Building-level load forecasting has become essential with the support of fine-grained data collected by widely deployed smart meters.It acts as a basis for arranging distributed energy resources,implementing demand re...Building-level load forecasting has become essential with the support of fine-grained data collected by widely deployed smart meters.It acts as a basis for arranging distributed energy resources,implementing demand response,etc.Compared to aggre-gated-level load,the electric load of an individual building is more stochastic and thus spawns many probabilistic forecasting meth-ods.Many of them resort to artificial neural networks(ANN)to build forecasting models.However,a well-designed forecasting model for one building may not be suitable for others,and manually designing and tuning optimal forecasting models for various buildings are tedious and time-consuming.This paper proposes an adaptive probabilistic load forecasting model to automatically generate high-performance NN structures for different buildings and produce quantile forecasts for future loads.Specifically,we cascade the long short term memory(LSTM)layer with the adjusted Differential ArchiTecture Search(DARTS)cell and use the pinball loss function to guide the model during the improved model fitting process.A case study on an open dataset shows that our proposed model has superior performance and adaptivity over the state-of-the-art static neural network model.Besides,the improved fitting process of DARTS is proved to be more time-efficient than the original one.展开更多
Existing load forecasting methods typically assume that recent load data are available for prediction.This is not in conformity with reality since there is a time gap between the flow date(when power is consumed)and w...Existing load forecasting methods typically assume that recent load data are available for prediction.This is not in conformity with reality since there is a time gap between the flow date(when power is consumed)and when measurement values are obtained.To this end,this letter proposes an online learning-based probabilistic load forecasting method considering the impact of the data gap.Specifically,an adaptive ensemble backpropagation-enabled online quantile regression algorithm is developed to optimize the parameters of the attention network recursively using the newly obtained load observations.To further improve the reliability and sharpness of prediction intervals under significant data gaps,we introduce an online interval calibration technique.The proposed online learning method allows us to adaptively capture the dynamic changes in load patterns and alleviate the information lags caused by data gaps.Comparative tests utilizing real-world datasets reveal the superiority of the proposed method.展开更多
基金supported in part by the Seed Fund for Basic Research for New Staff of The University of Hong Kong(202107185032)and in part by the Alibaba Innovative Research programme.
文摘Building-level load forecasting has become essential with the support of fine-grained data collected by widely deployed smart meters.It acts as a basis for arranging distributed energy resources,implementing demand response,etc.Compared to aggre-gated-level load,the electric load of an individual building is more stochastic and thus spawns many probabilistic forecasting meth-ods.Many of them resort to artificial neural networks(ANN)to build forecasting models.However,a well-designed forecasting model for one building may not be suitable for others,and manually designing and tuning optimal forecasting models for various buildings are tedious and time-consuming.This paper proposes an adaptive probabilistic load forecasting model to automatically generate high-performance NN structures for different buildings and produce quantile forecasts for future loads.Specifically,we cascade the long short term memory(LSTM)layer with the adjusted Differential ArchiTecture Search(DARTS)cell and use the pinball loss function to guide the model during the improved model fitting process.A case study on an open dataset shows that our proposed model has superior performance and adaptivity over the state-of-the-art static neural network model.Besides,the improved fitting process of DARTS is proved to be more time-efficient than the original one.
基金supported in part by National Natural Science Foundation of China under Grant 72401055in part by National Natural Science Foundation of China under Grant 52277083in part by the joint founding of Guangdong,and Dongguan under Grant 2023A1515110939.
文摘Existing load forecasting methods typically assume that recent load data are available for prediction.This is not in conformity with reality since there is a time gap between the flow date(when power is consumed)and when measurement values are obtained.To this end,this letter proposes an online learning-based probabilistic load forecasting method considering the impact of the data gap.Specifically,an adaptive ensemble backpropagation-enabled online quantile regression algorithm is developed to optimize the parameters of the attention network recursively using the newly obtained load observations.To further improve the reliability and sharpness of prediction intervals under significant data gaps,we introduce an online interval calibration technique.The proposed online learning method allows us to adaptively capture the dynamic changes in load patterns and alleviate the information lags caused by data gaps.Comparative tests utilizing real-world datasets reveal the superiority of the proposed method.
