Classification method of PV production patterns for energy flow analysis in design phase of energy-sharing community

To build an energy-sharing community, the energy flow, while considering the changes in the data within the community, must be identified in the planning and design phases. In this study, a procedure was suggested to assume PV production in a specific district and plan PV systems for the energy-sharing community using a method to identify and quantify daily patterns of PV production based on Discrete Fourier Transform (DFT). To verify its validity, the measured energy production of five PV systems and solar radiation within a district in Jincheon, South Korea, in 2019 were used. The proposed procedure based on the result of the study was as follows: First, apply DFT for daily PV production and solar radiation to convert continuously changing data with respect to time to simple expressions with amplitudes in frequency domain. Second, identify similar patterns in the data with different amplitudes by calculating the distribution of DFT results. Third, compare the DFT distribution of PV production and solar radiation for the same date and defined the relationship between them. Fourth, classify daily solar radiation types with respect to DFT distribution and define the characteristics of solar radiation based on the DFT distribution. Finally, assume annual PV production for a specific district from the past annual solar radiation. The proposed method is expected to accurately predict energy for a district via a simple calculation using past solar radiation data without complex simulation in the design phase of an energy-sharing community. Furthermore, this method can be used to estimate the efficiency of annual PV production for each district. Therefore, this can aid in the decision making for installing PV systems for specific districts.