Renewable energy needs to be utilized more heavily in the production of electricity since the United States government is to reach its goal of a net-zero emissions economy. Due to its quick installation growth and the quantity of solar energy, solar photovoltaics (PV) is one of the renewable energy sources that attract the most attention from the general public. Solar power generation, however, varies with the time of day and the seasons. It is necessary to estimate the solar power output over time so that we can assess the reliability of solar PV energy and make appropriate plans for when to rely on energy from other sources.
Since the solar PV power output depends on the temperature and solar irradiance, the temperature and irradiance data are to be obtained to determine the power output. However, the data collected over the years will take up a lot of space. Tensor decomposition is a technique that can be used to minimize data storage and forecast future data.
Without having to keep all of the original data, we may predict the temperature and solar irradiance at various places over time by computing Tucker decomposition using the Tensor Toolbox. Outputting the predictions to the two-stage PV integration in Simulink, power generation can then be forecasted over time in accordance with the fluctuating temperature and solar irradiation.