Centralised and decentralised control of active distribution systems: models, algorithms and applications.

Power system were traditionally planned and designed by assuming unidirectional power flows from power stations to loads. Nowadays, several factors (e.g., liberalization of the electricity market, need of increased reliability, and environmental issues) lead to a situation where electricity is produced also downstream the transmission level. Connecting generators to the distribution networks could provide several benefits to the whole system, but also technical and safety problems that must be faced. On the other hand, the loads are changing: new loads like electric vehicles and electric pumps are appearing in the network and they are going to modify the electricity consumption; while traditional loads are designed in order to be more efficient, but with additional functions or special features that require more energy. For all these reasons, since 2005, the interest on Smart Grid (electricity network that can intelligently integrate the actions of all users connected to it – generators, consumers – in order to efficiently deliver sustainable, economic and secure electricity supplies) increased. In this framework different techniques to control, operate and thereby integrate distributed energy resources into the network have been analysed and developed. The first technique designed is a centralised control, characterised by a central controller (Distribution Management System) that gathers information like the measures of the main electric parameters, energy price and indicates to DERs (Active Loads, Generators, Energy Storage) the optimal set points minimizing the system cost, subject to technical and economical constraints. The second technique developed is a decentralised control using Multi Agent Systems (MAS). This type of control has been designed and developed for the direct control of active demand and plug-in electric vehicles, managed by the Aggregator, entrusted by the end users to change their consumption habits according to their needs. Moreover, the proposed decentralised MAS, with the active participation of small consumers in the electricity system, support the integration of the Electric Vehicles in the LV distribution network and reduce its harmful impact on voltage regulation. The techniques and the algorithms proposed by the author are analysed and applied in representative Italian Distribution networks, by taking into account the development of the distribution system according to the load profile evolution, providing several examples to underline the importance of the Active Management for deferring the reinforcement of the existing grid infrastructures, increasing the hosting capacity of the network