To provide useful insights in the field of solar analytics using ground-based cameras and other weather sensors.
This is a collaborative work between:
- Nanyang Technological University Singapore
- Advanced Digital Sciences Center (ADSC), Singapore
- Copernicus Institute of Sustainable Development, Utrecht University, The Netherlands.
Owing to the growing concern of global warming and over-dependence on fossil fuels, there has been a huge interest in last years in the deployment of Photovoltaic (PV) systems for generating electricity. The output power of a PV array greatly depends, among other parameters, on solar irradiation. However, solar irradiation has an intermittent nature and suffers from rapid fluctuations. This creates challenges when integrating PV systems in the electricity grid and calls for accurate forecasting methods of solar irradiance.
In this research theme, we attempt to use a multi-modal approach of using data from various sensors to better understand such fluctuations in solar irradiance. We use time-series data of measured solar irradiance, together with clear-sky solar irradiance, to forecast solar irradiance upto a period of 20 minutes. We use techniques derived from time-series theory, to model the seasonality of solar irradiance, as captured in solar sensors.
Prediction of solar irradiance for shorter lead time.
In addition to using weather data, we also derive solar analytics information from images, captured via ground-based sky cameras. Unlike solar pyranometers and other regular meteorological sensors, ground-based sky images have additional information about the continuous evolution of cloud over time. We use these cloud/sky images to propose a solar radiation estimation model, that can accurately capture the short-term fluctuations of solar irradiance. We also use these sequence of images to estimate cloud motion fields. These motion fields are derived from optical flow formulation of two successive image frames. We compute the horizontal and vertical translation of pixels, and thereby predict future locations of cloud with a lead time of a few minutes.
Using two successive image frames, we compute the optical flow fields, and thereby predict the future image frames.
Furthermore, such multi-modal data can be useful for the purpose of detecting rainfall onset. We use ground-based sky cameras to detect the onset of rainfall. Using the luminance of the camera image as an indicator, we can accurately identify the precipitation onset. We define Clearness Luminance Index as the ratio of measured luminance value to the clear-sky luminance value.
We observe that the index value gradually decreases as it approaches a rain event, and reaches a minimum during rainfall. Post rain event, the clearness luminance index gradually increases with the passage of time. We also plot the Cumulative Distribution Plot (CDF) of clearness luminance index of images within and outside the time window of ±15 minutes.
Please refer to the publications.