Short term forecasting of solar radiation and power output of 89.6kWp solar PV power plant

Short term forecasting of solar radiation is useful for power plant operations, grid balancing, real-time unit dispatching, automatic generation control and trading. Solar forecasting is an essential tool in solar PV power plant to improve quality of energy delivery to the grid and to reduce weather dependent ancillary costs. In this paper short term forecasting of solar radiation and power output of 89.6 kWp solar PV power plant has been conducted. A new model has been proposed to conduct short term solar forecasting for different days of the year at Amity University Haryana (AUH) campus (28.4595 degrees N, 77.0266 degrees E) located in the northern region of India whereas auto-regressive integrated moving average (ARIMA) model is applied to forecast power output from the solar PV power plant. Root mean square error (RMSE) and Forecast Score (FS) has been used to for accessing the quality of forecasting models. The proposed model for prediction of solar radiation on tilted surface is simple and has very high accuracy. The model has ability to incorporate uncertainty due to environmental conditions. The proposed model is compared with Smart Persistence model and ARIMA model and it has been observed that it has better RMSE and Forecast Score than both Smart Persistence model and ARIMA model for both 15 min and 30 min time horizon. ARIMA model provides a reliable forecast for both solar radiation and solar PV power output. It is flexible enough to accept more information and its performance improves with the increase in number of data points. (C) 2019 Elsevier Ltd. All rights reserved.

Automated summary

This study focuses on short-term forecasting of solar radiation and power output of a solar PV power plant located in northern India. A new, simple, and accurate forecasting model has been proposed, which outperforms the Smart Persistence model and ARIMA model in terms of forecasting accuracy.