A novel pure data-selection framework for day-ahead wind power forecasting

Numerical weather prediction(NWP)data possess internal inaccuracies,such as low NWP wind speed corresponding to high actual wind power generation.This study is intended to reduce the negative effects of such inaccuracies by proposing a pure data-selection framework(PDF)to choose useful data prior to modeling,thus improving the accuracy of day-ahead wind power forecasting.Briefly,we convert an entire NWP training dataset into many small subsets and then select the best subset combination via a validation set to build a forecasting model.Although a small subset can increase selection flexibility,it can also produce billions of subset combinations,resulting in computational issues.To address this problem,we incorporated metamodeling and optimization steps into PDF.We then proposed a design and analysis of the computer experiments-based metamodeling algorithm and heuristic-exhaustive search optimization algorithm,respectively.Experimental results demonstrate that(1)it is necessary to select data before constructing a forecasting model;(2)using a smaller subset will likely increase selection flexibility,leading to a more accurate forecasting model;(3)PDF can generate a better training dataset than similarity-based data selection methods(e.g.,K-means and support vector classification);and(4)choosing data before building a forecasting model produces a more accurate forecasting model compared with using a machine learning method to construct a model directly.

Projection of temperature and precipitation under SSPs-RCPs Scenarios over northwest China

Climate change significantly affects the environmental and socioeconomic conditions in northwest China.Here we evaluate the ability of five general circulation models(GCMs)from 6th phase of the Coupled Model Inter-comparison Project(CMIP6)to reproduce regional temperature and precipitation over northwest China from 1961 to 2014,and project the future temperature and precipitation during 2021 to 2100 under SSPs-RCPs(SSP1-1.9,SSP1-2.6,SSP2-4.5,SSP3-7.0,SSP4-3.4,SSP4-6.0 and SSP5-8.5).The results show that the CMIP6 models can simulate temperature better than precipitation.Projections show that the annual mean temperature will further increase under different SSPs-RCPs scenarios in the 21st century.Future climate changes in the near-term(2021-2040),mid-term(2041-2060)and long-term(2081-2100)are analyzed relative to the reference period(1995-2014).In the long term,warming will be significantly higher than the near and mid-terms.In the long term,annual mean temperature will increase by 1.4℃,1.9℃,3.3℃,5.5℃,2.7℃,3.8℃ and 6.0℃ under SSP1-1.9,SSP1-2.6,SSP2-4.5,SSP3-7.0,SSP4-3.4,SSP4-6.0 and SSP5-8.5,respectively.Spatially,warming in the Junggar Basin will be higher than those in the Tarim Basin.Seasonally,the maximum warming zone will be in the mountainous areas of Tarim Basin during spring and autumn,in the southern basin during winter,and in the east during summer.Precipitation shows an increasing trend under different SSPs-RCPs in the 21st century.In the long term,increase in precipitation will be significantly higher than in the near and mid-terms.Increase in annual precipitation in the long term will be 4.1% under SSP1-1.9,13.9% under SSP1-2.6,28.4% under SSP2-4.5, 35.2% under SSP3-7.0, 6.9% under SSP4-3.4, 8.9% under SSP4-6.0, and 27.3% under SSP5-8.5 relative to the reference period of 1995-2014. Spatially, precipitation increase will be higher in the south than the north, especially higher in mountainous regions than the basin under SSP2-4.5, SSP3-7.0, and SSP5-8.5. Seasonally, highest increase can be expected for winter, followed by spring, with significant increase in mountainous regions of southern Tarim Basin. Summer precipitation will reduce in Tian Shan and basins but will significantly increase in the northern margin of the Kunlun Mountain.

Two‐stage short‐term wind power forecasting algorithm using different feature-learning models

Two-stage ensemble-based forecasting methods have been studied extensively in the wind power forecasting field. However, deep learning-based wind power forecasting studies have not investigated two aspects. In the first stage, different learning structures considering multiple inputs and multiple outputs have not been discussed. In the second stage, the model extrapolation issue has not been investigated. Therefore, we develop four deep neural networks for the first stage to learn data features considering the input-and-output structure. We then explore the model extrapolation issue in the second stage using different modeling methods. Considering the overfitting issue, we propose a new moving window-based algorithm using a validation set in the first stage to update the training data in both stages with two different moving window processes. Experiments were conducted at three wind farms, and the results demonstrate that the model with single-input–multiple-output structure obtains better forecasting accuracy compared to existing models. In addition, the ridge regression method results in a better ensemble model that can further improve forecasting accuracy compared to existing machine learning methods. Finally, the proposed two-stage forecasting algorithm can generate more accurate and stable results than existing algorithms.