A Method and System for Measuring the Activity Gamma Scan of Low- and Intermediate Radioactive Waste Cans Based on Neural Networks

Abstract

This invention provides a method and system for measuring the activity of low- and intermediate-level radioactive waste bins using gamma scanning based on neural networks, comprising: Step S1: The waste bin is vertically divided into several segments; Step S2: Radioactive point sources within each segment are assumed to be ring-shaped line sources, and all ring sources within the same segment are equivalent to one equivalent ring source relative to the detector; Step S3: A segment is scanned at the same height and three different eccentricity positions using a gamma spectrometer to obtain the corresponding count rates at the three detection positions; Step S4: The equivalent ring source radius and its detection efficiency are calculated using the count rates at the three detection positions combined with a neural network algorithm; Step S5: The nucleon activity of the current segment is calculated based on the equivalent ring source radius and its corresponding detection efficiency; Steps S1 to S5 are repeated to obtain the nucleon activity of each segment; Step S6: The total activity of the waste bin is obtained by summing the nucleon activities of each segment.

Description

Technical Field

[0001] This invention relates to the field of waste bin activity measurement technology, specifically to a method and system for measuring low- and intermediate-level radioactive waste bin activity using gamma scanning based on neural networks. Background Technology

[0002] With the development of nuclear power and nuclear technology, large quantities of low- to intermediate-level radioactive waste are generated and processed into solid drums for disposal. According to radioactive waste management regulations, before final disposal, the surface dose of the waste drums, the radionuclides inside, and their activity must be measured to meet the requirements of safe handling and classified disposal. Furthermore, with national requirements for reducing the volume of radioactive waste, the waste is compressed and packaged in large-volume drums; for example, a 200L initial container is compressed and then placed into a 400L metal drum. This results in high-density, large-volume waste drums, making accurate measurement of the waste even more difficult.

[0003] Measurements of radioactive waste containers are generally performed using non-destructive testing techniques, meaning that the activity of radionuclides within the container can be measured without damaging it. This includes gamma scanning technology. Currently, nuclear power plants commonly use segmented gamma scanning (SGS) technology and corresponding equipment for measuring radioactive waste containers. SGS technology was proposed in the 1970s by Los Alamos National Laboratory (LANL) in the United States. It assumes that the filling medium and radionuclides within the waste container are uniformly distributed, thus establishing a quantitative relationship between the activity of a uniformly distributed radioactive source and the detector's gamma photon count rate, enabling the measurement of radioactivity. However, when the radionuclides are not uniformly distributed within actual waste containers, it can cause significant errors in activity measurement. To address this, in the 1990s, LANL also proposed tomographic gamma scanning (TGS) technology. This technology divides the waste container into numerous block grids and reconstructs the spatial distribution of the medium density and radionuclides along the grid using tomographic techniques, thereby achieving relatively accurate measurements of non-uniformly distributed waste containers. However, this technology requires a sufficient number of measurements, such as scanning the waste bin process from different angles, eccentric positions, and heights, to meet the requirements of chromatographic reconstruction. Therefore, compared with SGS, the measurement process is more complex and the measurement time is dozens of times longer, making it difficult to meet the needs of practical applications.

[0004] To address these challenges and achieve rapid and accurate measurements, several new or improved technologies have been proposed. One such technology is the Improved Segmented Gamma Scanning (ISGS) method, which involves rotating the waste container at a constant speed, dividing it vertically into segments. Assuming a uniform distribution of the filling medium within each segment, the point sources within each segment are equated to ring sources. Based on the ring source activity, these are further equated to a single ring source. The radius of this equivalent ring source is reconstructed using two detectors, establishing the relationship between the ring source activity and the detector gamma count (detection efficiency) for radioactivity reconstruction. This method accurately determines the ring source radius through the ring source assumption. Compared to SGS, it takes approximately twice as long and can achieve relatively accurate measurements of radioactive waste with non-uniformly distributed nuclides. However, this method requires ensuring that the ratio of the count rates of the two detectors is monotonically related to the ring source radius. Therefore, extensive calculations are needed beforehand to determine the relative distance between the two detectors, which is insufficient for measuring large quantities of waste containers in practice.

