The project is fucused on the synthesis of a demand side-Energy Management System (EMS) of a grid connected-residential nanogrid for the application of real-time demand response services.
The EMS is in charge to schedule in real time the energy flows of a home or apartement considering a local prosumer, namely a user equipped with a local generation from renewable energy sources.
The EMS decision-making criterion must be interpretable, i.e. comparable to a gray/white box model. For this reason, a rule-based fuzzy inference system is going to be adopted as decision-making system and compared with more conventional EMS models such as Adaptive Control Systems or Rolling Time Horizon EMSs.
The nanogrid other than being connected to the grid, must manage and control the connection with conventional load devices (not controllable and not interruptible), shiftable loads (e.g. dishwasher, washing machine), an energy storage unit and the electric vehicle smart charging system.
The project proposed wants to continue the research started during my PhD and continued with my first postdoc. In the postdoc I personally coordinated and designed the EMS of a POR project regarding a modular multi-purpose nanogrid with successful results [1]. Regarding my research activity, three papers have been published on important journals in the last year, moreover a software about a generalized is free downloadable on Gitlab [2].
With this postdoc I would be able to continue the partnership between POMOS-DIET and Braga Moro on smart grid systems development, continue my teaching activity integrated with seminaries and therefore I'll be available to support student's exams and thesis, minor research and communications activities in collaboration with Prof. Antonello Rizzi, Prof. Fabio Massimo Frattale Mascioli and their students and collaborators which I worked with in the last years.
[1] https://www.bragamoro.com/it/progetto-moses-pr-1.html
[2] https://gitlab.com/labcoin/anfis-toolbox
The project will continue the work started with Braga Moro about a modular smart energy system.
In particular, the residential nanogrid is assumed to be composed by the following energy systems:
o A PV plant.
o Two controllable load units, a dishwasher, and a washing machine.
o The aggregation of conventional not controllable load devices.
o An electric vehicle which can be charged by a low power bidirectional smart charger.
o A suitable ESS unit.
o The connection with the grid.
A first work is addressed to the design of a suitable optimization problem including the objective function formulation which incorporates more performance indices (e.g. auto-consumption, peak shaving, valley filling, disservice to the user, local energy generation exploitation, stress on the distribution grid, stress on the energy storage etc.).
The performance indices included in the objective function would also allow to evaluate the impact and contribution of each controllable systems on the grid and the nanogrid.
The dataset used are already available thanks to the past works produced during my PhD and the postdoc [1,2].
Once the energy profile datasets about the PV generation and the FC station are properly generated and a useful level of abstraction of the problem under analysis is formulated it is possible to work on the EMS modelling.
The EMS is designed to define in real time how to efficiently distribute the energy flows of the nanogrid energy systems and the energy exchanged with the grid at the current time slot.
Therefore, the EMS synthesis will pass through a data driven approach.
Different EMS models will be implemented with the aim to proceed to a correct comparison of their performance and therefore to the selection of the best one.
One of the EMS model consist in a FIS synthesised by clustering techniques [3].
Furthermore, a Model Predictive Control EMS will be implemented for a suitable comparison.
The material produced, the simulation tests and results will be used for the writing of future publications.
Once is defined a good level of abstraction of the problem under analysis, the work will be focused on the EMS architecture. specifically, it will be defined the EMS input-output.
Continuing the work recently published in Applied Energy [1] a proper formulation for representing the MG energy flow exchange between the MG energy system and the grid will be studied to efficiently represent, analyse, and study the synthesised EMS decision making strategy.
Indeed, as the project will be specifically centred on the integration of the residential electric vehicle smart (home) charging and the management of the controllable loads. This aspect would drastically increase the number of free variables in the optimization problem, as well as the number of outputs in the EMS.
On the other hand, it is supposed that the involvement of these new flexibles loads and devices would bring more flexibility to the nanogrid.
[1] Microgrid Energy Management Systems Design by Computational Intelligence Techniques¿, Stefano Leonori, Alessio Martino, Antonello Rizzi and F. M. Frattale Mascioli, Applied Energy. DOI: 10.1016/j.apenergy.2020.115524.
[2] Intelligent Energy Flow Management of a Nanogrid Fast Charging Station Equipped with Second Life Batteries¿¿, Stefano Leonori, Giorgio Rizzoni, F. M. Frattale Mascioli and Antonello Rizzi, International Journal of Electrical Power and Energy Systems.
[3] A Generalized Framework for ANFIS Synthesis Procedures by Clustering Techniques¿, Stefano Leonori, Alessio Martino, Massimiliano Luzi, Antonello Rizzi and F. M. Frattale Mascioli, Applied Soft Computing. DOI: 10.1016/j.asoc.2020.106622.