摘要
天气预报的困难主要源于非线性、高维性和数据的不精确性。所以，我们利用人工神经元网络来处理气象问题。本文引入BP网络来预报鲁西南地区的西南涡降水强度。经过网络的自学习，它能自动地从历史资料中抽取预报因子和预报对象的关系，进而作出较正规的预报。
The difficuities in forecasting weather result from non－linearity、high—dimension and data’s lack of preciseness．Consequently’we take advantage of Artificial NeuralNetwork to solve meteorological problems． Back－Propagation network is used to forecast the intensity of Southwest Vortex’s rainfall over the southwestern part of Shandong province in China．After learning by itself，network automatically abstracts the relationship between predctors and predictand from historical data．Finally，we areable to make a comparably ocrrect prediction by means of BP network．
出处
《气象科学》
CSCD
北大核心
1999年第2期158-165,共8页
Journal of the Meteorological Sciences
