This study investigates the long-term changes of monthly sea surface wind speeds over the China seas from 1988 to 2015. The 10-meter wind speeds products from four major global reanalysis datasets with high resolution...This study investigates the long-term changes of monthly sea surface wind speeds over the China seas from 1988 to 2015. The 10-meter wind speeds products from four major global reanalysis datasets with high resolution are used: Cross-Calibrated Multi-Platform data set(CCMP), NCEP climate forecast system reanalysis data set(CFSR),ERA-interim reanalysis data set(ERA-int) and Japanese 55-year reanalysis data set(JRA55). The monthly sea surface wind speeds of four major reanalysis data sets have been investigated through comparisons with the longterm and homogeneous observation wind speeds data recorded at ten stations. The results reveal that(1) the wind speeds bias of CCMP, CFSR, ERA-int and JRA55 are 0.91 m/s, 1.22 m/s, 0.62 m/s and 0.22 m/s, respectively.The wind speeds RMSE of CCMP, CFSR, ERA-int and JRA55 are 1.38 m/s, 1.59 m/s, 1.01 m/s and 0.96 m/s,respectively;(2) JRA55 and ERA-int provides a realistic representation of monthly wind speeds, while CCMP and CFSR tend to overestimate observed wind speeds. And all the four data sets tend to underestimate observed wind speeds in Bohai Sea and Yellow Sea;(3) Comparing the annual wind speeds trends between observation and the four data sets at ten stations for 1988-1997, 1988–2007 and 1988–2015, the result show that ERA-int is superior to represent homogeneity monthly wind speeds over the China seaes.展开更多
Wind and wave data are essential in climatological and engineering design applications.In this study,data from 15 buoys located throughout the South China Sea(SCS)were used to evaluate the ERA5 wind and wave data.Appl...Wind and wave data are essential in climatological and engineering design applications.In this study,data from 15 buoys located throughout the South China Sea(SCS)were used to evaluate the ERA5 wind and wave data.Applicability assessment are beneficial for gaining insight into the reliability of the ERA5 data in the SCS.The bias range between the ERA5 and observed wind-speed data was-0.78-0.99 m/s.The result indicates that,while the ERA5 wind-speed data underestimation was dominate,the overestimation of such data existed as well.Additionally,the ERA5 data underestimated annual maximum wind-speed by up to 38%,with a correlation coefficient>0.87.The bias between the ERA5 and observed significant wave height(SWH)data varied from-0.24 to 0.28 m.And the ERA5 data showed positive SWH bias,which implied a general underestimation at all locations,except those in the Beibu Gulf and centralwestern SCS,where overestimation was observed.Under extreme conditions,annual maximum SWH in the ERA5 data was underestimated by up to 30%.The correlation coefficients between the ERA5 and observed SWH data at all locations were greater than 0.92,except in the central-western SCS(0.84).The bias between the ERA5 and observed mean wave period(MWP)data varied from-0.74 to 0.57 s.The ERA5 data showed negative MWP biases implying a general overestimation at all locations,except for B1(the Beibu Gulf)and B7(the northeastern SCS),where underestimation was observed.The correlation coefficient between the ERA5 and observed MWP data in the Beibu Gulf was the smallest(0.56),and those of other locations fluctuated within a narrow range from 0.82 to 0.90.The intercomparison indicates that during the analyzed time-span,the ERA5 data generally underestimated wind-speed and SWH,but overestimated MWP.Under non-extreme conditions,the ERA5 wind-speed and SWH data can be used with confidence in most regions of the SCS,except in the central-western SCS.展开更多
We propose a novel machine learning approach to reconstruct meshless surface wind speed fields,i.e.,to reconstruct the surface wind speed at any location,based on meteorological background fields and geographical info...We propose a novel machine learning approach to reconstruct meshless surface wind speed fields,i.e.,to reconstruct the surface wind speed at any location,based on meteorological background fields and geographical information.The random forest method is selected to develop the machine learning data reconstruction model(MLDRM-RF)for wind speeds over Beijing from 2015-19.We use temporal,geospatial attribute and meteorological background field features as inputs.The wind speed field can be reconstructed at any station in the region not used in the training process to cross-validate model performance.The evaluation considers the spatial distribution of and seasonal variations in the root mean squared error(RMSE)of the reconstructed wind speed field across Beijing.The average RMSE is 1.09 m s^(−1),considerably smaller than the result(1.29 m s^(−1))obtained with inverse distance weighting(IDW)interpolation.Finally,we extract the important feature permutations by the method of mean decrease in impurity(MDI)and discuss the reasonableness of the model prediction results.MLDRM-RF is a reasonable approach with excellent potential for the improved reconstruction of historical surface wind speed fields with arbitrary grid resolutions.Such a model is needed in many wind applications,such as wind energy and aviation safety assessments.展开更多
