A Deep Learning Model for Electricity Demand Forecasting Based on a Tropical Data

Abstract

Electricity demand forecasting is a term used for prediction of users’ consumption on the grid ahead of actual demand. It is very important to all power stakeholders across levels. The power players employ electricity demand forecasting for sundry purposes. Moreover, the government’s policy on its market deregulation has greatly amplified its essence. Despite numerous studies on the subject using certain classical approaches, there exists an opportunity for exploration of more sophisticated methods such as the deep learning (DL) techniques. Successful researches about DL applications to com¬puter vision, speech recognition, and acoustic computing problems are motivation. However, such researches are not sufficiently exploited for electricity demand forecasting using DL methods. In this paper, we considered specific DL techniques (LSTM, CNN, and MLP) to short-term load fore¬casting problems, using tropical institutional data obtained from a Transmission Company. We also test how accurate are predictions across the techniques. Our results relatively revealed models appropriateness for the problem.

Investigating the Drivers of Grid Electricity Demand in Nigeria: Harnessing the Power of Machine Learning and Artificial Intelligence Algorithms

“…The study concluded that the ANN model is more accurate. Also, Adewuyi et al (2020) carried out a long-term electricity demand forecast in Nigeria based on deep learning algorithms. The specific algorithms included convolutional neural networks (CNN), long short-term memory networks (LSTM), and multilayer perceptron (MLP).…”

“…For instance, Abdusalam et al (2016), Melodi et al (2017), Saglam et al (2022), andHasanah et al (2020) provided evidence that supports the optimal performance of neural network ML algorithm; however, Chapagain et al (2020), Yotto et al (2023 and Eya et al (2023) refuted their optimal performance. However, studies about electricity demand modelling using ML techniques for Nigeria, such as Abdusalam et al (2016), Melodi et al (2017) and Adewuyi et al (2020) adopted the neural network algorithms, but Eya et al (2023) that also used the neural network algorithms for Nigeria reported that the algorithm underperformed compared to artificial neuro-fuzzy inference system (ANFIS) algorithm.…”

Time series forecasting of electrical energy consumption using deep learning algorithm

“…However, scientific breakthroughs have made it possible to develop more accurate and precise forecasting techniques. Some of these techniques fall within the category of traditional forecasting approaches, and some have inaccurate load estimation problems (Adewuyi et al, 2020). An accurate shortterm forecasting strategy is therefore essential.…”

“…As time went on, Adewuyi et al (2020) created specific DL techniques (LSTM, CNN, and MLP) for short-term load forecasting issues using tropical institutional data obtained from a Transmission Company of Nigeria (TCN) and one-year weather data gathered from the Nigerian Meteorological Agency (NiMet) for an institutional customer. They evaluated the predictive accuracy of the various methodologies, and their findings indicated that, when comparing the three, the LSTM model performed on average in terms of the training, validation, and testing metrics.…”

“…A more sophisticated approach to STLF is using machine learning techniques that can learn from data (Adewuyi et al, 2020). However, each of these strategies has a unique set of issues, particularly when fitting a model to realize its innate abilities on a job while dealing with a large amount of applicable data.…”