Fast neutron multiplex nuclear material checking system based on physical field encryption

Abstract

The invention provides a physical field encryption-based fast neutron multiplex nuclear material checking system, which is used for checking and verifying the authenticity of a device or a sample containing a nuclear material on the premise of not obtaining sensitive information. The system is composed of a neutron source, a random encryption mask and a scintillator detector array. The neutron source and the detector array are used for acquiring multiplex information of the sample, and the mask is located between the detector and the sample and is made of polyethylene. The voxel thickness of the mask is distributed randomly, multiple fission neutrons emitted by fast neutrons in an induced mode are attenuated randomly, the neutrons in the open hole part completely penetrate out, the neutrons in the solid part are partially absorbed, and the neutrons modulated by the mask are finally measured by a detector. And then the attenuated neutron multiplex information is coded into a time sequence signal, so that unique encrypted information of a checking object is formed, and checking is carried out by comparing the unique encrypted information with a known template. The verification system provided by the invention can perform identity authentication on the premise of not obtaining sensitive information, and can be applied to related fields of nuclear material encryption verification, military control, nuclear security and the like.

Description

Technical Field

[0001] The present invention relates to the field of encrypted radiation measurement, in particular to an encrypted detection system based on neutron multiplicity and mask random coding. Background Art

[0002] Nuclear material verification is an important technical means of nuclear safety and nuclear non-proliferation. The main content of verification is the tracking and monitoring of nuclear materials. With the continuous development of nuclear technology, verification not only needs to track known nuclear materials, but also needs to be able to cope with new materials and technologies, such as advanced nuclear fuels, recycled and reprocessed waste, and potential new nuclear materials.

[0003] Illegal acquisition of nuclear materials is one of the core threats of nuclear terrorism. The risk of illicit nuclear material acquisition has increased significantly. Illegal organizations could potentially acquire highly enriched uranium and plutonium to manufacture improvised nuclear weapons or conduct "dirty bomb" attacks, posing a significant threat to global security and undermining global stability. Therefore, the management and monitoring of nuclear materials has become a key task in international counter-terrorism and nuclear security. The International Atomic Energy Agency and the UN Security Council have begun strengthening their monitoring and verification of nuclear materials. Cross-border transfer, storage, and waste management of nuclear materials all require precise verification and oversight.

[0004] However, to prevent the leakage of sensitive data during verification, verification cannot directly reflect relevant information about nuclear materials and components (such as spatial and elemental composition). The measurement process must be encrypted, a process known as encrypted verification. The difficulty lies in verifying the identity of nuclear materials without obtaining sensitive information about them. Traditional verification techniques rely primarily on passive detection, such as gamma rays, alpha particles, and neutrons from the spontaneous decay of nuclear materials. Current verification systems typically combine multiple detection methods, such as neutron imaging, gamma spectroscopy, and electron beam scanning, to detect and identify verification targets from multiple dimensions.

[0005] The main challenge currently facing verification technology is how to ensure high verification reliability while improving security. The present invention proposes an encrypted verification technology route based on fast neutron multiplicity. By introducing a random mask to perform random physical encryption on the neutron signal in the physical field, and introducing algorithmic encryption for secondary encryption, it further improves security while ensuring verification reliability. Summary of the Invention

[0006] The present invention aims to propose a fast neutron multiplicity nuclear material verification system based on physical field encryption, which verifies the authenticity of devices or samples containing nuclear materials without obtaining sensitive information. The system consists of a neutron source, a random encryption mask, and a scintillator detector array.

[0007] The neutron source uses DT neutrons with an energy of about 14 MeV, located 15 cm above the center of the sample; the detectors are 8 scintillator detectors, and the geometric dimensions of each detector are 6×6×5 cm. 3 The front end of the detector is 20 cm away from the central axis of the detection system. The switch is a random mask with a hollow cylinder located between the detector and the sample, with an inner and outer diameter of 10 cm and

[0008] The hollow cylinder is 13 cm wide and 10 cm high, made of polyethylene. To randomly attenuate passing fission neutrons, the hollow cylinder is divided into several voxels. The radial thickness of each voxel is a random number between 0 and 3 cm. This random thickness randomly attenuates the multiple fission neutrons emitted by fast neutrons. Neutrons in open areas completely pass through, while neutrons in solid areas are partially absorbed. Neutrons modulated by the mask are ultimately measured by the detector. The attenuated neutron multiplicity information is then encoded into the time-series signal, forming a unique encrypted signature of the verification object, which is then compared with a known template for verification.

