Search

The digital transformation of the energy sector toward the Smart Grid paradigm, intelligent energy management, and distributed energy integration poses new requirements for computer science. Issues related to the automation of power grid management, multidimensional analysis of data generated in Smart Grids, and optimization of decision-making processes require urgent solutions. The article aims to analyze the use of selected artificial intelligence (AI) algorithms to support the abovementioned issues. In particular, machine learning methods, metaheuristic algorithms, and intelligent fuzzy inference systems were analyzed. Examples of the analyzed algorithms were tested in crucial domains of the energy sector. The study analyzed cybersecurity, Smart Grid management, energy saving, power loss minimization, fault diagnosis, and renewable energy sources. For each domain of the energy sector, specific engineering problems were defined, for which the use of artificial intelligence algorithms was analyzed. Research results indicate that AI algorithms can improve the processes of energy generation, distribution, storage, consumption, and trading. Based on conducted analyses, we defined open research challenges for the practical application of AI algorithms in critical domains of the energy sector. [ABSTRACT FROM AUTHOR]

Accurate electricity price forecasting has become a substantial requirement since the liberalization of the electricity markets. Due to the challenging nature of electricity prices, which includes high volatility, sharp price spikes and seasonality, various types of electricity price forecasting models still compete and cannot outperform each other consistently. Neural Networks have been successfully used in machine learning problems and Recurrent Neural Networks (RNNs) have been proposed to address time-dependent learning problems. In particular, Long Short Term Memory (LSTM) and Gated Recurrent Units (GRU) are tailor-made for time series price estimation. In this paper, we propose to use multi-layer Gated Recurrent Units as a new technique for electricity price forecasting. We have trained a variety of algorithms with three-year rolling window and compared the results with the RNNs. In our experiments, three-layered GRUs outperformed all other neural network structures and state-of-the-art statistical techniques in a statistically significant manner in the Turkish day-ahead market. [ABSTRACT FROM AUTHOR]

In order to combat climate change, many countries have promised to bolster Renewable Energy (RE) production following the Paris Agreement with some countries even setting a goal of 100% by 2025. The reasons are twofold: capitalizing on carbon emissions whilst concomitantly benefiting from reduced fossil fuel dependence and the fluctuations associated with imported fuel prices. However, numerous countries have not yet made preparations to increase RE production and integration. In many instances, this reluctance seems to be predominant in energy-rich countries, which typically provide heavy subsidies on electricity prices. With such subsidies, there is no incentive to invest in RE since the time taken to recoup such investments would be significant. We develop a model using a Neural Network (NN) regression algorithm to quantitatively illustrate this conjecture and also use it to predict the reduction in electricity price subsidies required to achieve a specified RE production target. The model was trained using 10 leading metrics from 53 countries. It is envisaged that policymakers and researchers can use this model to plan future RE targets to satisfy the Nationally Determined Contributions (NDC) and determine the required electricity subsidy reductions. The model can easily be modified to predict what changes in other country factors can be made to stimulate growth in RE production. We illustrate this approach with a sample use case. [ABSTRACT FROM AUTHOR]

We evaluate and compare two common methods, artificial neural networks (ANN) and support vector regression (SVR), for predicting energy productions from a solar photovoltaic (PV) system in Florida 15 min, 1 h and 24 h ahead of time. A hierarchical approach is proposed based on the machine learning algorithms tested. The production data used in this work corresponds to 15 min averaged power measurements collected from 2014. The accuracy of the model is determined using computing error statistics such as mean bias error (MBE), mean absolute error (MAE), root mean square error (RMSE), relative MBE (rMBE), mean percentage error (MPE) and relative RMSE (rRMSE). This work provides findings on how forecasts from individual inverters will improve the total solar power generation forecast of the PV system. [ABSTRACT FROM AUTHOR]

Short-term load forecasting (STLF) is very important for planning and operating power systems and markets. Various algorithms have been developed for STLF. However, numerous utilities still apply additional correction processes, which depend on experienced professionals. In this study, an STLF algorithm that uses a similar day selection method based on reinforcement learning is proposed to substitute the dependence on an expert's experience. The proposed algorithm consists of the selection of similar days, which is based on the reinforcement algorithm, and the STLF, which is based on an artificial neural network. The proposed similar day selection model based on the reinforcement learning algorithm is developed based on the Deep Q-Network technique, which is a value-based reinforcement learning algorithm. The proposed similar day selection model and load forecasting model are tested using the measured load and meteorological data for Korea. The proposed algorithm shows an improvement accuracy of load forecasting over previous algorithms. The proposed STLF algorithm is expected to improve the predictive accuracy of STLF because it can be applied in a complementary manner along with other load forecasting algorithms. [ABSTRACT FROM AUTHOR]