权益分类	功能权益	普通用户	{{item.name}}会员
{{category.name}}	{{benefitItem.name}}

Estimation and Management of Drought Situation by Using Pattern Classification and furry Infererc

利用模式分类和毛推断的干旱形势估计和管理

基本信息

批准号：
03832023
负责人：
KOGIRI Toshiharu
金额：
$ 1.09万
依托单位：
Gifu University
依托单位国家：
日本
项目类别：
Grant-in-Aid for General Scientific Research (C)
财政年份：
1991
资助国家：
日本
起止时间：
1991 至 1992
项目状态：
已结题

项目摘要

We have been suffering from heavy drought for these years, which was caused by difference between patterns of discharge hydrograph and water demand. Basically hydrographs are gained through the run-off analysis applied to procipitation, and geological and geographical conditions. Though precipitations are occurred with the complex components, hyetographs and hydrographs consist of typically few patterns according to their characteristics. Therefore in this project, We have investigated the estimation methods of drought severity and real-time operation for drought control. To sum up, the following results were gained:i)ISODATA method classified historical hydrographs into the reasonable patterns with the proposed estimation function from the view point of low flow.ii)long-term run-off model with snow-melt was formulated by using mesh-typed multi-layer process. Discharge was simulated through fuzzy generation procedures under the conditions of assumed scenarios. The snow cover area was identified with discrimination function of satellite data of NOAA.iii)Probability of drought event for classified hydrograph was estimated through the concept of success percent. Moreover, the drought characteristics was formulated as reliability and resiliency.iv)In the real-time operation of reservoir for drought, the operational procedure was proposed by comparing the expert system and following the experience of operator.v)Target storage volume at each control time stage was calculated for each classified hydrograph under the condition of special operation rule.vi)Fuzzy inference theory gave the actual target storage volume in other words, the release at next control time stage with similarities of present drought against classified hydrographs.

多年来，我国遭遇了严重的干旱，这是由于径流过程线与需水量的差异造成的。水文过程线基本上是通过径流分析应用于预测，地质和地理条件。虽然降水的组成成分复杂，但雨线和过程线的特征决定了它们通常只包含少数几种模式。因此，本计画主要探讨干旱严重度的估测方法与实时抗旱作业。研究结果表明：（1）ISODATA方法从枯水流量的角度出发，利用所提出的估算函数对历史过程进行了合理的分型;（2）采用网格式多层处理，建立了考虑融雪的长期径流模型。在假设情景条件下，通过模糊生成程序模拟流量。利用NOAA卫星资料的判别函数识别积雪覆盖面积。3）利用成功率的概念估计了分类过程线的干旱事件概率。在水库实时调度中，通过与专家系统的比较，并根据调度员的经验，提出了水库实时调度的操作程序。（5）在特殊调度条件下，计算了各分类过程线在各控制时间段的目标库容rule.vi。下一个控制时间阶段的释放与当前干旱的相似性对分类过程线。