Radiation Portal Monitoring for Moving Source Identification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing radiation portal monitoring systems face challenges in accurately identifying moving radiation sources due to issues such as varying shielding configurations and background radiation, which complicate the analysis of detection signals.
Innovation Solution
A method involving a radiation portal monitoring system that uses a machine learning model trained with pseudo-dynamic identification data, generated by modifying static identification data to simulate detection signal alterations caused by source movement, allowing for accurate identification of moving radiation sources.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a machine learning model is trained with static identification data, then the model can identify stationary radiation sources accurately, but it fails to accurately identify moving radiation sources due to detection signal alterations caused by movement
Solution Approach 1:
The patent applies preliminary action by pre-processing static identification data to simulate dynamic detection conditions before training the machine learning model. Pseudo-dynamic identification data is generated by introducing modifications representing detection signal alterations caused by source movement, allowing the model to learn from pre-simulated dynamic scenarios without requiring actual moving source data during training
Solution Approach 2:
The patent employs parameter changes by modifying the static identification data with specific transformations that represent detection signal alterations. These modifications include changes in signal intensity, timing patterns, and spectral characteristics that simulate the effects of source movement, thereby adapting the training data parameters to reflect dynamic conditions
2Reliability
If the radiation portal monitoring system uses conventional detection methods, then the system structure remains simple, but the system cannot accurately distinguish moving radiation sources from background radiation and false alarms
Solution Approach 1:
The patent introduces an intermediary element in the form of pseudo-dynamic identification data that bridges the gap between simple static training data and complex moving source detection requirements. This intermediary training data enables the machine learning model to achieve high detection reliability without requiring complex hardware modifications or multiple detection systems
Solution Approach 2:
The patent replaces mechanical or physical detection system complexity with computational complexity. Instead of using multiple detectors, complex shielding arrangements, or sophisticated mechanical scanning systems, the invention uses a machine learning model trained on pseudo-dynamic data to achieve accurate moving source identification, substituting physical system complexity with algorithmic processing
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
The method enhances the accuracy of identifying moving radiation sources by effectively training the machine learning model to handle dynamic detection scenarios, reducing false alarms and improving detection efficiency.
Implementation Method 1
a radiation portal monitor with a plurality of radiation detectors configured to detect ionizing radiation of the moving radiation source and to generate a detection signal responsive to detection of the ionizing radiation
Data Source
AI summary
A method for identifying a moving radiation source by a radiation portal monitoring system is described. The radiation portal monitoring system includes a radiation portal monitor with a plurality of radiation detectors configured to detect ionizing radiation of the radiation source and to generate a detection signal responsive to detection of the ionizing radiation, and at least one processor executing the steps of providing an identification machine learning model; receiving labelled static identification training data generated by radiation detection of a plurality of known static radiation sources; introducing to the static identification training data modifications representing detection signal alterations caused by radiation source movement through the radiation portal monitor to obtain pseudo-dynamic identification training data; training the identification machine learning model using the pseudo-dynamic identification training data; and identifying the moving radiation source from the detection signal using the trained identification machine learning model.