A method for rapid nuclide identification and a device thereof

Abstract

The application discloses a kind of fast nuclide identification method and equipment, the method comprises the following steps: the target energy spectrum is normalized to energy spectrum;Energy spectrum pretreatment;Full spectrum fitting;Determine nuclide identification result.The present application directly saves the step such as peak searching, uses the way of bringing all energy spectrum into fitting to calculate the proportion of various nuclides, in the case of complex nuclide and low resolution detector, also has good nuclide identification accuracy.The present application uses monte carlo simulation to produce standard nuclide spectrum, avoids the environmental interference problem introduced by actual measurement energy spectrum, ensures the accuracy of benchmark energy spectrum, and uses snip method to reduce the influence of Compton plateau in the bottom of energy spectrum caused by Compton scattering of gamma energy spectrum, so as to reduce the interference of useless information in energy spectrum to data fitting.

Description

Technical Field

[0001] This invention relates to an identification method and its application, and more particularly to a rapid nuclide identification method and device, belonging to the field of nuclear radiation monitoring signal processing. Background Technology

[0002] Nuclide identification is a method used in the field of nuclear radiation monitoring to identify the types of radioactive isotopes in the surrounding environment. With the development of nuclear application technologies, nuclear science is increasingly demonstrating its power in various fields, most notably customs inspection, environmental monitoring, medical testing, food safety, and metal flaw detection. Nuclear radiation detection technology is directly or indirectly involved in all of these. Nuclide identification is a crucial aspect of nuclear radiation detection technology. However, due to the inherent statistical fluctuations in nuclear radiation energy spectra, which vary depending on the type and dose of radiation, despite decades of development, it is difficult to find a single method that can solve the nuclide identification problem in all scenarios.

[0003] Currently, the traditional nuclide identification algorithm process is as follows: find the characteristic peak—correspond the peak position to the energy—compare the characteristic peak energy with the nuclide library—calculate the confidence level based on the matching degree between the characteristic peak and the nuclide library. This method is perfectly capable of identifying a single nuclide. See also... Figure 1-2 As shown, the visualization results of the energy spectrum data obtained using the traditional nuclide identification algorithm and the peak finding results are presented; the characteristic peak information is as follows:

[0004] Nuclide library information extraction, such as Figure 8 As shown, the confidence level of various nuclides can be obtained by comparing the characteristic peaks with the nuclide library information. However, this method has inherent limitations in complex environments with multiple nuclides. When there are too many characteristic peaks of a nuclide or when the energy is so similar that the detector resolution is insufficient to distinguish them, the mutual influence of their peak shapes manifests as a series of irregular energy peaks in the energy spectrum. This is very unfriendly to peak-finding algorithms. This directly leads to the poor performance of traditional algorithms in cases of multiple nuclides and low detector resolution. In other words, traditional nuclide identification methods based on peak-finding are difficult to find the necessary characteristic peak information in cases of irregular peak shapes and peak overlap, resulting in deviations in the nuclide identification results. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention discloses a rapid radionuclide identification method, the technical solution of which is as follows: A rapid nuclide identification method, characterized by comprising the following steps: Step 1: Normalize the target energy spectrum into an energy spectrum; Step 2: Energy spectrum preprocessing; Step 3: Full spectrum fitting; Step 4: Determine the nuclide identification result.

[0006] The present invention also discloses a non-volatile storage medium, characterized in that: the non-volatile storage medium includes a stored program, wherein the program, when running, controls the device where the non-volatile storage medium is located to execute the above-described method.

[0007] The present invention also discloses a nuclear radiation monitoring signal processing device, characterized in that it includes a processor and a memory; the memory stores computer-readable instructions, and the processor is used to run the computer-readable instructions, wherein the computer-readable instructions execute the above-described method when they are run. Beneficial effects

[0008] 1. This invention uses full-spectrum decomposition of gamma energy for nuclide identification, unlike traditional nuclide identification methods that require first finding characteristic peaks, then comparing these peaks with a nuclide library, and finally calculating the nuclide confidence level. This avoids the loss of spectral information caused during intermediate data extraction steps, thus maximizing the preservation of spectral data integrity and improving the accuracy of nuclide identification.

[0009] 2. The reference vector energy spectrum used in the method of the present invention is the standard nuclide spectrum produced by Monte Carlo simulation, which avoids the environmental interference problem introduced by the measured energy spectrum and ensures the accuracy of the reference energy spectrum.