[0009] Neutron multiplicity refers to the number of neutrons produced during a nuclear fission or reaction. The number of neutrons produced in each fission process is not fixed but follows a specific statistical distribution, such as the Poisson distribution or a more complex probability distribution model. Fission neutrons are measured using scintillator detectors. Due to the presence of a significant gamma-ray background, a scintillator (plastic or liquid) with neutron and gamma discrimination is required. Based on the differences in the light pulse characteristics produced by the detector in response to the incident particle, scintillator detectors can accurately distinguish between neutron and gamma signals. Neutrons and gamma rays interact differently in matter, and scintillator detectors can exploit this difference for discrimination. When a neutron is incident, the resulting scintillation light pulse has a larger slow component and a longer pulse decay time. In contrast, gamma rays primarily produce a fast component with a relatively smaller slow component. This difference in pulse shape forms the basis for pulse shape discrimination.

[0010] The pulse signals generated by the detector are sampled as digital signals by the electronics system. Constant fraction flow (CFD) is used to precisely mark the triggering moment of each signal with picosecond accuracy. This time, called the signal's timestamp, indicates the moment the signal arrives at the system. All signals are recorded in chronological order, including the pulse waveform and trigger time. Neutron / gamma discrimination is performed by analyzing the pulse shape, and coincidence measurement is then performed based on the timing of the neutron signal.

[0011] Based on the arrival time distribution of fission neutrons, a time gate (typically tens of nanoseconds) is set, and the neutron signals are sorted chronologically. Starting with the first signal, the number of signals within the time gate is counted, recording the double coincidence count (D) and triple coincidence count (T). This process is then repeated, using each neutron signal as the starting time, until all signals have been counted. The final result is the total double coincidence count (D), the total triple coincidence count (T), and the total neutron count (S).

[0012] By adjusting the position of the encryption physical mask at equal time intervals—that is, changing the voxel structure of the effective encryption mask facing the neutron detector—and repeating the above measurement, the neutron multiplicity count distribution that varies with time (i.e., physical mask displacement) can be obtained. Each measurement is fixed in time, and the mask is randomly translated in the z-axis for the next measurement to ensure that the neutron beam passes through a different mask each time. A total of 100 measurements are performed to obtain a time series signal of neutron multiplicity for the final encryption verification.

[0013] To effectively compare the reference sample with the sample under test, we need to define a metric c to quantify the difference between the two. This metric not only reflects the degree of similarity between the sample under test and the reference sample, but also measures the bias and uncertainty of the sample under test to a certain extent. The metric c is defined as follows:

[0014]

[0015] where y obs is the measured value of the sample to be tested, μ(y exp ) is the mean of the template sample, σ(y exp ) is the standard deviation of the template sample. Specifically, the sample to be tested y obs By comparing the mean μ(y exp ) is used for comparison and to measure its deviation; at the same time, the deviation is divided by the standard deviation σ(y exp ) is used to eliminate the differences caused by the different fluctuation ranges of the reference samples and ensure that the scale of the measurement values is unified.

[0016] The metric value c is essentially a standardized error value. It reflects the deviation between the measured value of the test sample and the reference template sample. This deviation is normalized by the standard deviation, allowing different samples to be compared on the same scale. A smaller metric value c indicates that the test sample is more similar to the reference template sample; conversely, a larger metric value c indicates a significant difference between the two.

