Integrating Machine Learning into Modified CUK Converters for Household Solar Applications

The depletion of petroleum resources has resulted in a significant increase in the demand for renewable energy like solar which can be produced locally for household use. However, the efficiency of the solar panel is significantly lower, resulting in a decrease in the energy provided to the household, so to increase the power, a boosting network is added to the converter used to step up and down the voltage. The Machine Learning (ML) models are used in this work to predict the output voltage of the system with the varying parameters of the converter like switching frequency, and the solar panel. By predicting the output voltage in advance necessary measures can be taken so that the household appliances are provided with the required power when needed and to better ourselves for any sudden surge and drop in the power produced by the system, also adding feedback to the system can ensure that the power given the appliances always stay in the required amount to avoid any damage to them. This work also aims to find the best machine learning that can used for this application to predict the output voltage with less error thereby increasing the overall efficiency.