Method and system for solar power forecasting

A method for generating a solar power output forecast for a solar power plant by a hybrid solar power forecasting system, comprising:

• the hybrid solar power forecasting system including a physical subsystem being a master subsystem to generate the solar power output forecast and an artificial intelligence subsystem being a slave subsystem to optimize coefficients of the master subsystem;
• the hybrid solar power forecasting system using training, validation, and runtime modes for each of the physical subsystem and the artificial intelligence subsystem;
• using a clear sky model, a cloud model, and an irradiance-to-electrical-power model of the physical subsystem;
• using a spherical coordinate system-based multistate framework for diffuse irradiance in the cloud model of the physical subsystem for representing a three-dimensional (“3D”) matrix of cloud density;
• and, using remote sensing observations of the cloud density and cloud shadow location;
• using a processor of the hybrid solar power forecasting system, in a training mode, generating a trained artificial intelligence model to optimize the coefficients of the master subsystem using historical output data and historical input data, the historical output data including selected historical diffuse irradiance and direct normal irradiance forecasts and historical solar power output forecasts, the historical input data including selected historical remote sensing observations of cloud locations and density input data and historical artificial intelligence subsystem output forecasts data, the historical artificial intelligence subsystem output forecasts for bias correcting at least one of biases in cloudiness index of atmosphere in a trained physical subsystem;
• in a runtime mode, for a predetermined forecast horizon, applying the trained artificial intelligence model to current input data, the current input data including current remote sensing observations of the cloud locations and the density input data and current artificial intelligence subsystem output forecasts data, the trained physical subsystem to produce a diffuse irradiance forecast, a direct normal irradiance forecast, and the solar power output forecast, the current artificial intelligence subsystem output forecasts data for bias correcting at least one of the biases in the cloudiness index of the atmosphere in the trained physical subsystem;
• and,
• presenting the solar power output forecast on a display.