Gaussian process regression for thermal transport analysis in infrared imaging video bolometry

Area of Science:

• Plasma physics and diagnostics
• Infrared thermography
• Statistical modeling

Background:

• Accurate characterization of thermal diffusion and blackbody radiation is essential for infrared imaging video bolometry (IRVB).
• Inferring spatial distribution of plasma radiation relies on precise detector foil measurements.
• Existing methods may require temporal or spatial averaging, potentially limiting accuracy.

Purpose of the Study:

• To develop a novel inference framework for modeling blackbody radiation and thermal diffusion power densities.
• To apply Gaussian process regression to IRVB data for enhanced plasma radiation analysis.
• To validate the framework's reliability using both synthetic and experimental data.

Main Methods:

• Utilizing Gaussian process regression for modeling.
• Developing a new inference framework for power density estimation.
• Validating the model with synthetic and experimental infrared imaging video bolometry data.

Main Results:

• The proposed framework reliably models blackbody radiation and thermal diffusion power densities.
• Accurate results were obtained without the need for temporal or spatial averaging.
• Analysis identified the effects of noise level and foil material on performance.

Conclusions:

• The Gaussian process regression framework offers a robust method for IRVB data analysis.
• The approach provides reliable plasma radiation distribution inference.
• Limitations were identified, and strategies for future performance improvements were proposed.