[0005] In addition, to address the issue of long detection time in TGS (Total Gamma Scanning), semi-tomographic gamma scanning (STGS) has been proposed. This technique involves rotating the waste container at a constant speed during measurement. Under homogeneous conditions, the point source is equivalent to a ring source, thus transforming the distribution of the radioactive source from three-dimensional to two-dimensional, i.e., distributed along the radius and height directions. Since it eliminates the need to reconstruct the circumferential distribution and to divide the area into grids, the number of tomographic reconstruction grids is significantly reduced, as are the number of detections and the time required. This method can measure waste containers with large volumes and high densities; however, it typically requires dividing the area into four or more ring-shaped grids in the radial direction to obtain a relatively accurate reconstruction of the radionuclide distribution along the radius. Therefore, the measurement time is longer than that of the dual-detector segmented gamma scanning method.

[0006] Patent document CN115685299A (application number: 202211297254.5) discloses a gamma scanning method, system, device, and storage medium for radioactive waste bins. The method includes the following steps: dividing the waste bin into at least two equal segments, and each segment into at least two primary screening regions; receiving detectors measuring the count rate of the waste bin at the corresponding measurement points in each primary screening region; performing primary screening reconstruction based on the count rate of the corresponding measurement points in each primary screening region and the pre-obtained detection efficiency of each primary screening region at the measurement points, and obtaining the activity reconstruction value of each primary screening region; and summing the activity reconstruction values ​​of each primary screening region to obtain the total activity of the waste bin.

[0007] Therefore, it is necessary to further develop a rapid and high-precision measurement method for large-volume, high-density compressed waste bins, especially to study the problems existing in the current technology and make reasonable improvements to meet the measurement requirements of the new waste bins. Summary of the Invention

[0008] To address the shortcomings of existing technologies, the purpose of this invention is to provide a method and system for measuring the activity of low- and intermediate-level radioactive waste bins using a neural network-based gamma scanning method.

[0009] A method for measuring the activity gamma scan of low-to-intermediate radioactive waste bins based on a neural network, according to the present invention, includes:

[0010] Step S1: When measuring the waste bin, the medium inside the bin is assumed to be uniformly distributed, and the waste bin is divided into several segments in the vertical direction;

[0011] Step S2: The radioactive point sources within each segment are assumed to be ring-shaped line sources, and the radius of each ring source is the horizontal distance from the corresponding point source to the center of the barrel axis; all ring sources within the same segment are equivalent to one equivalent ring source relative to the detector.

[0012] Step S3: Using a gamma spectrometer, a segment is scanned at the same height and three different eccentricity positions. Through spectral analysis, the corresponding nuclide is identified based on the characteristic energy of the full-energy peak, and the corresponding count rate at the three detection positions is obtained.

[0013] Step S4: Use the count rate at the three detection locations combined with a neural network algorithm to solve for the equivalent ring source radius and its detection efficiency;

[0014] Step S5: Calculate the kernel activity of the current segment layer based on the equivalent ring source radius and its corresponding detection efficiency; repeat steps S1 to S5 to obtain the kernel activity of each segment layer.

[0015] Step S6: Sum the activity of the kernels in each segment layer to obtain the total activity of the waste bin.

[0016] Preferably, step S2 involves: dividing the waste bin into multiple segments along its axial direction; scanning the waste bin with a high-purity germanium detector equipped with a collimator; rotating the waste bin at a constant speed; and treating a point source at a certain radius within the bin as an equivalent to a ring-shaped line source at that radius. It is assumed that there are N point sources of the same type located at different radii within a certain segment, which are equivalent to N ring sources with different radii.

