Quantum Long Short-Term Memory Network With Adaptive Tasmanian Devil Optimizer for Earlier Prediction of Solar Irradiance Prediction

To address the volatility of photovoltaic (PV) power, this study introduces the OHA-QLSTM-ATDO-SIP framework for precise solar irradiance forecasting. The system utilizes a Multi-observation Fusion Kalman Filter (MOFKF) to pre-process data from the National Solar Radiation Data Base, followed by a Quantum Long Short-Term Memory (QLSTM) network for prediction. To enhance accuracy, an Adaptive Tasmanian Devil Optimizer (ATDO) is employed to fine-tune the QLSTM parameters. Tested across five cities