台风条件下海上风电场风速变化大、无明显周期性,这对海上风电场的风速预测造成了极大的困难。针对此问题,提出台风条件下海上风电场风速多步预测方法。首先,针对台风预报信息与风电场风速数据时间尺度不统一的问题,提出用嵌入层网络对...台风条件下海上风电场风速变化大、无明显周期性,这对海上风电场的风速预测造成了极大的困难。针对此问题,提出台风条件下海上风电场风速多步预测方法。首先,针对台风预报信息与风电场风速数据时间尺度不统一的问题,提出用嵌入层网络对台风预报信息进行动态插值。其次,基于Holland气压场模型和Batts梯度风模型构建融合物理信息的神经网络,将Holland模型和Batts模型中的经验参数替换成网络可学习的参数,并针对网络训练过程中可能出现的数值问题引入适当的近似方法。最后,对含时序模式注意力机制的长短期记忆网络(temporal pattern attention long short-term memory,TPA-LSTM)进行改进,嵌入融合物理信息的神经网络,利用近40年台风期间的数据进行训练和测试。结果表明,在引入较少参数的情况下,物理信息神经网络能减少TPA-LSTM网络的训练迭代次数以及提高预测精度,所提模型相比序列到序列(sequence to sequence,Seq2Seq)模型和TPA-LSTM网络具有更高的预测精度。展开更多
Altimetry data have been widely used in various fiehts of oceanography, including the extreme weather events such as tropical cyclones, typhoons, and hurricanes. The performance of JASON1 in Typhoon Shanshan is assess...Altimetry data have been widely used in various fiehts of oceanography, including the extreme weather events such as tropical cyclones, typhoons, and hurricanes. The performance of JASON1 in Typhoon Shanshan is assessed by examining the sensor geophysical data record and illustrates how the measured return waveform, significant wave height, and backscatter are all affected by various factors associated with the typhoon, with details by the rain are illustrated. The correction method to maintain accurate wave height and wind speed measurements in Typhoon Shanshan and the results are presented. Furthermore, the additional results of rain rate and typhoon eye diameter can be retrieved. Because of the lack of in-situ measurements of wind, wave, and rain rate at Typhoon Shanshan, results are compared with the forecasted typhoon data and a good agreement is found.展开更多
以欧洲中期天气预报中心ECMWF(European Centre for Medium Range Weather Forecasts)的ERA5风场数据为真实风速参考值,利用HY-2B卫星散射计L2A数据,使用反向传播神经网络方法对风速进行了反演,分别建立了中高风速、中低风速和全风速反...以欧洲中期天气预报中心ECMWF(European Centre for Medium Range Weather Forecasts)的ERA5风场数据为真实风速参考值,利用HY-2B卫星散射计L2A数据,使用反向传播神经网络方法对风速进行了反演,分别建立了中高风速、中低风速和全风速反演模型。与基于NSCAT-4地球物理模式函数得到的L2B风速相比,在训练集中,3种网络模型反演风速的均方误差(Mean Square Error,MSE)分别达到了0.18,0.14和0.32 m/s,平均绝对值误差(Mean Absolute Error,MAE)分别达到了0.27,0.24和0.34 m/s;在测试集中,3种网络模型反演风速的均方误差(MSE)分别达到了0.54,0.27和0.46 m/s,平均绝对值误差(MAE)分别达到了0.48,0.35和0.42 m/s。研究结果表明,中高和中低风速模型优于全风速模型,其中中低风速模型反演风速的MSE和MAE最低,中高风速模型反演风速的MSE和MAE下降幅度最大;3种模型都具有良好的泛化能力。展开更多
文摘台风条件下海上风电场风速变化大、无明显周期性,这对海上风电场的风速预测造成了极大的困难。针对此问题,提出台风条件下海上风电场风速多步预测方法。首先,针对台风预报信息与风电场风速数据时间尺度不统一的问题,提出用嵌入层网络对台风预报信息进行动态插值。其次,基于Holland气压场模型和Batts梯度风模型构建融合物理信息的神经网络,将Holland模型和Batts模型中的经验参数替换成网络可学习的参数,并针对网络训练过程中可能出现的数值问题引入适当的近似方法。最后,对含时序模式注意力机制的长短期记忆网络(temporal pattern attention long short-term memory,TPA-LSTM)进行改进,嵌入融合物理信息的神经网络,利用近40年台风期间的数据进行训练和测试。结果表明,在引入较少参数的情况下,物理信息神经网络能减少TPA-LSTM网络的训练迭代次数以及提高预测精度,所提模型相比序列到序列(sequence to sequence,Seq2Seq)模型和TPA-LSTM网络具有更高的预测精度。
基金The National Natural Science Foundation of China under contract No. 30671619
文摘Altimetry data have been widely used in various fiehts of oceanography, including the extreme weather events such as tropical cyclones, typhoons, and hurricanes. The performance of JASON1 in Typhoon Shanshan is assessed by examining the sensor geophysical data record and illustrates how the measured return waveform, significant wave height, and backscatter are all affected by various factors associated with the typhoon, with details by the rain are illustrated. The correction method to maintain accurate wave height and wind speed measurements in Typhoon Shanshan and the results are presented. Furthermore, the additional results of rain rate and typhoon eye diameter can be retrieved. Because of the lack of in-situ measurements of wind, wave, and rain rate at Typhoon Shanshan, results are compared with the forecasted typhoon data and a good agreement is found.
文摘以欧洲中期天气预报中心ECMWF(European Centre for Medium Range Weather Forecasts)的ERA5风场数据为真实风速参考值,利用HY-2B卫星散射计L2A数据,使用反向传播神经网络方法对风速进行了反演,分别建立了中高风速、中低风速和全风速反演模型。与基于NSCAT-4地球物理模式函数得到的L2B风速相比,在训练集中,3种网络模型反演风速的均方误差(Mean Square Error,MSE)分别达到了0.18,0.14和0.32 m/s,平均绝对值误差(Mean Absolute Error,MAE)分别达到了0.27,0.24和0.34 m/s;在测试集中,3种网络模型反演风速的均方误差(MSE)分别达到了0.54,0.27和0.46 m/s,平均绝对值误差(MAE)分别达到了0.48,0.35和0.42 m/s。研究结果表明,中高和中低风速模型优于全风速模型,其中中低风速模型反演风速的MSE和MAE最低,中高风速模型反演风速的MSE和MAE下降幅度最大;3种模型都具有良好的泛化能力。
