Editorial to the Special Issue “AI Applications to Power Systems”

Abstract

This Special Issue consists of the successful invited submissions to Energies on the very topical subject area of “AI applications to power systems”.

The energy system of the future is a work in progress. Many countries around the world have made a target for their energy system to be completely renewable. How the power system eventually forms, such as the business models, the key players, and its architecture, as well as how it works, will be dependent on the outcomes of trends, forces, regulations, and strategic actions by many diverse players in the energy sector.

Digitalisation of the traditional generation plan and transmission has been proposed. However, there is no existing work in the literature on the digitalisation of the wind turbine generation or solar photovoltaic farms. The future power system will be very complex due to the multiple forces affecting various levels of the system, especially the distribution network level. The systems are composed of thousands of local distribution areas operated by distribution operators (suppliers) on top of the consumers, etc. Thus, although it may not appear too different, the power flows are no longer just one way, going from the bulk power system to the consumer end. In fact, the power flow will be very hard to trace because it can be from one consumer to another, the consumer to the grid, or the grid to the consumer, and some will be mobile/random due to the charging/discharging of electric vehicles.

These types of renewable energy resources (solar PV and wind turbine generation) are incredibly dependent on nature (wind speed, wind direction, temperature, solar irradiation, humidity, etc.). Thus, the outputs are highly stochastic in nature. Data science techniques for handling real-time big data will ideally fit this stream. Furthermore, integrated systems modelling methods and concepts are needed for the study of the self-organisation, complexity, emergent properties, and dynamical behaviour of complex systems for their holistic understanding, management, and development based primarily on neural networks, fuzzy and soft systems/fuzzy cognitive maps, network modelling, and mathematics. Other advanced applications in computational early detection of mastitis and computer-based decision support systems for complex systems are also needed. Due to the scale of the network and the amount of data that need to be digitised, new techniques in data mining and AI approaches are needed in order to analyse and predict the behaviour of these complex systems.

I would like to thank the staff and reviewers for their efforts and input. The task of editing and selecting papers for this collection was found to be both stimulating and rewarding [1,2,3,4,5,6].