Forecasting of Power Demands Using Deep Learning

Abstract

The forecasting of electricity demands is important for planning for power generator sector improvement and preparing for periodical operations. The prediction of future electricity demand is a challenging task due to the complexity of the available demand patterns. In this paper, we studied the performance of the basic deep learning models for electrical power forecasting such as the facility capacity, supply capacity, and power consumption. We designed different deep learning models such as convolution neural network (CNN), recurrent neural network (RNN), and a hybrid model that combines both CNN and RNN. We applied these models to the data provided by the Korea Power Exchange. This data contains the daily recordings of facility capacity, supply capacity, and power consumption. The experimental results showed that the CNN model outperforms the other two models significantly for the three features forecasting (facility capacity, supply capacity, and power consumption).

Deep Learning Approach to Power Demand Forecasting in Polish Power System

“…For better generalization capacity of the network, the additional regularization is applied, which takes into account minimization of weights, the sensitivity of hidden neurons to minor changes in input signal values, as well as specialization of neurons in particular areas of input data [22]. This is an important advantage of autoencoder over other deep learning solutions, for example, CNN [17], in which such regularization is not provided. After finishing the learning procedure, the reconstructing part of the autoencoder is eliminated from the structure, and the hidden layer signals provide the input to the next stage of dimensionality reduction.…”

“…The paper of Correia et al [10] has presented multi-model methodology for forecasting the sales of liquefied petroleum gas cylinders. The problem of sensitivity and accuracy in the ensemble of predictors has been considered in the paper of Fernandez et al [11] In recent years deep learning techniques have attracted great interest in many branches of engineering applications, including short time series prediction [12][13][14][15][16][17]. Unlike traditional machine learning, the deep learning approach extracts features directly from data without the manual intervention of the human operator.…”

“…Tian C et al [7] and Kang et al [17] have proposed long short-term memory (LSTM) and convolutional neural network for short-term load forecasting, showing their superiority in time series prediction problems. Deep learning has shown its superiority.…”

Deep-Learning Forecasting Method for Electric Power Load via Attention-Based Encoder-Decoder with Bayesian Optimization

“…In many areas, deep-learning methods have made remarkable progress in image processing, video tracking, speech recognition, and natural language understanding [24][25][26][27]. Meanwhile, some researchers have pointed out that deep-learning methods including Deep Belief Networks (DBNs), Convolutional Neural Networks (CNNs), Recurrent Neural Networks (RNNs), Generative Adversarial Networks (GAN), Graph Neural Network (GNN), etc., are very valuable to carry out better power load prediction by making full use of massive time-series data [28][29][30]. In particular, the powerful type of deep-learning framework, RNNs, specially designed for temporal analysis and modeling, have already gained a great amount of concern due to their flexibility in obtaining underlying non-linear relationships and sequential rules [31,32].…”

Method for Determining the Optimal Capacity of Energy Storage Systems with a Long-Term Forecast of Power Consumption

“…Methods of medium-term forecasting based on historical data have proven themselves well for predicting the loads of large clusters of consumers, for example, the power system of a region [47,48], since the irregularities associated with the influence of micro-factors of each individual consumer are smoothed out, and the overall load schedule receives a clearly pronounced seasonality and a trend associated with a system-wide increase in electricity consumption. However, when considering a single enterprise, internal factors and business processes have a significant impact on energy consumption.…”