Converter Averaging Approach for Modeling a Residential Supply Subsystem

Residential Supply Subsystems (RSSs) are part of the local grid developed to integrate and manage Renewable Energy Sources (RES) and energy storage (ES) by utilization of the Power Electronics Converters (PECs). PECs used in RSS must be designed and evaluated for different operating conditions, including specified hardware and advanced control algorithms. At the beginning of the designing process, simulation tools allow for testing many aspects without carrying the costs of manufacturing prototypes. Depending on factors like the number of internal components of the model, their complexity, and time constants under consideration, a simulation can require a significant amount of computing power and time. This is especially important for inherently cross-domain systems, with time constants varying significantly between domains, likewise for multiple PECs cooperating and controlled by one complex control platform. Therefore, it is necessary to devise models appropriate for the scenario to be studied. In this work, the large-signal averaging technique has been used to an RSS model and simulate its dynamics compared to the full-switching approach. In the analyzed system, the integration of a photovoltaic (PV) array and ES connected to a three-phase, four-wire grid, all with dedicated PECs and a local asymmetrical load, have been included. Different modes of operation are considered: charging/discharging modes of the ES, and maximum power point tracking (MPPT), both under unbalanced load conditions. A comparison of simulation times is performed for the switching, and the averaged models and the benefits of each modeling technique are summarized.