scholarly journals Short-term load forecasting using Theta method

2019 ◽  
Vol 84 ◽  
pp. 01004 ◽  
Author(s):  
Grzegorz Dudek
Keyword(s):  
Empirical Studies ◽  
Load Forecasting ◽  
Seasonal Cycles ◽  
Short Term ◽  
Forecasting Accuracy ◽  
Forecasting Methods ◽  
Short Term Load Forecasting ◽  
Data Definition ◽  
System Data ◽  
Theta Method

The Theta method attracted the attention of researchers and practitioners in recent years due to its simplicity and superior forecasting accuracy. Its performance has been confirmed by many empirical studies as well as forecasting competitions. In this article the Theta method is tested in short-term load forecasting problem. The load time series expressing multiple seasonal cycles is decomposed in different ways to simplify the forecasting problem. Four variants of input data definition are considered. The standard Theta method is uses as well as the dynamic optimised Theta model proposed recently. The performances of the Theta models are demonstrated through an empirical application using real power system data and compared with other popular forecasting methods.

A comparative study of short-term load forecasting methods in distribution network

2021 ◽  
Vol 15 (1) ◽  
pp. 23-35
Author(s):  
Tuan Ho Le ◽  
◽  
Quang Hung Le ◽  
Thanh Hoang Phan
Keyword(s):  
Neural Network ◽  
Artificial Intelligence ◽  
Artificial Neural Network ◽  
Arima Model ◽  
Distribution Network ◽  
Load Forecasting ◽  
Short Term ◽  
Load Demand ◽  
Forecasting Methods ◽  
Short Term Load Forecasting

Short-term load forecasting plays an important role in building operation strategies and ensuring reliability of any electric power system. Generally, short-term load forecasting methods can be classified into three main categories: statistical approaches, artificial intelligence based-approaches and hybrid approaches. Each method has its own advantages and shortcomings. Therefore, the primary objective of this paper is to investigate the effectiveness of ARIMA model (e.g., statistical method) and artificial neural network (e.g., artificial intelligence based-method) in short-term load forecasting of distribution network. Firstly, the short-term load demand of Quy Nhon distribution network and short-term load demand of Phu Cat distribution network are analyzed. Secondly, the ARIMA model is applied to predict the load demand of two distribution networks. Thirdly, the artificial neural network is utilized to estimate the load demand of these networks. Finally, the estimated results from two applied methods are conducted for comparative purposes.

Application of SVM-RBF to Prediction of Short-Term Load

2012 ◽  
Vol 614-615 ◽  
pp. 811-814
Author(s):  
Hong Zhang ◽  
Sheng Zhu Li ◽  
Luan Song Yue ◽  
Zhao Yu Pian
Keyword(s):  
Radial Basis Function ◽  
Basis Function ◽  
Input Data ◽  
Rbf Neural Network ◽  
Load Forecasting ◽  
Matlab Simulation ◽  
Short Term ◽  
Forecasting Accuracy ◽  
Short Term Load Forecasting ◽  
Radial Basis

Short Term Load Forecasting is important to power system. It can be economic and reasonable to arrange start and stop of the Generator in wire net, The text adopt radial basis function neural networks. The GA-optimized multi-core radial basis function SVM is applied to extract useful data and short-term load forecasting accuracy based on RBF neural network has been improved. In this paper, The advantages of improving the algorithm is demonstrated by the application of the MATLAB simulation with the input data of the spring load collected from California, United States.

scholarly journals Use of Available Daylight to Improve Short-Term Load Forecasting Accuracy

Energies ◽  
2020 ◽  
Vol 14 (1) ◽  
pp. 95
Author(s):  
Miguel López ◽  
Sergio Valero ◽  
Carlos Sans ◽  
Carolina Senabre
Keyword(s):  
Load Forecasting ◽  
Linearization Method ◽  
Ar Model ◽  
Forecasting Model ◽  
Short Term ◽  
Forecasting Accuracy ◽  
Electric System ◽  
Forecasting Error ◽  
Short Term Load Forecasting ◽  
System Operator

This paper introduces a new methodology to include daylight information in short-term load forecasting (STLF) models. The relation between daylight and power consumption is obvious due to the use of electricity in lighting in general. Nevertheless, very few STLF systems include this variable as an input. In addition, an analysis of one of the current STLF models at the Spanish Transmission System Operator (TSO), shows two humps in its error profile, occurring at sunrise and sunset times. The new methodology includes properly treated daylight information in STLF models in order to reduce the forecasting error during sunrise and sunset, especially when daylight savings time (DST) one-hour time shifts occur. This paper describes the raw information and the linearization method needed. The forecasting model used as the benchmark is currently used at the TSO’s headquarters and it uses both autoregressive (AR) and neural network (NN) components. The method has been designed with data from the Spanish electric system from 2011 to 2017 and tested over 2018 data. The results include a justification to use the proposed linearization over other techniques as well as a thorough analysis of the forecast results yielding an error reduction in sunset hours from 1.56% to 1.38% for the AR model and from 1.37% to 1.30% for the combined forecast. In addition, during the weeks in which DST shifts are implemented, sunset error drops from 2.53% to 2.09%.

scholarly journals Short-Term Load Forecasting Using Neural Networks with Pattern Similarity-Based Error Weights

Energies ◽  
2021 ◽  
Vol 14 (11) ◽  
pp. 3224
Author(s):  
Grzegorz Dudek
Keyword(s):  
Time Series ◽  
Test Pattern ◽  
Empirical Studies ◽  
Load Forecasting ◽  
Learning Performance ◽  
Short Term ◽  
Training Samples ◽  
Wide Range ◽  
Pattern Similarity ◽  
Short Term Load Forecasting

Forecasting time series with multiple seasonal cycles such as short-term load forecasting is a challenging problem due to the complicated relationship between input and output data. In this work, we use a pattern representation of the time series to simplify this relationship. A neural network trained on patterns is an easier task to solve. Thus, its architecture does not have to be either complex and deep or equipped with mechanisms to deal with various time-series components. To improve the learning performance, we propose weighting individual errors of training samples in the loss function. The error weights correspond to the similarity between the training pattern and the test query pattern. This approach makes the learning process more sensitive to the neighborhood of the test pattern. This means that more distant patterns have less impact on the learned function around the test pattern and lead to improved forecasting accuracy. The proposed framework is useful for a wide range of complex time-series forecasting problems. Its performance is illustrated in several short-term load-forecasting empirical studies in this work. In most cases, error weighting leads to a significant improvement in accuracy.

Sign in / Sign up

Export Citation Format

Share Document