-based learning methods)电子工业出版社,北京,2004,3
[94]Vapnik V N. An overview of statistical learning theory[J].IEEE Trans on Neural Networks.1999,10(5):988-999
[95]杜熊禹.用于数据挖掘的支持向量机算法研究[D].电子科技大学,2007
[96]http://map.sogou.com/#city=%u6E56%u5317&lq=%u4E39%u6C5F%u5E93%u533 A&page=1&hb=1,1&c=12384500,3856500,8
[97]中华人民共和国行业标准SL277-2002.北京：中国水利水电出版社,2002：
[98]张科利,刘宝元,蔡永明.土壤侵蚀预报研究中得标准小区问题论证[J].地理研究,2000,19(3)
[99]黄煜煜,李勇志,刘予时.房县地区地形因子对水土流失的影响[J].湖北水利水电职业技术学院学报,2006
[100]张洪江.土壤侵蚀原理[M],北京：中国林业出版社,1999,10
[101]WischmeierW H, et al. A soil erodibility nomorgraph farm land and construction sites[J]. Journal of Soil andWater Conservation,1971,26:189-193
[102]杨子生.滇东北山箧坡耕地土壤流失方程研究[J].水土保持通报,1999,19(1)
[103]Jacobs R. Increased rates of convergence through learning rate adaptation. Neural Networks,1988,1:295-307
[104]Chauvin Y and Rumelhart D. E. Backpropagation:Theory,Architectures and Applications. Hillsdale, NJ:LawrenceErlbaum Associates,1995
[105]王春香,付云中,王永章,路华,基于BP神经网络的张力控制系统,中国机械工程,2001.12(8)：873-875
[106]D.C. Flanagan, J.C. Ascough Ⅱ, A.D. Nicks, M.A. Nearing and J.M. Laflen Overview of the WEPP erosion prediction model,1995
[107]王万忠,焦菊英.中国的土壤侵蚀因子定量评价研究[J].水土保持通报,1996,
16(5)：1-20
[108]李甚悟,李君莲.降雨及土壤湿度对水土流失的影响[J].土壤学报,1992,29(1)：94-103
[109]Y. k Bissonnais, B. Renaux, H. Delouche. Interactions between soil properties and moisture content in crust formation, runoff and interill erosion from tillde loess soils[J]. CATENA,1995,25:33-46
[110]Bernard B, Eric R. Aggregate stability as all indicator of soil susceptibility to runoff and erosion:validation at several levels [J]. Catena,2002,47:133-149
[111]Gyssels G, Poesen J. The importance of plant root characteristics in controlling concentrated flow erosion rates[J]. Earth Surface Processes and Landforms,2003, 28:371-384
[112]Hofzullah Aksoy, M. Levent Kawas. A review of hillslope and watershed scale erosion and sediment transport models [J]. Catena,2005,64:247-271
[113]Misra, R. K, Rose, C. W. Application and sensitivity analysis of process based erosion model-GUEST[J]. European Journal Soil Science,1996,10:593-604
[114]Casermeim M. A., Molina J. A., de la Cruz Caravaea M T. Influence of scrubs on runoff and sediment loss in soils of Medeterranean climate climate[J]. Catena, 2004,57:91-107
© 2004-2018 中国地质图书馆版权所有 京ICP备05064691号 京公网安备11010802017129号 地址：北京市海淀区学院路29号 邮编：100083 电话：办公室：(+86 10)66554848；文献借阅、咨询服务、科技查新：66554700 |