Fractal classification of typical meteorological day based on solar behavior (Case study: Karaj synoptic station)

Background and objectives: Today the main part of human`s needed energy is provided by fossil fuels. Due to the reduction of fossil fuel reserves and climate changes caused by increased emissions, the production and use of new sources of clean renewable energy with fewer emissions is a necessity. Due to the efficiency of energy production, solar energy is more pronounced; among other renewable energy sources.The utilization of the information of solar irradiance is in many industrial applications, photovoltaic systems, agriculture and solar collectors design. For this purpose, the fractal dimension is used as a classification criterion. In order to provide a model that allows classification of days to three types this study estimated the daily fractal index and the index of purity of sky. In this method, with the advantage of cumulative distribution function, the fractal dimension classifies the days of Karaj station in three classes: clear sky day, partially clouded sky day and clouded sky day.
Materials and methods: The experimental database contains global irradiance data and sunshain measured at Karaj site during 2014-2016 year. In the first stage after data quality control daily fractal dimension of solar radiation time series and clearness index was calculated. For each year two fractal thresholds was obtained using the cumulative distribution function (CDF). The days of year was classified in three classes of clear sky, partially cloudy sky and cloudy sky based on obtained thresholds. In the next step monthly analyses was done.
Conclusion: Results showed that high frequency of class 1 was occurred in August 2015, high frequency of class 2 was occurred in February 2016 and high frequency of class 3 was occurred in March 2015 respectively. Also these statistical properties show that our classification method leads to homogeneous groupings of the studied days since the standard deviations of D and KT are low in compared to their averages. The more important value of this standard deviation for class III (upper than 10%) is due to the fact that this class contains rainy days whose irradiance signals have a regular form thus a fractal dimension near to 1. However, the analysis of monthly values of D permits the detection of the months where the fluctuations of their radiances are most intense and those where these irradiances are very regular. This information is very beneficial to refine the assessing of photovoltaic systems and to reduce the initial costs by appropriate design and construction of solar energy systems suitable to the climate of the site of interest.
Results: This paper offers a method for the classification of radiation per day according to weather different classes, in order to exploit photovoltaic systems in using solar energy in the study area. This method has been proposed to classify the daily global irradiances into typical days using the fractal dimension as a basic criterion since it allows quantifying the irradiance fluctuations. This method defines fractal dimension thresholds using the cumulative distribution function. Then shown that it is possible to realize daily solar irradiances classification using the D thresholds obtained from the CDF method.Classification of the daily solar irradiance is important in design and installation of solar energy systems, especially PV arrays. Trends in the patterns of daily solar irradiance became significant information due to the recent interests in renewable technologies. This interest is essentially due to global warming and other negative effects to our environment. Such analyses presented in this purpose are of great interest as they reduce the initial costs by appropriate design and construction of solar energy systems suitable to the climate of the site of interest.