[0017] Preferably, step S4 employs:

[0018] Step S4.1: Assume the detector measures a segment of the layer at locations A, B, and C, with count rates C... A C B C C ;

[0019]

[0020]

[0021]

[0022] Where α is the probability of emitting gamma rays of that energy when a nuclide decays; A represents the activity; E(r A ) indicates that the detector is for a radius of r A The detection efficiency of the loop source; Δr represents the equivalent loop source radius when the detector is at position B, and r A The difference; ΔR represents the equivalent loop source radius when the detector is at position C and r A The difference;

[0023] Step S4.2: Based on the count rate ratio of detectors at different measurement locations, the equivalent loop source radius is obtained using a neural network algorithm;

[0024]

[0025] Preferably, the detection efficiency of the equivalent ring source is calculated using the non-circuit efficiency scaling method based on the radius of the equivalent ring source. A ).

[0026] Preferably, step S5 employs the following methods:

[0027]

[0028] A neural network-based gamma scanning measurement system for low-to-intermediate radioactive waste bin activity, provided by the present invention, includes:

[0029] Module M1: When measuring the waste bin, the medium inside the bin is assumed to be uniformly distributed, and the waste bin is divided into several segments in the vertical direction;

[0030] Module M2: Radioactive point sources within each segment are assumed to be ring-shaped line sources, with the radius of each ring source being the horizontal distance from the corresponding point source to the center of the barrel axis; all ring sources within the same segment are equivalent to one equivalent ring source relative to the detector.

[0031] Module M3: Using a gamma spectrometer, a segment is scanned at the same height and three different eccentricity positions. Through spectral analysis, the corresponding nuclides are identified based on the characteristic energy of the full-energy peak, and the corresponding count rates at the three detection positions are obtained.

[0032] Module M4: Utilizes the count rates at three detection locations combined with a neural network algorithm to solve for the equivalent loop source radius and its detection efficiency;

[0033] Module M5: Calculate the kernel activity of the current segment layer based on the equivalent ring source radius and its corresponding detection efficiency; repeat modules M1 to M5 to obtain the kernel activity of each segment layer.

[0034] Module M6: Summing the activity of kernels in each segment layer to obtain the total activity of the waste bin.

[0035] Preferably, module M2 employs the following method: dividing the waste bin into multiple segments along its axial direction, scanning the waste bin with a high-purity germanium detector equipped with a collimator, rotating the waste bin at a uniform speed, and equating a point source at a certain radius within the bin to a ring-shaped line source at that radius; assuming that there are N point sources of the same type located at different radii within a certain segment, these are equivalent to N ring sources with different radii.

[0036] Preferably, the module M4 adopts:

[0037] Module M4.1: Assume the detector measures a segment of the layer at locations A, B, and C, with count rates C... A C B C C ;

[0038]

[0039]

[0040]

[0041] Where α is the probability of emitting gamma rays of that energy when a nuclide decays; A represents the activity; E(r A ) indicates that the detector is for a radius of r A The detection efficiency of the loop source; Δr represents the equivalent loop source radius when the detector is at position B, and r A The difference; ΔR represents the equivalent loop source radius when the detector is at position C and r A The difference;

[0042] Module M4.2: The equivalent loop source radius is obtained using a neural network algorithm based on the count rate ratio of detectors at different measurement locations;

[0043]

[0044] Preferably, the detection efficiency of the equivalent ring source is calculated using the non-circuit efficiency scaling method based on the radius of the equivalent ring source. A ).

[0045] Preferably, the module M5 adopts:

[0046]

[0047] Compared with the prior art, the present invention has the following beneficial effects:

[0048] 1. In reconstructing the radionuclide activity from measurement data, this invention assumes that the radioactive source is equivalent to a ring source. All ring sources equivalent to the uniform rotation of the waste container are then further equivalent to a single "equivalent ring source." The radius of the "equivalent ring source" is directly output using the count rates at three detection positions combined with a neural network model, thereby calculating the accurate detection efficiency and ultimately achieving accurate measurement of the radioactive waste container activity. This method increases the number of detectors and uses three data points to determine the radius of the "equivalent ring source," thus allowing for determination even when the detector count rate ratio curve is not monotonic. Compared to ISGS, this method eliminates the need for extensive pre-calculation to determine the relative eccentricity of the detectors, resulting in a wider range of practical applications and simpler, more convenient operation.

