Chronological Probability Model of Photovoltaic Generation

Abstract

This paper proposes a chronological probability model of photovoltaic (PV) generation on the basis of conditional probability and nonparametric kernel density estimation. In addition to randomness of PV power, the correlation of PV powers between adjacent time points and the uncertainty of start and end moments of PV output can be represented. The proposed model can be employed to produce random PV power time series curve using a stochastic sampling method. The power data of three PV generators in different regions with distinct weather conditions and 34-node distribution network are used to demonstrate the correctness, effectiveness, and adaptability of the presented model.