Amid the randomness and volatility of wind speed, an improved VMD-BP-CNN-LSTM model for short-term wind speed prediction was proposed to assist in power system planning and operation in this paper. Firstly, the wind s...Amid the randomness and volatility of wind speed, an improved VMD-BP-CNN-LSTM model for short-term wind speed prediction was proposed to assist in power system planning and operation in this paper. Firstly, the wind speed time series data was processed using Variational Mode Decomposition (VMD) to obtain multiple frequency components. Then, each individual frequency component was channeled into a combined prediction framework consisting of BP neural network (BPNN), Convolutional Neural Network (CNN) and Long Short-Term Memory Network (LSTM) after the execution of differential and normalization operations. Thereafter, the predictive outputs for each component underwent integration through a fully-connected neural architecture for data fusion processing, resulting in the final prediction. The VMD decomposition technique was introduced in a generalized CNN-LSTM prediction model;a BPNN model was utilized to predict high-frequency components obtained from VMD, and incorporated a fully connected neural network for data fusion of individual component predictions. Experimental results demonstrated that the proposed improved VMD-BP-CNN-LSTM model outperformed other combined prediction models in terms of prediction accuracy, providing a solid foundation for optimizing the safe operation of wind farms.展开更多
High availability of wind power data is the basis for wind power research, but there are a large number of abnormal data in actual collected data, which seriously affects analysis of wind power law and reduces predict...High availability of wind power data is the basis for wind power research, but there are a large number of abnormal data in actual collected data, which seriously affects analysis of wind power law and reduces prediction accuracy. Measured power data of wind farm are analyzed, influence of wind speed fluctuation characteristics on wind power is discussed, and abnormal points are identified for data of different wind types. The Cluster-Based Local Outlier Factor (CLOF) algorithm based on K-means is used to identify outlier abnormal points, and conditional constraints based on physical background are used to identify accumulation abnormal points. Reconstructed data segment is divided according to fluctuation of wind speed. The Bidirectional Gate Recurrent Unit (BiGRU) model with wind speed as input reconstructs fluctuation segment data, and bi-directional weighted random forest model reconstructs stationary segment data. Based on analysis of measured data of a wind farm, results show the method can effectively identify various abnormal data, and complete high-quality reconstruction of data, thereby improving accuracy of wind power prediction.展开更多
基金The National Key R&D Program of China under contract No.2016YFC1401905the National Natural Science Foundation of China under contract No.41776004the Fundamental Research Funds for the Central Universities under contract No.2016B12514
文摘This study investigates the long-term changes of monthly sea surface wind speeds over the China seas from 1988 to 2015. The 10-meter wind speeds products from four major global reanalysis datasets with high resolution are used: Cross-Calibrated Multi-Platform data set(CCMP), NCEP climate forecast system reanalysis data set(CFSR),ERA-interim reanalysis data set(ERA-int) and Japanese 55-year reanalysis data set(JRA55). The monthly sea surface wind speeds of four major reanalysis data sets have been investigated through comparisons with the longterm and homogeneous observation wind speeds data recorded at ten stations. The results reveal that(1) the wind speeds bias of CCMP, CFSR, ERA-int and JRA55 are 0.91 m/s, 1.22 m/s, 0.62 m/s and 0.22 m/s, respectively.The wind speeds RMSE of CCMP, CFSR, ERA-int and JRA55 are 1.38 m/s, 1.59 m/s, 1.01 m/s and 0.96 m/s,respectively;(2) JRA55 and ERA-int provides a realistic representation of monthly wind speeds, while CCMP and CFSR tend to overestimate observed wind speeds. And all the four data sets tend to underestimate observed wind speeds in Bohai Sea and Yellow Sea;(3) Comparing the annual wind speeds trends between observation and the four data sets at ten stations for 1988-1997, 1988–2007 and 1988–2015, the result show that ERA-int is superior to represent homogeneity monthly wind speeds over the China seaes.