[0017] The overall encryption verification method uses the Receiver Operating Characteristic Curve (ROC) to evaluate its reliability. The ROC curve reflects the judgment results of different thresholds under the same stimulus. The horizontal axis of the ROC characteristic curve is the false positive probability (False Positive Rate, FPR), that is, the ratio of the number of false samples judged as true samples to the number of all false samples, and the vertical axis is (True Positive Rate, TPR), that is, the ratio of the number of true samples judged as true samples to the number of all true samples. It can relatively simply derive the recognition ability of any judgment threshold. The closer the curve is to the upper left corner, the higher the accuracy of the system. The point closest to the upper left corner is the optimal judgment threshold, and its proportion of false positives and false negatives is the smallest. AUC (Area Under Curve) is the area enclosed by the coordinate axis under the ROC curve. It is an important indicator to measure the system. Obviously, its maximum value is 1. When the AUC value is greater than 0.9, it proves that the system has excellent resolution.

[0018] In order to verify the effectiveness of the above metrics in classification and discrimination, an evaluation method based on the ROC curve was further constructed. The c values representing the true samples and the c values representing the false samples were uniformly mapped to confidence scores:

[0019] score=1-c, (2)

[0020] The higher the score, the more likely the sample is to be a true sample.

[0021] Then construct a binary classification label (true sample is 1, false sample is 0), calculate the FPR (false positive rate) and TPR (true positive rate) corresponding to the score, draw the ROC curve, and calculate its AUC value as the classification performance indicator. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] The accompanying drawings are used to provide further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention but do not constitute a limitation of the present invention.

[0023] Figure 1 This is a schematic diagram of three-dimensional encryption verification. DETAILED DESCRIPTION

[0024] The present invention proposes an encrypted detection system based on neutron multiplicity and mask random coding, which aims to verify the authenticity of samples without obtaining sensitive information. The specific implementation method is as follows:

[0025] 1. System Configuration and Installation

[0026] The encrypted detection system consists of three parts: a neutron source, an encryption mask, and a scintillator detector array. The neutron source uses a neutron beam with an energy of 14 MeV and is located 15 cm above the center of the sample to be tested, ensuring that the neutron beam is evenly irradiated on the sample surface. The detector array consists of eight scintillator detectors with a geometric size of 6 × 6 × 5 cm. 3 , and the front end of each detector is 20 cm away from the central axis of the system.

[0027] 2. Mask Design and Layout

[0028] The mask is a hollow cylinder made of polyethylene with an inner diameter of 10 cm, an outer diameter of 13 cm, and a height of 10 cm. It is located between the detector and the sample. The mask is divided into a number of voxels, each with a random radial thickness between 0 and 3 cm. This random mask design effectively modulates the neutron beam's path, producing unpredictable changes as the beam passes through the sample, providing critical support for sample encryption.

[0029] 3. Signal Sampling and Neutron / Gamma Identification

[0030] A 14 MeV neutron beam emitted by a neutron source first passes through a cryptographic mask. The hollow portions of the mask allow the beam to pass completely, while the solid portions partially absorb neutrons, forming a modulated neutron beam. The modulated neutron beam then strikes the sample under test, inducing fission reactions in fissile nuclides. The fission neutrons produced by these (n, f) reactions are received and recorded by a surrounding array of scintillator detectors. Each detector generates a scintillation light pulse upon receiving a fission neutron. The detector output pulse signal is sampled and converted into a digital signal by the electronics system. Constant fraction flow (CFD) is used to precisely mark the triggering moment of each pulse signal, achieving picosecond-level timing accuracy. All signals are recorded in chronological order, including the pulse waveform and trigger time. By analyzing the pulse waveforms, the system can accurately distinguish neutron and gamma-ray signals. Neutron signals are characterized by a large slow component, while gamma rays are characterized by a shorter decay time. Based on these differences in pulse shape, the system uses pulse shape discrimination to effectively distinguish neutron and gamma-ray signals.

[0031] 4. Multiplicity Measurement

[0032] First, a standard sample is determined as a template against which the sample to be tested is compared for authentication. During each measurement, the mask is randomly translated in the Z-axis to ensure that the neutron beam passes through a different mask each time. Using the detector neutron count as a fingerprint, multiple measurements are performed to generate multiple images of the sample, creating a unique fingerprint for the object under test and verifying its authenticity.

