摘要
将人工神经网络方法与统计方法相结合 ,采用分步预报的方法 ,首先用经验统计方法进行晴雨预测 ,再用 BP人工神经网络建立降水量级预报模型 ,探讨神经网络方法在中期降水预报中的应用。试验表明 ,神经网络方法能提高中到大雨降水量级的中期预报能力 ,但对暴雨的预报不理想。
The application of artificial neural network to rainfall mid range forecasting was discussed by combining it with statistics means. The statistics means was used to forecast rain or not. Then ANN was used to develop rainfall forecast model. Studies suggest that ANN is an effectual way to improve rainfall forecast, but it was unfavorable for heavy rain forecast.
出处
《气象》
CSCD
北大核心
2001年第6期40-42,46,共4页
Meteorological Monthly
