360 Degree Solar Animal Repeller

What is sky images & photovoltaic power generation dataset?

To fill these gaps, we introduce SKIPP'D—a SKy Images and Photovoltaic Power Generation Dataset. The dataset contains three years (2017–2019) of quality-controlled down-sampled sky images and PV power generation data that is ready-to-use for short-term solar forecasting using deep learning.

Can a suncast model predict solar power generation?

The results show that the SUNSET nowcast model can effectively extract the information in the sky images and correlate it with the local PV panel generation. It can well approximate the sun angle equations in the sunny days with clear sky conditions and reasonably estimate the states of PV power generation under different cloudy conditions.

What are the future research directions for predicting PV power generation?

These two aspects of errors direct to the following future research directions: (1) generating probabilistic prediction rather than point prediction to deal with the uncertainty in the sudden changes of PV power generation; (2) developing models for better prediction of the cloud movement to deal with the systematic temporal lags in forecasting.

Where are the PV power generation data from?

The PV power generation data are from PV panel approximately 125 m away from the camera on the roof of the Jen-Hsun Huang Engineering Center at Stanford University, which are logged by Stanford Utility and shared to us.

Why is solar PV becoming a major source of power generation?

Solar PV is rapidly becoming a significant source of power generation. Fluctuations in solar power generation due to short-term events (like moving clouds) can have large impacts in areas with high solar PV penetration.

Can a CNN model measure solar irradiance?

The potential use of the nowcast model is to serve as an alternative to the traditional sensor measurement of solar irradiance or PV panel power output, which is generally expensive. A similar study by Jiang et al. (2020) has also examined the potential of using an end-to-end CNN model to estimate the state of solar irradiance. Table 5.1.