A Novel Short-Term Load Forecasting Method by Combining the Deep Learning With Singular Spectrum Analysis

Abstract

One of the major issues about the operation of power systems is the prediction of load demand. Moreover, load forecasting is of prime concern to system operators. Recently, the integration of power system elements, such as renewable energy sources, energy storages and electricity vehicle, brings more challenges, particularly when there are large fluctuations in forecasting cycle. This study concentrates on short-term load demand forecasting and proposes a hybrid method that combines Singular Spectrum Analysis (SSA) with deep-learning Neural Network (NN) techniques. In the beginning, the SSA technique is applied as an initial filter to remove noises. Next, a hybrid neural network, including Backpropagation Neural Network (BPNN) and Long Short-Term Memory (LSTM), is developed and trained. Then, the trained network is used as the core forecasting algorithm. Each SSA has different forms to combine with neural networks. The performance of the proposed forecasting algorithm is demonstrated using the power demand data recorded in Taiwan. Furthermore, this study compares the forecasting results by five models, including SSA, SSA-BPNN, ANN, SSA-LSTM, Wavelet Neural Network (WNN) and LSTM. The forecasting results reveal that the proposed forecasting model using Singular Spectrum Analysis provides the best performance on load forecasts.

Research on Peak Load Prediction of Distribution Network Lines Based on Prophet-LSTM Model

“…Li et al [33] developed a combined forecasting model that uses LSTM and XGBoost methods to improve the accuracy of electric load forecasting. Pham et al [34] developed a hybrid forecasting method integrating singular spectrum analysis (SSA) and deep neural network techniques. Similarly, Ciechulski et al [35] used artificial neural networks and SSA techniques to investigate efficient and accurate methods for 1 h and 24 h power load pattern forecasting.…”

Multi‐level refined power system operation mode analysis: A data‐driven approach

“…Given the strong periodicity of power system operation data, the SSA technique is chosen to process the operation data set and mine data information with different time resolutions. The matrix D formed by the power system operation data set decomposes the original time series into highfrequency, low-frequency, and noise matrices through normalization, matrix decomposition, and data reconstruction [20,21]. Different sampling point parameters, trend components, periodic components, and noise components with different time resolutions in each dimension of the power system operation data matrix D can be extracted to form a trend component matrix TD and a periodic component matrix SD.…”

Day-Ahead Spot Market Price Forecast Based on a Hybrid Extreme Learning Machine Technique: A Case Study in China

“…According to the above analysis, some key elements are missing from the current studies on electricity price forecasting, namely: (a) Many current studies ignore the useless information brought by the large amount of electricity price data when screening data features, which not only causes a reduction in forecasting accuracy but also affects the operational efficiency of forecasting models [34]. (b) The existing prediction models are generally based on a single sample set composed of features, which leads to the extraction of too much data, resulting in poor prediction accuracy [35,36].…”