Renewable energy systems such as photovoltaic (PV) and wind energy systems are widely designed grid connected or autonomous. This is a problem especially in small powerful system due to the restriction on the inverter markets. Inverters which are utilised in these kinds of energy systems operate on grid or off grid. In this study, a novel power management strategy has been developed by designing a wind-PV hybrid system to operate both as an autonomous system and as a grid-connected system. The inverter used in this study has been designed to operate both on-grid and off-grid. Due to the continuous demand for energy, gel batteries are used in the hybrid system. The designed Power Management Unit performs measurement from various points in the system and in accordance with this measurement; it provides an effective energy transfer to batteries, loads and grid. The designed control unit provided the opportunity to work more efficiently up to 10% rate.
Since the harmful effects of climate warming on our planet were first observed, the use of renewable energy resources has been significantly increasing. Among the potential renewable energy sources, photovoltaic (PV) system installations keep continuously increasing world‐wide due to its economic and environmental contributions. Despite its significant benefits, the inherent variability of PV power generation due to meteorological parameters can cause power management/planning problems. Thus, forecasting of PV output data (directly or indirectly) in an accurate manner is a critical task to provide stability, reliability, and optimisation of the grid systems. In considering the literature reviewed, there are various research items utilizing PV output power forecasting. In this study, a systematic literature review based on the search of primary studies (published between 2010 and 2020), which forecast PV power generation using machine learning and deep learning methods, is reported. The studies are evaluated based on the PV material used, their approaches, generated outputs, data set used, and the performance evaluation methods. As a result, gaps and improvable points in the existing literature are revealed, and suggestions which include novelties are offered for future works.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.