Predicting the Stabilization of Electric Power in Solar Panel Systems using a Support Vector Machine

Abstract

Energy efficiency in the utilization of electrical energy has attracted significant attention in academia and industry for many years, efforts to develop hybrid systems, especially with solar power, continue to be developed. In this study, the support vector machine (SVM) method was used to predicting stabilization the electric power from the PLTS which was implemented for ATS switching. The data process is recorded when the angle is shifted by the servo motor according to the time of the sun's shift and the angle range of the solar panels between 60°,90° and 120°. Based on the research that has been done, it can be concluded that the accuracy increases up to 98% in data set testing and 97% in real time testing. The SVM model is used in real time as a relay control prediction for switching between PLTS and PLN, this is evidenced by testing data taken from conditions when tested on a relay with an error prediction error of 3% during the day.