模糊均生函数残差模型在地下水数值模拟降水量预报中的应用

Application of residual error model of fuzzy average-generated function in precipitation forecasting of groundwater numerical simulation

地下水主要补给来源为大气降水的入渗和地表水体的渗漏。在地下水模拟预报模型中需要预先知道降水量。迄今为止,年降水量的预测仍然是一个不易解决的难题。在模糊均生函数模型(FAFM)基础上,利用其残差数据序列对FAFM进行校正,提出了模糊均生函数残差模型(REMFAF),给出了模型预报精度的检验方法。实例研究表明,REMFAF模型应用于吉林省西部地区地下水数值模拟中的降水量预报,比FAFM的预报精度更高,取得了较为理想的结果。

Groundwater recharge mainly comes from infiltration of precipitation and seepage of surface water. Precipitation must be known in advance in groundwater modeling. So far, it is difficult to forecast annual precipitation. On the basis of fuzzy average-generated function model (FAFM), its residual error data series has been used to calibrate FAFM, and residual error model of fuzzy average-generated function (REMFAF) has been presented and the method of model accuracy validation has also been given. A case study shows that REMFAF has been used to forecast precipitation in the western Jilin Province with good results and higher accuracy than FAFM.