Fast nuclide identification method based on hybrid dynamic Bayesian network.

Area of Science:

• Nuclear Physics and Engineering
• Radiation Detection and Measurement
• Applied Bayesian Statistics

Background:

• Accurate and efficient nuclide identification is critical for various applications.
• Traditional methods can be limited by speed and accuracy in complex environments.
• Detector output pulse analysis offers a rich data source for identification.

Purpose of the Study:

• To develop an efficient and precise nuclide identification method.
• To model detector output pulse events using hybrid dynamic Bayesian networks.
• To enhance nuclide identification speed and accuracy through probabilistic propagation and sequential testing.

Main Methods:

• Treated detector output pulse events as sequences of energy event models.
• Employed hybrid dynamic Bayesian network modeling based on detector energy spectrum response.
• Utilized Monte Carlo simulations and a probabilistic propagation algorithm for parameter refinement and sequential testing.

Main Results:

• Achieved >91.3% accuracy in single-nuclide identification with a 1:7 background-to-nuclide intensity ratio.
• Demonstrated high detection rates for 60Co (~100%) and 137Cs (>81%) in dual-nuclide scenarios with noise.
• Showcased excellent multi-nuclide identification rates (e.g., 99.6% for 133Ba, 84.4% for 137Cs) with low false alarm rates (<0.8%).

Conclusions:

• The proposed hybrid dynamic Bayesian network method is feasible for rapid nuclide identification.
• The method demonstrates robustness and high accuracy in diverse radiation detection scenarios, including noisy environments.
• Significant potential exists for this approach in real-time radiation monitoring and analysis.