An Ensemble Model Based on Machine Learning Methods and Data Preprocessing for Short-Term Electric Load Forecasting

Lin, Yanbing; Luo, Hongyuan; Wang, Deyun; Guo, Hao; Zhu, Kunpeng

doi:10.3390/en10081186

Cited by 55 publications

(29 citation statements)

References 33 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Further, we compared the DALM with two other commonly used load decomposition algorithms, namely WT [6] and VMD [8]. We used the fast discrete WT based on the Mallat algorithm to decompose each column of the HLM L t into one approximation component and three detail components independently.…”

Section: Data Preprocessing Algorithmmentioning

confidence: 99%

“…Various load forecasting techniques have emerged over the past decades [5][6][7][8][9], among which artificial intelligence based methods have become promising solutions because they excel at mapping the relationship between dependent and independent variables [10,11]. Reis et al [6] embedded the discrete wavelet transform (WT) into the multilayer perceptron and proposed a multi-model short term load forecasting scheme.…”

Section: Introductionmentioning

confidence: 99%

“…In [7], the authors employed support vector regression (SVR) machines to build a parallel forecasting architecture for hourly load in a day and used particle swarm search to optimize the SVR hyperparameters. Lin et al [8] proposed an ensemble model based on variational mode decomposition (VMD) and extreme learning machine (ELM) for multi-step ahead load forecasting. Ko et al [12] developed a model based on SVR and radial basis function networks and leveraged the extended Kalman algorithm to filter the model parameters.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Spatiotemporal Feature Learning Based Hour-Ahead Load Forecasting for Energy Internet

Zhang

Wang

2020

Electronics

View full text Add to dashboard Cite

In this paper, we analyze the characteristics of the load forecasting task in the Energy Internet context and the deficiencies of existing methods and then propose a data driven approach for one-hour-ahead load forecasting based on the deep learning paradigm. The proposed scheme involves three aspects. First, we formulate a historical load matrix (HLM) with spatiotemporal correlation combined with the EI scenario and then create a three-dimensional historical load tensor (HLT) that contains the HLMs for multiple consecutive time points before the forecasted hour. Second, we preprocess the HLT leveraging a novel low rank decomposition algorithm and different load gradients, aiming to provide a forecasting model with richer input data. Third, we develop a deep forecasting framework (called the 3D CNN-GRU) featuring a feature learning module followed by a regression module, in which the 3D convolutional neural network (3D CNN) is used to extract the desired feature sequences with time attributes, while the gated recurrent unit (GRU) is responsible for mapping the sequences to the forecast values. By feeding the corresponding load label into the 3D CNN-GRU, our proposed scheme can carry out forecasting tasks for any zone covered by the HLM. The results of self-evaluation and a comparison with several state-of-the-art methods demonstrate the superiority of the proposed scheme.

show abstract

Section: Data Preprocessing Algorithmmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Spatiotemporal Feature Learning Based Hour-Ahead Load Forecasting for Energy Internet

Zhang

Wang

2020

Electronics

View full text Add to dashboard Cite

show abstract

“…VMD, developed by Dragomiretskiy and Zosso (2014), is a novel decomposition method that has been frequently used in many fields, such as biomedical image denoising (Lahmiri and Boukadoum, 2014), mechanical fault diagnosis (Huang et al, 2016), and seismic time-frequency analysis (Xue et al, 2016). Many previous studies have proven that the VMD method is superior to other decomposition approaches such as wavelet transform and EMD for signal denoising (Lin et al, 2017). The details of VMD are presented as Algorithm 1.…”

Section: Variational Mode Decomposition (Vmd)mentioning

confidence: 99%

Hybrid wind energy forecasting and analysis system based on divide and conquer scheme: A case study in China

Yang

Wang

et al. 2019

Journal of Cleaner Production

129

View full text Add to dashboard Cite

Wind energy, acknowledged as a promising form of renewable energy and the fastestgrowing clean method for electricity generation, has attracted considerable attention from many scientists and researchers in recent decades. However, wind energy forecasting is still a challenging task owing to its inherent features of non-linearity and randomness. Therefore, this study develops a hybrid wind energy forecasting and analysis system including a deterministic forecasting module and an uncertainty analysis module to mitigate the challenges in existing studies. In particular, these challenges are as follows: (1) It is difficult to guarantee that the data characteristics underlying the time series are effectively extracted; (2) in the modeling of each subseries, i.e., when the original data is decomposed into some time series, forecasting accuracy and stability are not simultaneously considered, and thus, they are not properly modeled; and (3) the best function to perform a deterministic forecasting and uncertainty analysis based on the forecaster of each subseries is unknown. The developed hybrid system consists of three steps: First, data preprocessing is conducted to capture and mine the main feature of the wind energy time series and weaken the noises' negative effects; second, multi-objective optimization is proposed to achieve the forecasting of each subseries with improvements in accuracy and stability; finally, search for the best function, which obtains the deterministic forecasting and uncertainty analysis results using an optimized extreme learning machine based on different modeling objectives, is conducted. Experimental simulations are performed using data from three sites in a real wind farm, which indicate that the developed system has a better performance in engineering applications than that of other methods. Furthermore, this system could not only be used as an effective tool for wind energy deterministic forecasting and uncertainty analysis, but also for other engineering application areas in the future.

show abstract

“…However, EMD lacks a mathematical definition and has weaknesses that diverge at end-points when decomposing the signal. To overcome the weakness of EMD, load forecasting studies using variational mode decomposition (VMD) have been proposed [22][23][24][25]. Existing regression methods with various decompositions, clustering algorithms, and probabilistic analyses have been investigated, as they can be used to identify load characteristics; however, they increase the dimension of the input [26][27][28].…”

Section: Introductionmentioning

confidence: 99%

Deep Learning Based on Multi-Decomposition for Short-Term Load Forecasting

et al. 2018

View full text Add to dashboard Cite

Load forecasting is a key issue for efficient real-time energy management in smart grids. To control the load using demand side management accurately, load forecasting should be predicted in the short term. With the advent of advanced measuring infrastructure, it is possible to measure energy consumption at sampling rates up to every 5 min and analyze the load profile of small-scale energy groups, such as individual buildings. This paper presents applications of deep learning using feature decomposition for improving the accuracy of load forecasting. The load profile is decomposed into a weekly load profile and then decomposed into intrinsic mode functions by variational mode decomposition to capture periodic features. Then, a long short-term memory network model is trained by three-dimensional input data with three-step regularization. Finally, the prediction results of all intrinsic mode functions are combined with advanced measuring infrastructure measured in the previous steps to determine an aggregated output for load forecasting. The results are validated by applications to real-world data from smart buildings, and the performance of the proposed approach is assessed by comparing the predicted results with those of conventional methods, nonlinear autoregressive networks with exogenous inputs, and long short-term memory network-based feature decomposition.

show abstract

An Ensemble Model Based on Machine Learning Methods and Data Preprocessing for Short-Term Electric Load Forecasting

Cited by 55 publications

References 33 publications

Spatiotemporal Feature Learning Based Hour-Ahead Load Forecasting for Energy Internet

Spatiotemporal Feature Learning Based Hour-Ahead Load Forecasting for Energy Internet

Hybrid wind energy forecasting and analysis system based on divide and conquer scheme: A case study in China

Deep Learning Based on Multi-Decomposition for Short-Term Load Forecasting

Contact Info

Product

Resources

About