PV Maximum Power-Point Tracking by Using Artificial Neural Network

Abstract

The utilization of solar panels as an effort to implement new and renewable energy still has challenges in optimizing the performance of solar panels to produce electrical power efficiently. Therefore, a careful planning and operational process is needed to maximize the performance of solar panels. One of the important steps to plan carefully the process of installing solar panels is to estimate the production of electrical power generated by solar panels. This research was conducted to design a maximum voltage estimation system on solar panels by considering solar radiation and air temperature. The Mean Squared Error (MSE) parameter is used as a performance parameter of the artificial neural network used. The results showed a significant relationship between solar radiation, air temperature and voltage on solar panels made from polycrystalline. Testing and evaluation of the estimation system is done using independent data that has never been seen before. The MSE result in the test is 9.9602e-12 which has met the MSE target set in this study.