[0010] 3. In the method of the present invention, when performing full-spectrum decomposition, the target energy spectrum is normalized to the same energy spectrum as the standard reference spectrum before fitting, and the snip method is used to reduce the influence of the Compton plateau at the bottom of the energy spectrum caused by Compton scattering of the gamma energy spectrum, thereby reducing the interference of useless information in the energy spectrum on data fitting. Attached Figure Description

[0011] Figure 1 This is a schematic diagram of the energy spectrum data visualization results in the existing technology; Figure 2 This is a schematic diagram of the peak finding results in the existing technology; Figure 3 This is a schematic diagram illustrating the process of converting the target energy spectrum into an energy spectrum according to the present invention; Figure 4 This is a flowchart of the preprocessing of the energy spectrum using the ln function before snip processing in this invention; Figure 5 This is a schematic diagram of the energy spectrum snip filtering process of the present invention; Figure 6 This is a schematic diagram illustrating the process of anti-inverse quantization of the energy spectrum after snip filtering using the exp function in this invention. Figure 7 This is a schematic diagram of the process by which the present invention determines the confidence level of a nuclide by calculating the similarity between the synthesized energy spectrum and the target energy spectrum; Figure 8 Data is extracted for the nuclide library information. Detailed Implementation

[0012] This invention discloses a rapid method for identifying nuclides, the steps of which are as follows: Step 1: Normalize the target energy spectrum into an energy spectrum Step 1-1: Obtain the current cumulative energy spectrum of the device as the target energy spectrum. The target energy spectrum data format is a one-dimensional array with a length of 4096 ( / 2048 / 1024).

[0013] Steps 1-2: Perform peak finding in the target energy spectrum and determine the area of ​​the corresponding energy peak: Use the zero-symmetric area method (zero-symmetric area can be understood as a variable window digital filter algorithm that can calculate the value of the zero-symmetric area corresponding to each channel address in the energy spectrum) to process the energy spectrum. Find the point in the energy spectrum processing result where the zero-symmetric area value is greater than 3 and the value is the maximum value within ±1 half-width and height as the found energy peak. Calculate the total value within the ROI range of ±1 half-width and height for the corresponding channel address based on the peak finding result, and consider it as the area of ​​the energy peak.

[0014] Steps 1-3: Determine the position of the 1460 keV characteristic peak of the natural nuclide K40 based on the information of the found energy peak: Determine the position of the 1460 keV characteristic peak of the natural nuclide K40 based on whether the energy value is greater than 1460 keV*99% and less than 1460 keV*101%.

[0015] Steps 1-4: Determine the drift ratio of the target energy spectrum relative to the factory-calibrated energy spectrum based on the position of the 1460 keV characteristic peak of the natural nuclide K40. The drift ratio is a parameter used to correct the energy calibration curve, which is calibrated by scaling the K40 characteristic peak position to channel 2048. When the actual measured channel address of the K40 characteristic peak in the target energy spectrum is x, the drift ratio is calculated as 2048 / x.

[0016] Steps 1-5: Linear stretching of the target energy spectrum according to the drift ratio: Let the channel address of the characteristic peak of K40 obtained by peak finding of the target energy spectrum be x. Then, by stretching with a drift ratio of 2048 / x, the peak position of K40 in the stretched energy spectrum will be 2048. For example, if the drift ratio is 2, it means that the position of K40 in the target energy spectrum is 1024. The stretching process is to evenly distribute the count of channel x of the target energy spectrum into the array cells of 2x-1 and 2x in the stretched energy spectrum. The count of each cell in the stretched energy spectrum is half of the count value of channel x of the target energy spectrum.

[0017] Steps 1-6: Convert the target energy spectrum into an energy spectrum using a pre-calibrated address-energy curve. The conversion steps are as follows: Figure 3 As shown: The process involves calculating the energy corresponding to each channel address of the target energy spectrum using an energy calibration curve. It is assumed that the counts between two adjacent channels are uniformly distributed with respect to the energy. These counts are then assigned to the corresponding energy spectrum array, thus completing the conversion from the target energy spectrum to the energy spectrum. The curve formula is: y = a*x^3 + b*x^2 + c*x^+d Where x represents the channel address of the energy spectrum; if it is an array of length 4096, then x takes values ​​from 0 to 4095. y represents the energy value corresponding to the channel address, in keV. a, b, c, and d are pre-calibrated fixed coefficient values.