[0049] 2. This invention directly outputs the radius of the "equivalent loop source" through the constructed neural network model. Compared with ISGS, which uses interpolation to obtain the radius of the "equivalent loop source", this invention can determine the "equivalent loop source" more accurately and quickly.

[0050] 3. The gamma scanning method proposed in this invention can achieve a relatively accurate measurement of the radioactivity of large-volume, high-density waste bins formed after overpressure. Since the equivalent ring source radius is calculated first and then the efficiency is calibrated, compared with SGS, TGS, STGS and other methods that calibrate the efficiency based on the artificially set nucleus distribution in the waste bin, the accuracy of this invention is higher than SGS and STGS, reaching the level of TGS. The measurement time is three times that of SGS and 1 / 20 of that of TGS. Attached Figure Description

[0051] Other features, objects, and advantages of the present invention will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings:

[0052] Figure 1a and Figure 1b This is a schematic diagram showing how a point source can be equivalent to a ring source when the waste bin rotates at a constant speed.

[0053] Figure 2 For the nuclide Cs-137 in a medium with a density of 1 g / cm³ 3 The distribution curve of detection efficiency as a function of the ring source radius.

[0054] Figure 3 This is a schematic diagram of the measurement (top view).

[0055] Figure 4 A schematic diagram illustrating how to quickly output the equivalent ring source radius using a neural network.

[0056] Figure 5This is a schematic diagram showing the count rate ratios of different detectors. Detailed Implementation

[0057] The present invention will now be described in detail with reference to specific embodiments. These embodiments will help those skilled in the art to further understand the present invention, but do not limit the invention in any way. It should be noted that those skilled in the art can make several changes and improvements without departing from the concept of the present invention. These all fall within the protection scope of the present invention.

[0058] Example 1

[0059] This invention addresses the problems of existing dual-detector segmented gamma scanning methods by proposing a neural network-based gamma scanning measurement method and system for low- and intermediate-level radioactive waste containers. This method enables relatively accurate measurement of radionuclide activities in overpressured, high-density waste in 400L metal containers. The measurement accuracy is significantly higher than that of SGS and STGS scanning methods, the measurement time is much shorter than that of TGS, and it eliminates the need for extensive calculations to determine the relative distance between detectors. This solves the problem that existing gamma scanning methods cannot simultaneously guarantee high measurement accuracy and short measurement time.

[0060] According to the present invention, a gamma scanning method for measuring low- to medium-level radioactive waste bins includes: During waste bin measurement, the medium inside the bin is assumed to be uniformly distributed; the waste bin is vertically divided into several segments; the waste bin rotates at a constant speed; and the radioactive point sources within each segment are assumed to be annular line sources, with the radius of each annular source being the horizontal distance from the corresponding point source to the bin's axis. All annular sources within the same segment can be equivalent to one "equivalent annular source" relative to the detector. A gamma spectrometer (high-purity germanium detector and matching multichannel analyzer) scans one segment at the same height and three different eccentricity positions. Through spectral analysis, the corresponding nuclides are identified based on the characteristic energy of the full-energy peak, and the count of gamma rays absorbed by the detector crystal per unit time is obtained. The radius of the "equivalent annular source" is obtained using the count rate at the three detection positions combined with a neural network algorithm, its detection efficiency is calibrated, and the nuclide activity of the segment is reconstructed based on the measured gamma ray count rate. The above steps are repeated to scan each segment from bottom to top using a gamma spectrometer, calculating the activity of each segment separately. The sum of the nuclide activities of all segments gives the activity of the waste bin. Compared with traditional gamma scanning methods, this invention does not require assumptions about the nuclide distribution within the waste bin, thus better reflecting reality. Furthermore, because the equivalent ring source radius is calculated first before efficiency calibration, the calculated detection efficiency value of this invention is more accurate, thereby improving the precision of activity reconstruction.

