【摘要】 以异常天气条件下的实测潮位过程为研究对象,根据其为非平稳时间序列的特点,以人工神经网络BP算法作为预测工具,对潮位时间序列中缺失的数据进行补遗建立:差分方法人工神经网络模型;同一海域邻近潮汐测站潮位过程对应预测的去平均值法人工神经网络模型;对增减水现象潮汐过程预测的气象资料数据库人工神经网络模型。以实测资料验证上述方法的可行性,并取得了很好的预测结果。更多还原

【Abstract】 Based on BP algorithm and the nonstationary characteristics of tidal field data under unusual weather,this paper presents three artificial neural network models: Difference artificial neural networks model for supplement forecasting;Minus-mean artificial neural networks model for corresponding forecasting between different tidal gauge stations;Weather database artificial neural networks model for setup and setdown.The numerical results show that these models perform well for tidal field data supplemental forecasting.更多还原