This comprehensive open access book enables readers to discover the essential techniques for load forecasting in electricity networks, particularly for active distribution networks. From statistical methods to deep learning and probabilistic approaches, the book covers a wide range of techniques and includes real-world applications and a worked examples using actual electricity data (including an example implemented through shared code). Advanced topics for further research are also included, as well as a detailed appendix on where to find data and additional reading. As the smart grid and low carbon economy continue to evolve, the proper development of forecasting methods is vital. This book is a must-read for students, industry professionals, and anyone interested in forecasting for smart control applications, demand-side response, energy markets, and renewable utilization.

The overarching aim of this open access book is to present self-contained theory and algorithms for investigation and prediction of electric demand peaks. A cross-section of popular demand forecasting algorithms from statistics, machine learning and mathematics is presented, followed by extreme value theory techniques with examples. In order to achieve carbon targets, good forecasts of peaks are essential. For instance, shifting demand or charging battery depends on correct demand predictions in time. Majority of forecasting algorithms historically were focused on average load prediction. In order to model the peaks, methods from extreme value theory are applied. This allows us to study extremes without making any assumption on the central parts of demand distribution and to predict beyond the range of available data. While applied on individual loads, the techniques described in this book can be extended naturally to substations, or to commercial settings. Extreme value theory techniques presented can be also used across other disciplines, for example for predicting heavy rainfalls, wind speed, solar radiation and extreme weather events. The book is intended for students, academics, engineers and professionals that are interested in short term load prediction, energy data analytics, battery control, demand side response and data science in general.

This book is a printed edition of the Special Issue "Short-Term Load Forecasting by Artificial Intelligent Technologies" that was published in Energies

"This reference offers a wide-ranging selection of key research in a complex field of study,discussing topics ranging from using machine learning to improve the effectiveness of agents and multi-agent systems to developing machine learning software for high frequency trading in financial markets"--Provided by publishe

Since electricity plays a crucial role in industrial infrastructures of countries, power companies are trying to monitor and control infrastructures to improve energy management, scheduling and develop efficiency plans. Smart Grids are an example of critical infrastructure which can lead to huge advantages such as providing higher resilience and reducing maintenance cost. Due to the nonlinear nature of electric load data there are high levels of uncertainties in predicting future load. Accurate forecasting is a critical task for stable and efficient energy supply, where load and supply are matched. However, this non-linear nature of loads presents significant challenges for forecasting. Many studies have been carried out on different algorithms for electricity load forecasting including; Deep Neural Networks, Regression-based methods, ARIMA and seasonal ARIMA (SARIMA) which among the most popular ones. This thesis discusses various algorithms analyze their performance for short-term load forecasting. In addition, a new hybrid deep learning model which combines long short-term memory (LSTM) and a convolutional neural network (CNN) has been proposed to carry out load forecasting without using any exogenous variables. The difference between our proposed model and previously hybrid CNN-LSTM models is that in those models, CNN is usually used to extract features while our proposed model focuses on the existing connection between LSTM and CNN. This methodology helps to increase the model's accuracy since the trend analysis and feature extraction process are accomplished, respectively, and they have no effect on each other during these processes. Two real-world data sets, namely "hourly load consumption of Malaysia" as well as "daily power electric consumption of Germany", are used to test and compare the presented models. To evaluate the performance of the tested models, root mean squared error (RMSE), mean absolute percentage error (MAPE) and R-squared were used. The results show that deep neural networks models are good candidates for being used as short-term prediction tools. Moreover, the proposed model improved the accuracy from 83.17\% for LSTM to 91.18\% for the German data. Likewise, the proposed model's accuracy in Malaysian case is 98.23\% which is an excellent result in load forecasting. In total, this thesis is divided into two parts, first part tries to find the best technique for short-term load forecasting, and then in second part the performance of the best technique is discussed. Since the proposed model has the best performance in the first part, this model is challenged to predict the load data of next day, next two days and next 10 days of Malaysian data set as well as next 7 days, next 10 days and next 30 days of German data set. The results show that the proposed model also has performed well where the accuracy of 10 days ahead of Malaysian data is 94.16\% and 30 days ahead of German data is 82.19\%. Since both German and Malaysian data sets are highly aggregated data, a data set from a research building in France is used to challenge the proposed model's performance. The average accuracy from the French experiment is almost 77\% which is reasonable for such a complex data without using any auxiliary variables. However, as Malaysian data and French data includes hourly weather data, the performance of the model after adding weather is evaluated to compare them before using weather data. Results show that weather data can have a positive influence on the model. These results show the strength of the proposed model and how much it is stable in front of some challenging tasks such as forecasting in different time horizons using two different data sets and working with complex data.

Maintaining the balance between generation and consumption is at the heart of electricity grid operation. A disruption to this balance can lead to grid overloads, outages, system damage, rising electricity costs or wasted electricity. For this reason, accurate forecasting of load behavior is crucial. In this work, two classes of ML-based algorithms were used for load forecasting: the Hidden Markov Models (HMMs) and the Deep Neural Networks (DNNs), both of which provide stable and more accurate results than the considered benchmark methods. HMMs could be successfully used as a stand-alone predictor with a training based on Maximum Likelihood Estimation (MLE) in combination with a clustering of the training data and an optimized Viterbi algorithm, which are the main differences to other HMM-related load forecasting approaches in the literature. Adaptive online training was developed for DNNs to minimize training times and create forecasting models that can be deployed faster and updated as often as necessary to account for the increasing dynamics in power grids related to the growing share of installed renewables. In addition, the flexible and powerful encoder-decoder architecture was used, which helped to minimize the forecast error compared to simpler DNN architectures such as Convolutional Neural Networks (CNNs), Recurrent Neural Networks (RNNs), Long Short-Term Memory networks (LSTMs) and others.