基金Supported by the Southern Marine Science and Engineering Guangdong Laboratory(Zhuhai)(No.SML2021SP102)the Key Laboratory of Marine Environmental Survey Technology and Application+2 种基金Ministry of Natural Resources(Nos.MESTA-2020-C003,MESTA-2020-C004)the Key Research and Development Project of Guangdong Province(No.2020B1111020003)the Science and Technology Research Project of Jiangxi Provincial Department of Education(No.GJJ200330)。
文摘Wind and wave data are essential in climatological and engineering design applications.In this study,data from 15 buoys located throughout the South China Sea(SCS)were used to evaluate the ERA5 wind and wave data.Applicability assessment are beneficial for gaining insight into the reliability of the ERA5 data in the SCS.The bias range between the ERA5 and observed wind-speed data was-0.78-0.99 m/s.The result indicates that,while the ERA5 wind-speed data underestimation was dominate,the overestimation of such data existed as well.Additionally,the ERA5 data underestimated annual maximum wind-speed by up to 38%,with a correlation coefficient>0.87.The bias between the ERA5 and observed significant wave height(SWH)data varied from-0.24 to 0.28 m.And the ERA5 data showed positive SWH bias,which implied a general underestimation at all locations,except those in the Beibu Gulf and centralwestern SCS,where overestimation was observed.Under extreme conditions,annual maximum SWH in the ERA5 data was underestimated by up to 30%.The correlation coefficients between the ERA5 and observed SWH data at all locations were greater than 0.92,except in the central-western SCS(0.84).The bias between the ERA5 and observed mean wave period(MWP)data varied from-0.74 to 0.57 s.The ERA5 data showed negative MWP biases implying a general overestimation at all locations,except for B1(the Beibu Gulf)and B7(the northeastern SCS),where underestimation was observed.The correlation coefficient between the ERA5 and observed MWP data in the Beibu Gulf was the smallest(0.56),and those of other locations fluctuated within a narrow range from 0.82 to 0.90.The intercomparison indicates that during the analyzed time-span,the ERA5 data generally underestimated wind-speed and SWH,but overestimated MWP.Under non-extreme conditions,the ERA5 wind-speed and SWH data can be used with confidence in most regions of the SCS,except in the central-western SCS.
基金supported by the Strategic Priority Research Program of the Chinese Academy of Sciences(Grant No.XDA19030402)the Key Special Projects for International Cooperation in Science and Technology Innovation between Governments(Grant No.2017YFE0133600the Beijing Municipal Natural Science Foundation Youth Project 8214066:Application Research of Beijing Road Visibility Prediction Based on Machine Learning Methods.
文摘We propose a novel machine learning approach to reconstruct meshless surface wind speed fields,i.e.,to reconstruct the surface wind speed at any location,based on meteorological background fields and geographical information.The random forest method is selected to develop the machine learning data reconstruction model(MLDRM-RF)for wind speeds over Beijing from 2015-19.We use temporal,geospatial attribute and meteorological background field features as inputs.The wind speed field can be reconstructed at any station in the region not used in the training process to cross-validate model performance.The evaluation considers the spatial distribution of and seasonal variations in the root mean squared error(RMSE)of the reconstructed wind speed field across Beijing.The average RMSE is 1.09 m s^(−1),considerably smaller than the result(1.29 m s^(−1))obtained with inverse distance weighting(IDW)interpolation.Finally,we extract the important feature permutations by the method of mean decrease in impurity(MDI)and discuss the reasonableness of the model prediction results.MLDRM-RF is a reasonable approach with excellent potential for the improved reconstruction of historical surface wind speed fields with arbitrary grid resolutions.Such a model is needed in many wind applications,such as wind energy and aviation safety assessments.
文摘Amid the randomness and volatility of wind speed, an improved VMD-BP-CNN-LSTM model for short-term wind speed prediction was proposed to assist in power system planning and operation in this paper. Firstly, the wind speed time series data was processed using Variational Mode Decomposition (VMD) to obtain multiple frequency components. Then, each individual frequency component was channeled into a combined prediction framework consisting of BP neural network (BPNN), Convolutional Neural Network (CNN) and Long Short-Term Memory Network (LSTM) after the execution of differential and normalization operations. Thereafter, the predictive outputs for each component underwent integration through a fully-connected neural architecture for data fusion processing, resulting in the final prediction. The VMD decomposition technique was introduced in a generalized CNN-LSTM prediction model;a BPNN model was utilized to predict high-frequency components obtained from VMD, and incorporated a fully connected neural network for data fusion of individual component predictions. Experimental results demonstrated that the proposed improved VMD-BP-CNN-LSTM model outperformed other combined prediction models in terms of prediction accuracy, providing a solid foundation for optimizing the safe operation of wind farms.
基金supported by the National Key R&D Program of China“Technology and application of wind power/photovoltaic power prediction for promoting renewable energy consumption”(2018YFB0904200).
文摘High availability of wind power data is the basis for wind power research, but there are a large number of abnormal data in actual collected data, which seriously affects analysis of wind power law and reduces prediction accuracy. Measured power data of wind farm are analyzed, influence of wind speed fluctuation characteristics on wind power is discussed, and abnormal points are identified for data of different wind types. The Cluster-Based Local Outlier Factor (CLOF) algorithm based on K-means is used to identify outlier abnormal points, and conditional constraints based on physical background are used to identify accumulation abnormal points. Reconstructed data segment is divided according to fluctuation of wind speed. The Bidirectional Gate Recurrent Unit (BiGRU) model with wind speed as input reconstructs fluctuation segment data, and bi-directional weighted random forest model reconstructs stationary segment data. Based on analysis of measured data of a wind farm, results show the method can effectively identify various abnormal data, and complete high-quality reconstruction of data, thereby improving accuracy of wind power prediction.
