Application of improved artificial neural networks in short-term power load forecasting

Sun, Weiqiang; Liu, Mohan

doi:10.1063/1.4926771

Cited by 11 publications

(7 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The k ‐means algorithm defines centroids through an iterative procedure and then assigns input data to the nearest centroid. It has been applied to group atmospheric patterns produced by SOMs in other works, such as Weia and Mohanb, Ohba et al, and Vesanto and Alhoniemi and is chosen here for its computational efficiency and interpretability. Here, we are only interested in “hard” assignments of observations to clusters, in contrast to probabilistic methods which make “soft” assignments based on the probability of belonging to each mixture component, such as Gaussian mixture models.…”

Section: Forecasting Frameworkmentioning

confidence: 99%

“…() Unsupervised learning, or clustering, techniques may be used to codify these large‐scale atmospheric circulation patterns in terms of a relatively small number of distinct modes() defined based on the fields of mean sea‐level pressure (SLP) and geopotential height, for example, for each time instant of interest. To reduce the number of modes for specific applications, a second clustering stage may be applied, akin to Weia and Mohanb and Ohba et al In the meteorological literature, this process is sometimes referred to as “classification,” but we avoid the use of that term here as in broader usage, this term implies a form of supervised learning. Given the length scale of the synoptic‐scale features, modes are typically determined on a daily temporal resolution; however, the same methods can be used to cluster patterns on any temporal scale from subdaily to seasonal depending on the application.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Improved very short‐term spatio‐temporal wind forecasting using atmospheric regimes

2018

View full text Add to dashboard Cite

We present a regime-switching vector autoregressive method for very short-term wind speed forecasting at multiple locations with regimes based on large-scale meteorological phenomena.Statistical methods for wind speed forecasting based on recent observations outperform numerical weather prediction for forecast horizons up to a few hours, and the spatio-temporal interdependency between geographically dispersed locations may be exploited to improve forecast skill. Here, we show that conditioning spatio-temporal interdependency on "atmospheric modes" derived from gridded numerical weather data can further improve forecast performance. Atmospheric modes are based on the clustering of surface wind and sea-level pressure fields, and the geopotential height field at the 5000-hPa level. The data fields are extracted from the MERRA-2 reanalysis dataset with an hourly temporal resolution over the UK; atmospheric patterns are clustered using self-organising maps and then grouped further to optimise forecast performance. In a case study based on 6 years of measurements from 23 weather stations in the UK, a set of 3 atmospheric modes are found to be optimal for forecast performance. The skill of 1-to 6-hour-ahead forecasts is improved at all sites compared with persistence and competitive benchmarks. Across the 23 test sites, 1-hour-ahead root mean squared error is reduced by between 0.3% and 4.1% compared with the best performing benchmark and by an average of 1.6% over all sites; the 6-hour-ahead accuracy is improved by an average of 3.1%.

show abstract

Section: Forecasting Frameworkmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Improved very short‐term spatio‐temporal wind forecasting using atmospheric regimes

2018

View full text Add to dashboard Cite

show abstract

“…Improving the quality of load prediction has always been the focus and difficulty of researchers. At present, the commonly used methods include time series method, regression analysis method [9], similar day method, grey prediction method [10], artificial neural network, etc., and machine learning methods such as support vector machine (SVM) [11] also have good applications.…”

Section: Micro Grid Load Predictionmentioning

confidence: 99%

Load Prediction of Microgrid Optimal Operation Based on Improved Algorithm in Machine Learning

2020

IJOMAM

View full text Add to dashboard Cite

Load prediction is an important problem in microgrid operation. In this study, the kernel extreme learning machine (KELM) algorithm in machine learning was used for load prediction. In order to improve the prediction accuracy, the KELM algorithm was improved by combining with Ant Colony Optimization (ACO) algorithm and Particle Swarm Optimization (PSO) algorithm, and then experiments were carried out on the property load of two communities in May-July. The experimental results showed that MAPE of the improved algorithm in predicting the load of the two communities was 1.43% and 1.57% respectively, and the operation time was 215 s and 223 s respectively, which were better than than the support vector machine (SVM) and KELM algorithms; the prediction results were close to the actual values, and the error changes were stable, which verified the effectiveness of the improved algorithm. The experimental results make some contributions to improve the accuracy of load prediction and promote the optimal operation of microgrid, which is conducive to the further development of microgrid.

show abstract

“…Short-term load forecasting has very important significance: firstly, it provides a guarantee for the safe and economic operation of the power system; secondly, it provides the basis for grid dispatch, power supply plan, and transaction plan in the market environment; [10][11][12] thirdly, it provides a basis for the discovery of potential faults and helps the power grid to operate reliably; at last, it gives a basis for users to manage power consumption plans. [13][14][15] The impact of short-term load forecasting depends on the accuracy of the forecast.…”

Section: Introductionmentioning

confidence: 99%

Short‐term load forecasting based on LSTNet in power system

Liu

Chen

et al. 2021

Int Trans Electr Energ Syst

View full text Add to dashboard Cite

Accurate short-term power load forecasting is very important in power grid decision-making operations and users power management. However, due to the nonlinear and random behavior of users, the electrical load curve is a complex signal. Although a lot of research has been done in this field, a predictive model with good accuracy and stability is still needed. To further improve this situation, this article proposes a novel model: long-term and short-term time series network (LSTNet) to predict load. This article firstly analyzes the correlation between other variables and load and uses Spearman correlation coefficient to measure the impact of other variables on the load and does autocorrelation analysis of load itself. Then, this article designs a load forecasting model based on LSTNet. The model uses a convolutional layer composed of a convolutional neural network (CNN) to capture short-term characteristics of load and short-term dependencies of variables, while a recurrent layer and recurrent-skip layer composed of long-term short-term memory network (LSTM) to capture long-term characteristics and variables of load long-term dependence, the adaptive regression part composed of autoregressive model (AR) to improve the robustness of the model. The experiment results show that the LSTNet model has better performance.

show abstract

Application of improved artificial neural networks in short-term power load forecasting

Cited by 11 publications

References 22 publications

Improved very short‐term spatio‐temporal wind forecasting using atmospheric regimes

Improved very short‐term spatio‐temporal wind forecasting using atmospheric regimes

Load Prediction of Microgrid Optimal Operation Based on Improved Algorithm in Machine Learning

Short‐term load forecasting based on LSTNet in power system

Contact Info

Product

Resources

About