基于信息扩散近似推理的年径流量预测模型  

Annual Runoff Prediction Model Based on Information-Diffusion Approximate Reasoning

作　　者：陈海涛[1] 冯刚 段春青[1] 邱林[1] 

机构地区：[1]华北水利水电学院水利学院,河南郑州450011 [2]河北省邢台金成建筑安装有限公司,河北邢台054000

出　　处：《节水灌溉》2008年第9期29-31,共3页Water Saving Irrigation

基　　金：河南省高校创新人才培养工程项目(HNCX2003-17)

摘　　要：通过分析年径流时间序列的特性,利用信息扩散近似推理描述年径流量间的复杂非线性关系,建立起基于信息扩散近似推理的年径流预测模型。信息扩散近似推理将样本点转换成模糊集,部分弥补了由于数据的不完备性所造成的信息空白,并可以将矛盾模式转换成兼容模式。通过与传统预测方法相比较,发现该模型能够很好地光滑样本数据以及能够较好地发掘知识,有较高的预测精度和推广应用价值。Through analysis of annual runoff time-series characteristic, the information-diffusion approximate reasoning method is used to describe the complex nonlinear relationship among annual runoff, and the annual runoff prediction model based on information-diffusion approximate reasoning is established in this paper. The information-diffusion approximate reasoning method converts the sample poims to fuzzy set, partly compensates the information vacancy caused by incomplete data, and can convert incompatible mode to compatible mode. Compared with traditional prediction method, the information-diffusion approximate reasoning model can well smooth the difference among the sample data and mine uncertain knowledge. It also has good prediction precision and promotion value.

关 键 词：信息扩散近似推理 预测 年径流 

分 类 号：P333.1[天文地球—水文科学]