نظارت فتوولتائیک هوشمند مبتنی بر داده های بزرگ تابش خورشیدی و شبکه های حسگر بی سیم

Clean energy technologies, especially photovoltaic, have recently become more and more popular and important due to their substantial benefits for environment, economy, and energy security. How to improve the management and usage efficiency of photovoltaic power stations is a challenging problem that needs to be investigated deeply. In this paper, Wireless sensor networks (WSNs) are utilized to efficiently deliver the monitoring data of the photovoltaic (PV) modules from power stations to the monitoring center located in Cloud datacenter. With the aim of detecting the problems of PV modules from the monitoring big data, a two-class data fusion method is firstly developed to integrate the monitoring data at sensor nodes of WSNs; then an innovative semi-supervised Support vector machine (SVM) classifier is designed and trained by existing solar irradiance big data at the monitor center. With the prediction model provided by the trained classifier, an outlier detection algorithm is devised to classify and locate the problems of PV modules through calculating the average value of the questionable data. In order to evaluate the performance of the proposed methods, a comprehensive experimental platform is set up. The experimental results show that the predicted values match well with the theoretical value of power generation.