Photovoltaic power generation forecasting model based on IFCM

Abstract The forecasting of photovoltaic power generation is of important meaning to both security and stability runs for the electrified wire netting. In this paper, a photovoltaic energy generation forecasting model is proposed called IFCM, which is combined with isolated forest (IForest), fuzzy c-means, and support vector machine (SVM). Firstly, the data is cleaned through the isolation forest anomaly detection algorithm; Secondly, the historical data of similar days is used for Training. With the isolation forest anomaly detection algorithm, the proposed IFCM has achieved great results. In the experimental analysis, the proposed method is compared with the prediction result of the traditional IForest-SVM and IForest-BP models. The experimental results demonstrate that the proposed IFCM algorithm exhibits superior performance.