Research on complementary algorithm of photovoltaic power missing data based on improved cloud model
Mao Yang, Dingze Liu, Yang Cui, Xin Huang, Gangui Yan · International Transactions on Electrical Energy Systems · 2020
The completion of the missing data of photovoltaic (PV) power is of great significance for the development of PV power related research. Based on the analysis of the main influencing factors of PV power, the PV power data of different illumination characteristics were classified by the goodness of fit. Combined with Copula theory and abnormal data discriminant criteria, the PV power anomaly data were effectively identified. We changed the normal random entropy in the traditional cloud model to the random entropy based on Copula theory, and established an improved cloud model of irradiance and power. Based on the probability of PV power conditions generated by the cloud model, according to the volatility of PV power, this article constructed a conditional interpolation algorithm. The comparison of the results of different random missing proportions proved the limitation of the traditional cloud model and the instability of the mean method. It also proved that the accuracy of the method was improved and the error was the lowest.