Multi-Hybrid STL-LSTM-SDE-MA Model Optimized with IWOA for Solar PV-Power Forecasting

The solar power forecasting has its importance in renewable energy future. In fact, there are infinite challenges behind accurate forecasting of solar PV power. This state-of-the-art research work proposes a hybrid model brings in the key strengths of Seasonal-Trend decomposition layer (STL), Long Short-Term Memory (LSTM), Stochastic Differential Equation (SDE), and Moving Average (MA) methodologies. So as to capture the complex patterns and dynamics of solar PV-power data. To add more concentration on accuracy forecasting the model is optimized using the Improved Whales Optimization Algorithm (IWOA). IWOA is an advanced optimization algorithm inspired by the social behaviour of humpback whales, known for their collaborative hunting strategies. The IWOA brings in improved accuracy to the model by enhancing search capabilities and convergence speed. The proposed hybrid STL-LSTM-SDE-MA model, optimized with IWOA, is evaluated using real time solar power data. The model's superiority in forecasting accuracy and robustness is proved by Comparative analyses with other forecasting approaches and conventional models. This optimized hybrid model represents a cutting-edge technological solution for solar PV-power forecasting and making it a valuable asset in shaping a greener and more resilient future.