Кластеризация метеорологических условий для повышения точности прогнозирования генерации фотоэлектрической станции (Weather condition's clustering for improvement of PV power plants forecasting accuracy)

Хальясмаа Кристина Ильмаровна

Аннотация


The modern world is increasingly focusing on renewable energy sources and setting carbon-free electricity targets for cleaner production. According to statistics, solar technology has now become one of the most popular methods of generating electricity throughout the world. The expansion of renewable energy is not limited to regions where fossil fuels are scarce; countries such as China and Russia are also actively adopting these technologies.
Solar power plants play a significant role in the context of zero-emission approaches. Despite their importance in improving energy availability and reducing carbon emissions, solar power plants face challenges such as weather dependence and component disposal issues. One of the main challenges in terms of weather dependence is the uncertainty of power generation, which greatly affects the operation of power transmission networks as well as electricity tariffs. All of the above confirms the relevance of the chosen topic.
The object of the study is the generation of electricity based on photovoltaic stations.
The subject of the research is algorithms for clustering data on weather conditions and methods for forecasting electricity generation based on photovoltaic stations.
The purpose of the study is to tune a clustering algorithm for initial meteorological conditions data to improve the forecasting accuracy of photovoltaic power generation using machine learning methods.