Study on fast algorithm of neutron radiation field under complex terrain scenario based on ensemble learning approach

Area of Science:

• Nuclear physics and radiation transport
• Computational modeling and simulation
• Geospatial analysis and environmental science

Background:

• Traditional radiation transport simulations struggle with complex terrain scenarios for nuclear explosions.
• Understanding neutron and secondary gamma transport is crucial for nuclear explosion assessment.
• Digital Elevation Models (DEM) can characterize terrain features influencing radiation fields.

Purpose of the Study:

• To develop a rapid prediction algorithm for neutron radiation fields in complex terrains using ensemble learning.
• To analyze the impact of topographic features on neutron and secondary gamma transport.
• To validate the prediction model's performance against Monte Carlo (MC) simulations in realistic scenarios.

Main Methods:

• Ensemble learning approach applied to predict neutron radiation fields.
• Extraction of characteristic parameters from DEM to represent terrain influence.
• Construction of a training dataset using MC simulations of randomly generated terrain samples.
• Model validation using predictions for neutron flux, neutron tissue dose, and secondary gamma tissue dose.

Main Results:

• The prediction algorithm accurately estimated neutron flux and tissue doses in urban and mountainous terrains.
• Fast predictions showed good agreement with MC simulation results across various evaluation metrics.
• The developed models demonstrate preliminary engineering application value for radiation field prediction.

Conclusions:

• The ensemble learning-based prediction algorithm offers a viable alternative to traditional methods for complex terrain scenarios.
• The approach successfully characterizes terrain influence on radiation transport for nuclear explosions.
• Further research can build upon this work to enhance model generalization for diverse radiation environments.