[0061] More specifically, the waste bins were measured, as shown in Figure 1, where, Figure 1a Point sources distributed vertically; Figure 1b This is to be equivalent to a ring source within each segment. Specifically, it includes:

[0062] The waste bin is divided into multiple segments along its axial direction. A high-purity germanium (HPGe) detector with a collimator scans the waste bin, which rotates at a constant speed. A point source at a certain radius within the bin can be equivalently represented as a ring-shaped line source at that radius. An actual waste bin may contain multiple nuclides, and each nuclide may emit rays of various energies. To illustrate the reconstruction principle, we will now analyze the rays of a specific characteristic energy emitted by a particular nuclide.

[0063] Suppose there are N point sources of the same type located at different radii within a fault, which can be equivalent to N ring sources of different radii. When measuring this fault, for gamma rays with energy e emitted by the decay of a nuclide, the detector count rate is...

[0064]

[0065] Where α is the probability that a nuclide emits gamma rays of that energy when it decays, A n E represents the activity of the nth nuclide. n Let be the detector efficiency for detecting gamma rays with energy e emitted by the nth nuclide. Assume the total activity of this type of nuclide within the fault is A, and the activity of the nth nuclide is An. n The ratio of ε to total activity A is n Then the detector's counting rate can be expressed as:

[0066]

[0067] All ring sources in a plane are equivalent to a single ring source. For E n Perform a weighted average, with weights ε. n The result is E′. Therefore, equation (2) can be expressed as...

[0068] C=αAE'(r') (3)

[0069] like Figure 2 As shown, due to E n As the radius r of the ring source monotonically increases, there exists a radius r′ such that the detector's detection efficiency for the ring source at that location is E′. In this case, all ring sources within the fault are equivalent to a single ring source, and the activity of this equivalent ring source is equal to the sum of the activities of all ring sources.

[0070] Solve for the equivalent ring source radius. For example... Figure 3 As shown, assuming the detector measures a fault at locations A, B, and C, with count rates C0 and C1 respectively. A C B C C According to equation (3), we have

[0071]

[0072]

[0073]

[0074] From equations (4), (5), and (6), we can see that

[0075]

[0076] The count rate ratio of detectors at different locations has a functional relationship f with the equivalent loop source radius. This functional relationship is very complex and difficult to obtain by general methods.

[0077] like Figure 4 As shown, a model is established using neural network technology, with the count rate ratio of detectors at different measurement positions as input and the radius of the equivalent loop source as output. A mapping relationship f between the input and output is established, thereby accurately and quickly outputting the radius of the equivalent loop source.

[0078] Based on the ring source radius r A The detection efficiency E'(r) of the equivalent ring source was accurately calculated using the unwired efficiency scaling method. A The radioactivity A of this segment is calculated according to the following formula.

[0079]

[0080] Perform steps 1 to 4 for each segment, and sum the activity values ​​of each segment to obtain the total activity value of the waste bin.

[0081] The present invention also provides a neural network-based gamma-ray scanning measurement system for the activity of low-to-medium radioactive waste bins. The neural network-based gamma-ray scanning measurement system for the activity of low-to-medium radioactive waste bins can be implemented by executing the process steps of the neural network-based gamma-ray scanning measurement method for the activity of low-to-medium radioactive waste bins. That is, those skilled in the art can understand the neural network-based gamma-ray scanning measurement method for the activity of low-to-medium radioactive waste bins as a preferred embodiment of the neural network-based gamma-ray scanning measurement system for the activity of low-to-medium radioactive waste bins.

[0082] Example 2

[0083] Example 2 is a preferred example of Example 1.

[0084] The activity gamma scan measurement method for low-to-intermediate radioactive waste bins based on neural networks provided by this invention, along with other gamma scan methods, were simulated and measured using Monte Carlo software. Activity reconstruction was performed on a 400L metal bin with a radius of 35cm. The internal medium of the waste bin was set to cement with a density of 2.3g / cm³. 3 The nuclide selected is Co-60, a common nuclide found in nuclear power plant waste bins.

[0085] Fifty nuclides were randomly generated in each waste bin, corresponding to the single-ring source case. Then, ten groups were randomly generated, each containing five nuclides in the waste bin, corresponding to the case where multiple nuclides were present in the waste bin.

