摘要
将BP人工神经网络引入到风暴潮数值预报的解释应用中,并以惠州站为例,根据台风参数与增水的关系建立3套神经网络模型,对风暴潮的数值预报结果进行订正,计算结果显示:BP人工神经网络可以改进风暴潮数值模式的预报精度,可以作为惠州站数值预报结果解释应用的一种方法,同时也为台风风暴潮数值预测的解释应用提供了新思路。
In order to apply the BP artificial neural networkd（ANN） method in the forecasting system of storm surge model, three kinds of BP ANN models were established based on the relationship between typhoon parameters and increasing water. Taken Huizhou station as an example, the forecasting results from storm surge model were rectified. The results show that the BP ANN method could improve the forecasting accuracy from storm surge model, and the modeled results have a good agreement with the observed data. Hence, The BP ANN can be a useful measure method to reduce the bias of the simulation.
出处
《海洋预报》
2016年第2期60-65,共6页
Marine Forecasts
基金
国家海洋局海洋公益性行业科研专项(201305031-4)
关键词
BP人工神经网络
风暴潮
解释应用
数值预报
the BP artificial neural network
storm surge
interpretation
numerical prediction
