Artificial Intelligence Applications Within the context of Digital Transformation

This article discusses the benefits, impact and challenges of the digital transformation on human life and industry. It focuses on the use of Artificial Intelligence (AI) techniques in order to address these challenges, in particular the ones related to the energy transition. Addressing these challenges requires the development of scalable advanced methods and tools able to manage and process efficiently and online the huge data streams produced by heterogeneous technologies and systems in order to extract useful knowledge, recommendations, or rules. The latter are then used within decision support tools in order to solve multiple problems such as enhancing the energy efficiency of traditional/renewable power generators through user participation, facilitating the penetration (integration) of distributed/centralized renewable energy systems into electrical grids, reducing the peak load by the use of efficient demand response strategies, balancing and optimizing generation and consumption, reinforcing the electrical grid protection (grid resilience, fault diagnosis and prognosis, grid self-healing and recovery, etc.), ensuring cybersecurity and privacy issues, etc. This paper overviews the AI methods and techniques used to solve these problems. These methods will be classified according to some meaningful criteria such as: the market players interactions (service providers, producers, grid operators, consumers, integrators, etc.), the available data (supervised, unsupervised, semi-supervised), their objective (prediction, optimization and control), their challenges and requirements, the application domain (demand side management, load monitoring, microgrids, energy consulting (residents, utilities), energy saving, dynamic pricing revenue management, wind/solar energy management, mobility and electrical vehicles, etc.), the system scale (appliance, home, building, neighborhood, etc.), the built decision support system (passive with no control only recommendation, active with control, online, offline), the used features (consumption/energy features, statistical features, temporal features, civic, house or building features, etc.), the type of data (active power consumption, reactive power, RMS (Root Mean Square) current, RMS voltage, power factor, energy prices, weather conditions, user activities and behaviors, etc.), the level of intrusiveness, etc. The goal of this classification is to provide readers with an overview of the use of machine learning methods and techniques in order to address the different challenges and problems related to the digital transformation within the context of energy transition.