[0086] like Figure 5 As shown, C1, C2, C3, C4, and C5 represent the count rates of the detector at eccentricities of 0, 7, 14, 21, and 28 cm, respectively. Since the count rate ratio curves no longer monotonic with the ring source radius, the ISGS method is no longer applicable under this condition.

[0087] When using TGS and STGS for activity reconstruction, TGS is divided into 60 grids and STGS is divided into 3 grids.

[0088] The activity reconstruction error is:

[0089]

[0090] Where Ar is the reconstructed value of the nuclide activity, and A is the true value of the nuclide activity.

[0091] The average error, maximum error, and standard deviation of the error for activity reconstruction using different methods were calculated to compare the accuracy of the different methods. The results are shown in Tables 1 and 2.

[0092] Table 1 Results of Single-Source Activity Reconstruction

[0093]

[0094] Table 2 Results of Multi-Source Activity Reconstruction

[0095]

[0096] Assuming the measurement time of the detector at each measurement location is a dimensionless unit 1, the total measurement time for different methods is shown in Table 3.

[0097] Table 3 Time Measurement Methods

[0098]

[0099] As can be seen, for the case of a single-ring source, the measurement accuracy of this method reaches that of TGS. Compared with SGS and STGS, the accuracy of the new method is improved by an order of magnitude. When there are multiple sources in the waste bin, the average relative error of the new method is only half that of SGS and STGS, and close to that of TGS.

[0100] This method achieves accuracy far exceeding that of SGS and STGS, reaching the accuracy of TGS, while its measurement time is only 1 / 20th that of TGS. It significantly reduces measurement time while maintaining high measurement accuracy.

[0101] Those skilled in the art will understand that, besides implementing the system and its various devices, modules, and units provided by this invention in the form of purely computer-readable program code, the same functions can be achieved entirely through logical programming of the method steps, making the system and its various devices, modules, and units of this invention function in the form of logic gates, switches, application-specific integrated circuits, programmable logic controllers, and embedded microcontrollers. Therefore, the system and its various devices, modules, and units provided by this invention can be considered as a hardware component, and the devices, modules, and units included therein for implementing various functions can also be considered as structures within the hardware component; alternatively, the devices, modules, and units for implementing various functions can be considered as both software modules implementing the method and structures within the hardware component.

[0102] Specific embodiments of the present invention have been described above. It should be understood that the present invention is not limited to the specific embodiments described above, and those skilled in the art can make various changes or modifications within the scope of the claims, which do not affect the essence of the present invention. Unless otherwise specified, the embodiments and features described in this application can be arbitrarily combined with each other.

Claims

1. A method for measuring the activity of low-to-medium radioactive waste bins using gamma scanning based on neural networks, characterized in that, include: Step S1: When measuring the waste bin, the medium inside the bin is assumed to be uniformly distributed, and the waste bin is divided into several segments in the vertical direction; Step S2: The radioactive point sources within each segment are assumed to be ring-shaped line sources, and the radius of each ring source is the horizontal distance from the corresponding point source to the center of the barrel axis; all ring sources within the same segment are equivalent to one equivalent ring source relative to the detector. Step S3: Using a gamma spectrometer, a segment is scanned at the same height and three different eccentricity positions. Through spectral analysis, the corresponding nuclide is identified based on the characteristic energy of the full-energy peak, and the corresponding count rate at the three detection positions is obtained. Step S4: Use the count rate at the three detection locations combined with a neural network algorithm to solve for the equivalent ring source radius and its detection efficiency; Step S5: Calculate the kernel activity of the current segment layer based on the equivalent ring source radius and its corresponding detection efficiency; repeat steps S1 to S5 to obtain the kernel activity of each segment layer. Step S6: Sum the activity of the kernels in each segment layer to obtain the total activity of the waste bin.

2. The method for measuring the activity gamma scan of low-to-medium radioactive waste bins based on neural networks according to claim 1, characterized in that, Step S2 involves dividing the waste bin into multiple segments along its axial direction, scanning the waste bin with a high-purity germanium detector with a collimator, rotating the waste bin at a constant speed, and equating a point source at a certain radius within the bin to a ring-shaped line source at that radius. It is assumed that there are N point sources of the same type located at different radii within a certain segment, which are equivalent to N ring sources with different radii.

