Artificial intelligence can support the transition towards renewable energies by managing and coordinating between the electricity supply and demand. AI algorithms combined with smart metering infrastructure and demand response programs, showed a great success for tackling the challenges of continuous balancing of electricity supply and demand, despite the intermittency and uncertainty of renewable energies. In this thesis, AI algorithms were used to predict consumer behaviour, find optimal pricing for electricity, and coordinate between different elements of supply, demand, and storage.
Smart grid is a central concept to safely lead the transition from traditional, central, high carbon emission electricity sources to distributed, clean and renewable electricity sources. The later are usually intermittent, depend on weather conditions and highly volatile, which makes them unsuitable for following the vagaries of electricity demand. Smart grid is the combination of realtime electricity measuring infrastructure (known as smart metering systems), telecommunication technologies and energy storage solutions, with adequate software and algorithms to optimize the electricity network and improve its flexibility. Demand side management, automatic control of production, consumption and storage, accurate demand and supply foecasting, and digitalization of the electricity market are the main objectives of the smart grid.
The research presented in this Thesis consists of four main contributions:
1. An empirical analysis of residential demand flexibility in response to dynamic electricity prices. The proposed model predicts the response patterns of electric appliances that can be shifted or reduced in a single household. Artificial neural networks are used to “learn” these patterns based on historical data of prices and corresponding consumption.
2. A novel concept of combining between price-responsiveness and heating/cooling systems (known as TCLs) as a flexibility component and their empirical evaluation. TCLs are studied from different angles: their behavior in response to prices, optimization of their pricing schemes, and their contribution to the flexibility of a microgrid.
3. A microgrid simulation based on realistic scenarios and real data from Finland. This microgrid model includes various flexibility components alongside the typical microgrid elements. The simulation includes a wind farm, an energy storage system, directly controllable TCLs and price responsive loads.
4. A comprehensive and empirical comparison of AI algorithms for automatic energy management in the context of a microgrid with flexibility. The most successful algorithms have learned to optimally coordinate between the components of the microgrid and perform appropriate actions according to the state of TCLs, the outdoor emperatures, the wholesale electricity prices, and the generated energy.
Smart grid technologies powered by AI algorithms can predict, optimize, and coordinate between elements of the power grid, adding flexibility to the grid and allowing a large-scale integration of renewable energies.
The doctoral dissertation of Master of Science Taha Nakabi, entitled Computational Intelligence for Smart Grid’s Flexibility - Prediction, Coordination, and Optimal Pricing will be examined at the Faculty of Science and Forestry on the 1st of December in Kuopio. The opponent in the public examination will be Professor Matti Lehtonen of Aalto University, and the custos will be Professor Pekka Toivanen of the University of Eastern Finland. The public examination will be held in English.
