Dataset For Pix2pix CSP Simulation

This dataset is designed to advance research on predicting solar flux density distribution in Concentrated Solar Power (CSP) tower plants, offering a comprehensive foundation for optimizing performance and ensuring stable operations. Directly measuring solar irradiance at the receiver site is highly challenging due to the extreme temperatures involved. To address this, the dataset enables the development of predictive models using indirect, data-driven approaches.

The dataset comprises the following components:

1. Calibration Images: High-resolution images of concentrated solar spots formed by heliostats on a Lambertian target. These images are essential for training and validating deep learning models.
2. Monte Carlo Ray Tracing Results: Detailed simulations of solar flux density distributions generated through Monte Carlo ray tracing. These results serve as a guide for training Generative Adversarial Networks (GANs).
3. Actual Solar Spot Data: Real-world measurements of solar spots, used to validate and compare the accuracy of model predictions.
4. Feature Data: A collection of parameters, including heliostat position, angle, time, and weather conditions, which are critical inputs for predictive modeling.

Key Features:

• High Quality: The dataset has been meticulously cleaned and preprocessed to ensure consistency and reliability.
• Multimodal: It integrates image data with numerical simulations, providing a rich and diverse input for advanced modeling techniques.
• Scalability: The dataset is designed to be adaptable, allowing for extension to other heliostats or entire heliostat fields for broader applications.