摘要
利用人工神经网络和逐步回归方法对松花江汛期最高水位进行模拟 ,并对 1996、1997、1998年的最高水位进行了试报 ,效果极佳。试报 1998年松花江最高水位 119 6 9m ,实况 12 0 89m ,仅差 1 2m ;报出大洪水 ;而且在1999、2 0 0 0年业务预报中也准确报出松花江出现的枯水位。
Simulate the higher water level of the Songhua River with Artifical Neural Network and successive regression.Experimental forecast 1996?1997?1998year high water level,the are all very well.The Songhua River happened flood that was one beyond a century in 1998 year,experimental forecasting value is 119 69 meter,the fact is 120.89 meter,the balance is only 1.2 meter,so forecasted out the flood water level of 1998 year,and that accurately forecasted low water level of the Songhua River in 1999 and 2000 year forecasting operation.The both method may perfect use in acually working.
出处
《黑龙江气象》
2001年第3期9-11,15,共4页
Heilongjiang Meteorology
基金
黑龙江省科委下达"松花江洪水对经济环境影响评估及预防对策研究"课题 (1999~ 2 0 0 0年攻关项目 )