Short‐term photovoltaic power prediction based on MDCM‐GA‐LSTM model

Abstract

In recent years, photo-voltaic (PV) system has become one of the most potential renewable energy power generation technologies because of its many advantages. Considering the influence of the randomness of PV system on the operation and dispatching of power system, the necessity of a comprehensive forecasting model is increased rapidly. This paper proposes a short-term PV power forecasting method based on MDCM-GA-LSTM model to solve the problem of low accuracy of traditional and single forecasting model. First, the meteorological data are pre-processed based on wavelet denoising algorithm to improve the data quality. Second, a meteorological model is established based on genetic algorithm (GA) to optimize long short-term memory (LSTM) neural network. Finally, the network is trained with the modified data of the meteorological model as the input and the actual power as the output, and the power forecasting model is established. In this study, the forecasting accuracy is evaluated based on mean absolute error (MAE) and root mean square error (RMSE). The experimental results demonstrated that the MDCM-GA-LSTM model outperforms the conventional and single model by a difference of about 98.43% of the MAE and 98.97% of the RMSE error.