Kernel Density Estimation for Stochastic Modeling of PV Power Output

Penetration of Distributed Energy Resources (DERs) stands the problem of taking into account uncertainty and volatility of power output. Such problem becomes crucial for choosing capacities of DERs and storages, estimating reliability indexes and selecting energy management strategies. This paper presents research based on data from 34-kWp grid-connected PV system, located in the campus of Federal University of Goias (Brazil). The data were used for modeling of stochastic generation output for photovoltaic panels during the year period based on kernel distributions, obtained from fitting values of actual power output and amount of produced energy.