Share
Export Citation
Predictive Maintenance System Using Support Vector Machine Algorithm for Dust Cleaning on Solar Panels
Mukasir H.
Proceedings Icmeralda 2023 International Conference on Modeling and E Information Research Artificial Learning and Digital Applications
Abstract
Solar panels, which are the core component of a solar energy system, have great potential to produce clean and sustainable energy. However, unstable performance and potential interference with solar panels can result in a significant reduction in energy output. One of the main disruptions in solar panel operations is the early detection of disruption, especially dust deposition on solar panels. To overcome this challenge, smarter and more economical solutions are needed. This research focuses on developing an early prediction method for solar panel disruption due to dust deposition using the Support Vector Machine (SVM) algorithm in the monitoring system. SVM has proven to be a very effective tool for classification and detection problems. By integrating this model into a solar energy monitoring system, we hope to reduce preventive maintenance costs and increase the operational efficiency of solar energy systems. Evaluation results show that the model has an accuracy of 80%, indicating a good overall correct prediction rate. By implementing the early prediction method for solar panel disruption using the Support Vector Machine (SVM) algorithm in the solar energy monitoring system, preventive maintenance can be carried out in a more timely and efficient manner. This not only reduces operational costs but also significantly increases the productivity and reliability of the solar energy system.