AI Summary & Key Takeaways
AI InsightIt is found that the hour wise model performs better for hour wise prediction of diffused radiation than available prediction models and hence is recommended further for determination of diffused and direct components of available data.
📌 3 Key Takeaways
Abstract
In order to carry out energy simulation using software, a file containing various hourly daily weather parameters including solar global, diffused and direct radiations must be known and these parameters shall be averaged for a minimum of three years. In the present case, the data having values of other parameters except diffuse and direct solar radiations was available from Centre of Energy and Environmental Engineering from September 2011 onwards. A yearly data of the mentioned radiations was measured at same Centre using SOLYS 2 Sun Tracker and the relationship of diffuse and global components is established using clearness index. This paper outlines three developed hourly linear regression models for data classified as annual, seasonal and hour wise (6AM to 6PM) measured data. The developed models along with various available prediction models for hour wise analysis are evaluated for performance statistical test methods of mean bias error, root mean square error and correlation coefficients. The hour wise model is found better over the other developed models in mentioned tests and hence is used along with various prediction models for prediction of measured data at 6AM, 11 AM, 1 PM and 6 PM. It is found that the hour wise model performs better for hour wise prediction of diffused radiation than available prediction models and hence is recommended further for determination of diffused and direct components of available data.