REMTCN-Informer A Hybrid Approach to Photovoltaic Power Forecasting

As the focus on photovoltaic (PV) energy intensifies due to its role in clean energy production, accurate forecasting of PV power generation has become crucial. The inherent volatility and complexity of photovoltaic energy output pose considerable challenges to the forecasting task, particularly when dealing with long-term correlations and considering multiple influencing factors such as weather fluctuations over time. This necessitates the development of highly accurate and adaptable forecasting models to comprehensively address and analyse these complex and changeable factors. To tackle these issues, this paper introduces a novel hybrid model, REMTCN-Informer, which synergistically integrates TCN with the Informer model. The REMTCN-Informer model incorporates a range of sophisticated technologies, including residual connection and normalisation mechanisms. These have been shown to enhance the accuracy and stability of photovoltaic power prediction. The experimental results demonstrate that the model exhibits excellent performance across all forecasting time periods. In comparison to traditional forecasting models, the REMTCN-Informer model demonstrates superior forecasting ability and robust performance.