[0033] The specific test steps are as follows:

[0034] (1) Measure the template sample.

[0035] (2) Measure the sample to be tested.

[0036] (3) Normalize the measurement time.

[0037] (4) Calculate the metric value and ROC curve.

[0038] 5. Sample Fingerprint Generation and Verification

[0039] After multiple measurements, the neutron detector counting results are statistically analyzed to generate a neutron multiplicity image of the sample to be tested. This image serves as the "fingerprint" of the sample to be tested, is unique, and is directly related to the authenticity of the sample. By comparing it with the reference sample, the measurement value c is calculated using formula (1):

[0040]

[0041] where y obs is the measured value of the sample to be tested, μ(y exp ) is the mean of the template sample, σ(y exp ) is the standard deviation of the template sample. Based on the calculated metric value c, the similarity between the test sample and the reference sample is quantified, and then the authenticity verification is performed.

[0042] 6. ROC Curve and Classification Evaluation

[0043] The classification performance of the system can be further evaluated by constructing the ROC curve. The metric value c is converted into a confidence score.

[0044] score=1-c (2)

[0045] The false positive rate (FPR) and true positive rate (TPR) are calculated. Based on different judgment thresholds, the ROC curve is plotted and the area under the curve (AUC) is calculated as an indicator of classification performance. The closer the AUC value is to 1, the higher the classification accuracy of the system.

[0046] Through the above implementation method, the present invention can effectively verify the authenticity of the sample and provide a safe and efficient detection method by using neutron multiplicity and mask random coding technology without obtaining sensitive information.

Claims

1. A fast neutron multiplicity nuclear material verification system based on template comparison physical field encryption, characterized by: a. The system consists of a neutron source, a physical encryption mask, and a scintillator detector array. The neutron source and detector array in the system can obtain neutron multiplicity information of the sample to be tested; b. The neutron source energy is 14 MeV and is located 15 cm above the center of the sample. The detectors are 8 scintillator (plastic or liquid) detectors with neutron and gamma discrimination. The geometric dimensions of each detector are 6 × 6 × S cm 3 , the front end of the detector is 20cm away from the central axis of the detection system; c. The random mask is located between the detector and the sample. It is a hollow cylinder made of polyethylene with an inner diameter of 10 cm, an outer diameter of 13 cm, and a height of 10 cm. The hollow cylinder is divided into a number of voxels, each with a random thickness ranging from 0 to 3 cm in the radial direction. Voxels with a thickness of less than 0.1 cm are hollowed out. d. Physical encryption of the neutron multiplicity time series signal. After each neutron multiplicity measurement, the mask is randomly translated in the z-axis direction. Repeating the above measurement can obtain the neutron multiplicity count distribution that changes with time (i.e., the physical mask displacement), forming the unique encrypted information of the verification object; e. Encrypted verification based on template comparison: By comparing the physical encrypted neutron multiplicity timing signals of the measured object with those of a known template, it is possible to verify whether the measured object is consistent with the template.

2. The template comparison-based encryption verification method according to claim 1, characterized in that: a. First, a reference sample (i.e., a confirmed true sample) needs to be determined and measured to obtain the physically encrypted neutron multiplicity time series signal of the reference sample. b. Measure the sample under test under the same experimental conditions, obtain its physically encrypted neutron multiplicity time series signal, and calculate the difference metric value c between the sample under test and the reference sample. The metric value c is defined as follows: where y obs is the measured value of the sample to be tested, μ(y exp ) is the mean of the template sample, σ(y exp ) is the standard deviation of the template sample; c. Set an optimal threshold. When the metric value c is greater than the threshold, the sample to be tested is judged to be false. When the metric value c is less than or equal to the threshold, the sample to be tested is judged to be true. d. The optimal threshold is determined using the following steps. The ROC curve is plotted based on the calculated metrics. When the AUC value is greater than 0.9, the system's resolution is excellent. The optimal threshold is calculated when the ROC curve is plotted. Based on this threshold, the multiplicity information of the test sample and the template sample is compared to authenticate the identity.

Images