人工神经网络技术在台风浪预报中的应用

APPLICATION OF THE NEURAL NETWORK IN OCEAN WAVE FORECAST OF TYPHOON

利用人工神经网络中的BP算法 ,结合南海硇洲岛海区近30年的台风及台风浪资料 ,经预报因子的选择并作对比试验 ,建立了本海区较为理想的台风浪人工神经网络预报模型。结果表明:人工神经网络方法在台风浪的预报上 ,有较好的拟合历史台风浪高的能力 ,利用该模型对台风浪高的预报也达到了一定的精度。为实际台风海浪的预报增加了新方法、新思路。

Use BP net of the neural network to establish the quantitative sea wave forecast model of typhoon in the sea area of Zhan Jiang harbor.The result indicated:The neural network forecast model can simulate the historical wave high and use the wind etc. as a forecast factors to forecast the wave high of typhoon.As the BP net easily product oscillation ,we select different kinds of wave forecast factors to test the results and we get four factors which can ideally forecast the real sea wave.