3. The method for measuring the activity gamma scan of low-to-medium radioactive waste bins based on neural networks according to claim 1, characterized in that, Step S4 employs the following: Step S4.1: Assume the detector measures a segment of the layer at locations A, B, and C, with count rates of respectively... , , ; （4） （5） （6） in, This represents the probability that a nuclide emits gamma rays of this energy during decay. Indicates activity; Indicates that the detector has a radius of The detection efficiency of the ring source; The equivalent loop source radius when the detector is at position B is expressed as... The difference; The equivalent loop source radius when the detector is at position C is expressed as... The difference; Step S4.2: Based on the count rate ratio of detectors at different measurement locations, the equivalent loop source radius is obtained using a neural network algorithm; （7）。 4. The method for measuring the activity gamma scan of low-to-medium radioactive waste bins based on neural networks according to claim 3, characterized in that, The detection efficiency of the equivalent ring source is calculated using a passive efficiency calibration method based on the radius of the equivalent ring source. .

5. The method for measuring the activity gamma scan of low-to-medium radioactive waste bins based on neural networks according to claim 4, characterized in that, Step S5 employs the following: 。 6. A neural network-based gamma scanning measurement system for the activity of low-to-medium radioactive waste bins, characterized in that, include: Module M1: When measuring the waste bin, the medium inside the bin is assumed to be uniformly distributed, and the waste bin is divided into several segments in the vertical direction; Module M2: Radioactive point sources within each segment are assumed to be ring-shaped line sources, with the radius of each ring source being the horizontal distance from the corresponding point source to the center of the barrel axis; all ring sources within the same segment are equivalent to one equivalent ring source relative to the detector. Module M3: Using a gamma spectrometer, a segment is scanned at the same height and three different eccentricity positions. Through spectral analysis, the corresponding nuclides are identified based on the characteristic energy of the full-energy peak, and the corresponding count rates at the three detection positions are obtained. Module M4: Utilizes the count rates at three detection locations combined with a neural network algorithm to solve for the equivalent loop source radius and its detection efficiency; Module M5: Calculate the kernel activity of the current segment layer based on the equivalent ring source radius and its corresponding detection efficiency; repeat modules M1 to M5 to obtain the kernel activity of each segment layer. Module M6: Summing the activity of kernels in each segment layer to obtain the total activity of the waste bin.

7. The neural network-based gamma scanning measurement system for low-to-intermediate radioactive waste bin activity as described in claim 6, characterized in that, The module M2 employs the following method: dividing the waste bin into multiple segments along its axial direction, scanning the waste bin with a high-purity germanium detector equipped with a collimator, rotating the waste bin at a uniform speed, and treating a point source at a certain radius within the bin as an equivalent to a ring-shaped line source at that radius; assuming that there are N point sources of the same type located at different radii within a certain segment, these are equivalent to N ring sources with different radii.

8. The neural network-based gamma scanning measurement system for low-to-medium radioactive waste bin activity according to claim 6, characterized in that, The module M4 adopts: Module M4.1: Assume the detector measures a segment of the layer at locations A, B, and C, with count rates of respectively... , , ; （4） （5） （6） in, This represents the probability that a nuclide emits gamma rays of this energy during decay. Indicates activity; Indicates that the detector has a radius of The detection efficiency of the ring source; The equivalent loop source radius when the detector is at position B is expressed as... The difference; The equivalent loop source radius when the detector is at position C is expressed as... The difference; Module M4.2: The equivalent loop source radius is obtained using a neural network algorithm based on the count rate ratio of detectors at different measurement locations; （7）。 9. The neural network-based gamma scanning measurement system for low-to-medium radioactive waste bin activity according to claim 8, characterized in that, The detection efficiency of the equivalent ring source is calculated using a passive efficiency calibration method based on the radius of the equivalent ring source. .

10. The neural network-based gamma scanning measurement system for low-to-medium radioactive waste bin activity according to claim 9, characterized in that, The module M5 adopts: 。

Images