Forecasting Global Solar Insolation Using the Ensemble Kalman Filter Based Clearness Index Model

This paper describes a novel approach in developing a model for forecasting of global insolation on a horizontal plane. In the proposed forecasting model, constraints, such as latitude and whole precipitable water content in vertical column of that location, are used. These parameters can be easily measurable with a global positioning system (GPS). The earlier model was developed by using the above datasets generated from different locations in India. The model has been verified by calculating theoretical global insolation for different sites covering east, west, north, south and the central region with the measured values from the same locations. The model has also been validated on a region, from which data was not used during the development of the model. In the model, clearness index coefficients (K-T) are updated using the ensemble Kalman filter (EnKF) algorithm. The forecasting efficacies using the K-T model and EnKF algorithm have also been verified by comparing two popular algorithms, namely the recursive least square (RLS) and Kalman filter (KF) algorithms. The minimum mean absolute percentage error (MAPE), mean square error (MSE) and correlation coefficient (R) value obtained in global solar insolation estimations using EnKF in one of the locations are 2.4%, 0.0285 and 0.9866 respectively.

Automated summary

This paper presents a novel approach to forecast global insolation and utilizes measurements from a global positioning system (GPS) to determine parameters such as latitude and precipitable water content. The model is verified and validated using data from various locations, and the performance is compared with other popular algorithms. The results show high accuracy and effectiveness in estimating global solar insolation.