MFAM模型在河流水质污染模拟及预测中的应用

文中以时间序列分析为基础，介绍了均值生成函数这一崭新概念，并且经成份因子提取分析推导建立了模拟序列的数字模型（简记为MFAM），经对黄河下游花园口断面的1988-1989年实测水质污染指标溶解氧（DO），氨氧，化学耗氧量（COD），五日生化需氧量（BOD5）等序列模拟，结果表明MFAM模型能较好地模拟河流水质污染指标的变化趋势，拟合平均误差只有5.2-6.4%，MFAM模型应用于预测1990-1991年水质污染指标变化，结果表明预测精度达85%以上，文中最后得出结论：MFAM模型应用于河流污染模拟和预测，是完全可行且十分方便。

The Application of MFAM Model to River Water Quality Pollution Predication

Based on time series analysis,a new concept-average values generation function is introdued and a mathematic model(remarked as MFAM)is represened through tollecting major factors involed in time series The results from the application ofMFAM to the simlation of water qualitu pollution indexes gotten from 1985 to 1991 at Huayuankou cross section in the loweer reach ofYellow River show that MFAM could be ised to predict the trcnds of river water quality pollution index. Here. water quality indexes in-clude DO,ammonia nitrogen. COD. BOD_5. The mean simulated relative error and predicted precision are 5.8 % and over 85% respec-tively. Thus MFAM could be used to predic the trends of river water quality pollution indexes conveniently.