부산대학교 도서관

The vision of smart cities are being realized gradually by converting each building into a smart building. These are cyber-physical systems meant to provide best comfort to its habitant at most economical and environmentally sustainable way. There can be multiple loads for heating, cooling, lighting purposes. Renewables such as rooftop solar photovoltaic and small-scale wind turbines are also integrated for generating electricity that reduces the grid-dependency. This forms a Building Integrated Microgrid (BIMG) system consisting of its own sources, storage, and loads. Handling intermittent generation, variable load conditions, and external grid dynamics are the three essential factors to increase buildings’ smartness. In this work, an adaptive Distributed Model Predictive Control (DMPC) principle for energy management of renewables based grid-tied BIMG is presented. All possible arbitrages, viz., generation intermittency of renewables, time-variable load profiles, storage unit's charging and discharging cycles, interruption of non-critical loads, and dynamic pricing of grid energy are considered. The feasibility of the given strategy is validated by realizing extensive test cases over 24 hrs duration using variable profiles of solar irradiance, continuous dynamic load variation at small and large scale, interruption of non-critical loads, two-tier tariff, and battery longevity by accounting for smooth charging/discharging cycles on the MATLAB software and Real-Time Digital Simulator (RTDS) based Hardware-in-Loop (HIL) setup.