节点文献

模糊神经网络模型及其在水土资源中的应用

Fuzzy Neural Network and Its Applications in Water and Soil Resources

【作者】 杨先野

【导师】 付强;

【作者基本信息】 东北农业大学 , 农业水土工程, 2008, 硕士

【摘要】 水土资源领域中的时间序列总是受天文、地理、人为活动等众多因素的影响,因而存在高维、非线性、实测数据信息不完整等特征,但同时也具有多年变化的周期性和趋势性。常规的线性模型很难解决这些问题,因此,为解决水土资源中的复杂问题,本文建立了几种模糊神经网络模型,还通过MATLAB7.0软件进行编程,建立数学模型,并应用于实践中,结果都取得了良好的效果。本文取得了以下三方面的成果:1、构建了三江平原别拉洪河站多年降雨量预测和三江平原富锦市永富乡耕作土壤评价的模糊BP网络模型。(1)三江平原降雨量的多少不但直接影响当地水资源的开发和利用,而且与干旱、洪涝等自然灾害的发生、发展及其强度密切相关,因此,准确地预测降水量,既可为三江平原制定水资源开发利用策略提供科学依据,又可为防灾、抗灾、救灾提供有效的指导。(2)土壤质量的变化直接影响着粮食的增产。三江平原大面积开荒后,随着开垦年限的增长,土壤质量一股都呈现下降趋势,所以三江平原土壤的下降应当引起有关部门的高度重视。2、建立了三江平原创业农场的地下水预测的自适应模糊神经网络模型。三江平原地下水位的下降主要是因为以井灌水稻为主的农业仍将是该地区的主导产业,单纯追求粮食产量,盲目无序地开发利用地下水所造成的。如果仍按目前的发展模式,井灌水稻面积仍逐年加大,地下水资源开采量仍会增加,势必造成地下水供需动态平衡被进一步破坏,降雨及周围河水的侧向渗漏来不及补给,导致该地区地下水水位下降。因此,提醒该地区有关部门采取有效措施,如对机井进行合理布局、充分利用好地表水资源、采用新的节水灌溉措施等,以控制地下水资源的开采量,同时也应适当缩减井灌水稻的种植面积,退耕还林,使地下水动态平衡得以逐渐恢复,实现水资源的可持续利用。3、建立基于减法聚类的自适应模糊神经网络的三江平原富锦市农业水资源利用状况评价模型。三江平原农业水资源的供需矛盾日益加剧,从对三江平原富锦市2000年农业水资源供需状况看,富锦市水资源利用不合理,所以相关管理部门要大力采取节水措施,加强水资源管理,做到科学用水,计划用水,要根据本地的实际情况加强立法,制定切实有效的水法。从未来发展看,由于松花江水系水量的逐年减少,三江平原灌区的用水危机极大,必须狠抓节水和水资源管理尤,其是农作物布局结构的调整,使有限的水资源合理分配和利用以达到水资源永续利用的目的。基于以上几种的模型的建立,本论文实现了理论与实践的有机结合。同时本文将所建立的模型用于水土资源的多元复杂性、非线性研究,取得满意结果,为解决当前某些实际问题提供新技术,为水土资源系统研究提供了新思路。

【Abstract】 Time series of water and soil resources is always effected by the astronomy, the geography and the artificial activity,so there are characteristics such as high dimensional ,nonlinear and measured data with incomplete information and so on,but there is periodicity and trendy simultaneously of multi-annual variation. It is very difficult to solve these problems by the conventional linear model, therefore, to solve the complex problems in the water and soil resources, several kinds of fuzzy neural network model (FNN)is established in this paper and carried on the programming through the MATLAB7.0 software, established the mathematical model, and used in the practice, the results are good.The following results of three aspects are made in this paper:1.Fuzzy BP network model is established which is used in the multi-annual rainfall forecast of Bielahong River and cultivated soil appraisal of Yongfu town Fujin city in Sanjiang Plain.(1) The local water resources development and use is effected by the rainfall of Sanjiang Plain, and the drought and flood and other natural disaster’s occurrence, the development and the intensity is also closely related to the rainfall of Sanjiang Plain.So, accurate rainfall forecasts is the scientific basis of water resources development of Sanjiang Plain,and the effective instruction is provided for the disaster prevention, resist natural calamities and the disaster relief.(2) The grain production increase is effected by soil quality change. The soil quality is growing worse, after Sanjiang Plain was opened up with big area, therefore the deterioration soil quality of Sanjiang Plain must be taken seriously by the related government.2. Establishment the model of Adaptive Neuro-Fuzzy Inference System(ANFIS) of prediction groundwater in the Chuang Ye farm in San Jiang plain.The reason the level of groundwater in San Jiang plain is that well irrigation rice is the leading industry in this area, simply the pursuit of food production, blind and disorder in development and utilization of the groundwater. If the current mode of development will continue, the area of well irrigation rice will increase year by year, the exploitation of groundwater resources will continue to increase, thus the dynamic balance of groundwater supply and demand is bound to be further damaged, for the lateral leakage of rainfall and rivers around there was no time to recharge the groundwater, and cause the decline in the level of groundwater. Therefore, remind the relevant department to take effective measures, such as distribute wells in a rational way, make full use of surface water resources, and introduce the new water-saving irrigation measures and so on, to control the exploitation of groundwater resources, at the same time the plant area of well-irrigation rice should also be appropriately reduced, and returning farmland to forest, so the dynamic balance of groundwater can gradually restore, finally achieve sustainable use of water resources.3. Based on the Adaptive Neuro-Fuzzy Inference System, the agricultural utilization of water resources evaluation model in Sanjiang Plain Fujin city was established.The conflict between supply and demand of water resources in agriculture is growing in Sanjiang Plain. From agricultural water supply and demand situation in Sanjiang Plain Fujin city in 2000 to see, the use of water resources is irrational, so the relevant administrative departments should vigorously adopt water-saving measures to strengthen the management of water resources, to use water science and planned, according to the local conditions to strengthen legislation, speeding up the development of effective water law. From future development, because of the progressive decline of the Songhua River water, the Sanjiang Plain Irrigation District water crisis is great. We must pay close attention to water-saving and water resources management, in particular the restructuring of the distribution of crops to ensure that the limited rational allocation and use of water resources to meet the purpose of sustainable use of water resources. Based on the above several model, in the paper, the theory and practice are organically combined. At the same time, this will be the model for the multiple complex, nonlinear study on soil and water resources and obtain satisfactory results, in order to solve some practical problems currently provide new technologies for soil and water resources system provides a new idea.

  • 【分类号】O159;TV213.4
  • 【被引频次】13
  • 【下载频次】392
  • 攻读期成果
节点文献中: 

本文链接的文献网络图示:

本文的引文网络