节点文献

基于数据挖掘的水库供水调度规则提取

Water Supply Reservoir Operating Rules Extraction Based on Data Mining

  • 推荐 CAJ下载
  • PDF下载
  • 不支持迅雷等下载工具,请取消加速工具后下载。

【作者】 尹正杰王小林胡铁松吴运卿

【Author】 YIN Zheng-jie~1,WANG Xiao-lin~2,HU Tie-song~2,WU Yun-qing~2(1.Water Resource Department,Yangtse River Scientific Research Institute,Wuhan 430010,China;2.State Key Laboratory of Water Resources and Hydropower Engineering Science,Wuhan University,Wuhan 430072,China)

【机构】 长江科学院水资源研究所武汉大学水资源与水电工程科学国家重点实验室武汉大学水资源与水电工程科学国家重点实验室 湖北武汉430010湖北武汉430072

【摘要】 以供水调度为例对数据挖掘用于水库调度规则提取进行了研究.经分析主要选取了水库蓄水量、调度时段编号、需水量、径流量和水文年型5个特征属性构成数据集,通过数据挖掘从中发掘水库供水调度规则模式.采用径向基函数网络作为数据挖掘算法,将复杂的属性空间上的数据样本,映射为几种离散的供水调度模式,从而完成供水调度规则的模式划分.为了验证数据挖掘方法在调度规则提取上的效果,给出了调度图和调度函数方法用于供水调度的计算结果,三种方法的调度结果对比分析显示,数据挖掘方法在供水调度模式分类正确率和缺水指数两方面都是最好的,这反映出数据挖掘方法用于水库调度是合理有效的.

【Abstract】 This paper explores the application of data mining in reservoir operating rules extraction with a case of water supply operation.Five characteristic attributes of reservoir storage,within-year operating period number,water demand,reservoir inflow and hydrologic year type are selected to compose databases from which reservoir operating rule patterns are identified using data mining.Radial basic function(RBF) network is utilized as date mining algorithm to perform mapping of data samples from complex attribute space to several discrete water supply operating patterns,hence completing pattern division of water supply operating rules.To validate the effectiveness of data mining using for operating rules extraction,we also provide operation results of traditional operating chart and operating functions for comparison.And the result comparison analysis shows that data mining is the best method for reservoir operating rules extraction in terms of both correct operating pattern division ratio and water deficiency index,which demonstrates that application of data mining in reservoir operation is practicable and effective.

【基金】 国家重点基础发展计划(‘973’计划)(2003CB415202-5)
  • 【文献出处】 系统工程理论与实践 ,Systems Engineering-Theory & Practice , 编辑部邮箱 ,2006年08期
  • 【分类号】TV697.1
  • 【被引频次】71
  • 【下载频次】1040
节点文献中: 

本文链接的文献网络图示:

本文的引文网络