17 May 2018
Published online 13 September 2017
A new software-based tool can forecast the intensity of sunlight across different geographical locations.
A Saudi Arabian scientist has developed a new tool that can efficiently predict availability of solar rays at different locations in the kingdom and elsewhere in the Middle East.
His model can help fellow scientists and solar operators make the best possible plans to harvest energy from the sun.
This new software-based tool predicts the incidence of solar radiation at a given place through feeding parameters, such as temperature, humidity, wind speed and sunlight availability, into a computer-based system known as a "fuzzy system". The system then provides projections for solar radiation availability for the next day1.
“Solar plants depend on the level of incident solar radiation on solar panels,” explains the study's author Majid Almaraashi, assistant professor of artificial intelligence at the University of Jeddah, Saudi Arabia. Knowing an expected level of radiation beforehand can help manage different aspects of power plants including storage, operations, and finally, hybridization with other power resources. “A reliable prediction method with small errors is of great importance," he says.
Demand for electricity in Saudi Arabia is increasing at a rate of 7.5% per year. According to a study conducted at the King Abdullah City for Atomic and Renewable Energy, Saudi Arabia, the demand is expected to reach 120 gigawatts by 2028.
There is a degree of uncertainty in forecasting solar radiation, so the researcher resorted to what he called a fuzzy-logic-based system to overcome this.
Unlike digital systems that deal with only two values (1 or 0), “fuzzy systems” analyze all possible values between 1 and 0. It uses historical data from eight research weather stations in Saudi Arabia, counting in the sum of direct incident sunlight and diffused sunlight, including other parameters such as temperature, wind speed and humidity.
During testing, the model was able to reliably predict availability of solar radiation at specific locations, up to 24 hours later. Almaraashi claims it can be used anywhere in the Middle East.
Jr-Hau He at the KAUST Solar Centre of King Abdullah University of Science and Technology, who’s not involved in this study, pointed to the potential of this tool, calling it both flexible and reliable.
The proposed model is not restricted to those locations mentioned in the study but should allow for predicting solar radiation at other sites in Saudi Arabia, says He.
- Almaraashi, M. Short-term prediction of solar energy in Saudi Arabia using automated-design fuzzy logic systems. PLOS One. https://doi.org/10.1371/journal.pone.0182429 (2017)