Short-term load probabilistic forecasting based on quantile regression convolutional neural network and Epanechnikov kernel density estimation
Electricity load forecasting plays an indispensable role in the electric power systems. However, its characteristics of uncertainty and complexity are hard to handle. This paper proposes a probabilistic load forecasting approach named QRCNN-E. Specifically, the deep convolutional neural network is a...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2020-12-01
|
| Series: | Energy Reports |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S2352484720314062 |
| id |
doaj-e4e17a1ed0e7429a801a4d923778ab29 |
|---|---|
| record_format |
Article |
| spelling |
doaj-e4e17a1ed0e7429a801a4d923778ab292020-12-23T05:01:52ZengElsevierEnergy Reports2352-48472020-12-01615501556Short-term load probabilistic forecasting based on quantile regression convolutional neural network and Epanechnikov kernel density estimationHui He0Junting Pan1Nanyan Lu2Bo Chen3Runhai Jiao4School of Control and Computer Engineering, North China Electric Power University, Beijing 102206, China; Corresponding author.Department of Computer Science, University of Manchester, Manchester M13 9PY, United KingdomSchool of Control and Computer Engineering, North China Electric Power University, Beijing 102206, ChinaChina Unicom Big Data Co., Ltd, Beijing 100011, ChinaSchool of Control and Computer Engineering, North China Electric Power University, Beijing 102206, ChinaElectricity load forecasting plays an indispensable role in the electric power systems. However, its characteristics of uncertainty and complexity are hard to handle. This paper proposes a probabilistic load forecasting approach named QRCNN-E. Specifically, the deep convolutional neural network is applied to model the non-linear relationship with the electricity load and its influencing factors. By replacing the traditional loss function with pinball loss, the network can eventually forecast loads in quantiles. Then, kernel density estimation takes quantile forecasts as inputs and produces deterministic and probabilistic results. Case studies on GEFCom2014 show that the proposed method presents better performance than other cutting-edge models.http://www.sciencedirect.com/science/article/pii/S2352484720314062Probabilistic load forecastingQuantile regressionConvolutional neural networkKernel density estimation |
| collection |
DOAJ |
| language |
English |
| format |
Article |
| sources |
DOAJ |
| author |
Hui He Junting Pan Nanyan Lu Bo Chen Runhai Jiao |
| spellingShingle |
Hui He Junting Pan Nanyan Lu Bo Chen Runhai Jiao Short-term load probabilistic forecasting based on quantile regression convolutional neural network and Epanechnikov kernel density estimation Energy Reports Probabilistic load forecasting Quantile regression Convolutional neural network Kernel density estimation |
| author_facet |
Hui He Junting Pan Nanyan Lu Bo Chen Runhai Jiao |
| author_sort |
Hui He |
| title |
Short-term load probabilistic forecasting based on quantile regression convolutional neural network and Epanechnikov kernel density estimation |
| title_short |
Short-term load probabilistic forecasting based on quantile regression convolutional neural network and Epanechnikov kernel density estimation |
| title_full |
Short-term load probabilistic forecasting based on quantile regression convolutional neural network and Epanechnikov kernel density estimation |
| title_fullStr |
Short-term load probabilistic forecasting based on quantile regression convolutional neural network and Epanechnikov kernel density estimation |
| title_full_unstemmed |
Short-term load probabilistic forecasting based on quantile regression convolutional neural network and Epanechnikov kernel density estimation |
| title_sort |
short-term load probabilistic forecasting based on quantile regression convolutional neural network and epanechnikov kernel density estimation |
| publisher |
Elsevier |
| series |
Energy Reports |
| issn |
2352-4847 |
| publishDate |
2020-12-01 |
| description |
Electricity load forecasting plays an indispensable role in the electric power systems. However, its characteristics of uncertainty and complexity are hard to handle. This paper proposes a probabilistic load forecasting approach named QRCNN-E. Specifically, the deep convolutional neural network is applied to model the non-linear relationship with the electricity load and its influencing factors. By replacing the traditional loss function with pinball loss, the network can eventually forecast loads in quantiles. Then, kernel density estimation takes quantile forecasts as inputs and produces deterministic and probabilistic results. Case studies on GEFCom2014 show that the proposed method presents better performance than other cutting-edge models. |
| topic |
Probabilistic load forecasting Quantile regression Convolutional neural network Kernel density estimation |
| url |
http://www.sciencedirect.com/science/article/pii/S2352484720314062 |
| work_keys_str_mv |
AT huihe shorttermloadprobabilisticforecastingbasedonquantileregressionconvolutionalneuralnetworkandepanechnikovkerneldensityestimation AT juntingpan shorttermloadprobabilisticforecastingbasedonquantileregressionconvolutionalneuralnetworkandepanechnikovkerneldensityestimation AT nanyanlu shorttermloadprobabilisticforecastingbasedonquantileregressionconvolutionalneuralnetworkandepanechnikovkerneldensityestimation AT bochen shorttermloadprobabilisticforecastingbasedonquantileregressionconvolutionalneuralnetworkandepanechnikovkerneldensityestimation AT runhaijiao shorttermloadprobabilisticforecastingbasedonquantileregressionconvolutionalneuralnetworkandepanechnikovkerneldensityestimation |
| _version_ |
1724373419577835520 |