[0018] Step 2: Energy spectrum preprocessing Step 2-1: Background Energy Spectrum Subtraction. The background energy spectrum is the measured energy spectrum obtained by the equipment in a passive environment when powered on. At the end of the passive measurement, the accumulated energy spectrum needs to be normalized to the energy spectrum and then saved for later use. Background energy spectrum subtraction simply involves subtracting the background energy spectrum from the target energy spectrum.

[0019] Step 2-2: Use the snip method to eliminate the influence of the Compton plateau on the energy spectrum. The steps are as follows: Figure 4-6 As shown: Figure 4 The target energy spectrum is first logarithmically transformed using the ln function, and then... Figure 5 The filtering algorithm shown will filter the logarithmic data and finally use... Figure 6 The exp function shown performs anti-inverse scalarization on the energy spectrum.

[0020] Step 3. Full spectrum fitting Step 3-1. Allocate matrix data space. If the device nuclide library contains N nuclides, then the reference vector space X of the nuclide library needs to be allocated as n*4096. Simultaneously, the target energy spectrum space Y needs to be allocated as 4096. The weighting coefficient space W is 4096 in size. The space for the nuclide contribution coefficient C is n. The specific fitting equation is as follows.

[0021] [X1*C1+X2*C2+.....+Xn*Cn]*W T =Y Step 3-2. Initialize the data. X1~Xn are assigned standard nuclide library arrays. W is assigned the reciprocal of the detection efficiency corresponding to each energy level, thus ensuring a high fitting weight in the high-energy region of the energy spectrum. Y is assigned the target energy spectrum. C1~Cn are the contribution coefficients of each nuclide to be determined, which are also the targets for data fitting.

[0022] Step 3-3. Allocate 4096*n space as the data space for this linear fit.

[0023] Steps 3-4. Drive the linear fitting engine to fit the values ​​of C1 to Cn: Linear fitting is a fitting algorithm used in engineering. The calculation process uses an open-source mathematical library. It requires input of known quantities X1 to Xn arrays, W array, and Y array. Then, the calculation can be started to obtain the optimal solution of C1 to Cn. Step 4. Determine the nuclide identification result Step 4-1. Calculate the final fitted superimposed energy spectrum data based on X1*C1+X2*C2+.....+Xn*Cn.

[0024] Step 4-2. Determine whether the nuclide is trustworthy by calculating the 90% characteristic area region of the energy spectrum for each nuclide and comparing the similarity between the synthesized energy spectrum and the target energy spectrum in this region. Specific steps are as follows: Figure 3 As shown.

[0025] This invention differs from traditional nuclide identification methods by directly eliminating peak-finding steps and calculating the proportions of various nuclides by fitting the entire energy spectrum. It maintains good nuclide identification accuracy even with complex nuclides and low-resolution detectors. This invention uses Monte Carlo simulation to produce standard nuclide spectra, avoiding environmental interference introduced by measured energy spectra and ensuring the accuracy of the reference spectrum. Furthermore, it uses the snip method to mitigate the impact of Compton scattering at the bottom of the gamma spectrum on the Compton plateau, thereby reducing interference from useless information in the spectrum on data fitting.

[0026] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the claimed invention. The scope of protection claimed by the appended claims and their equivalents is defined.

Claims

1. A rapid nuclide identification method, characterized in that, Includes the following steps: Step 1: Normalize the target energy spectrum into an energy spectrum; Step 2: Energy spectrum preprocessing; Step 3: Full spectrum fitting; Step 3-1: Allocate matrix data space: When there are N nuclides in the device nuclide library, it is necessary to allocate a reference vector space X of n*4096; at the same time, allocate a target energy spectrum space Y of 4096, a weighting coefficient space W of 4096, and a space of n for the nuclide contribution coefficient C; the specific fitting equation is as follows: [X1*C1+X2*C2+.....+Xn*Cn]*W T =Y; Where: X1~Xn are assigned the standard nuclide library array, W is assigned the reciprocal of the detection efficiency corresponding to each energy level, Y is assigned the target energy spectrum, and C1~Cn are the contribution coefficients of each nuclide to be determined; Step 3-2: Allocate 4096*n space as the data space for this linear fit; Step 3-3: Drive the linear fitting engine to fit the values ​​of C1 to Cn; Step 4: Determine the nuclide identification result.

2. The rapid nuclide identification method according to claim 1, characterized in that step 1 includes the following: Step 1-1: Obtain the current cumulative energy spectrum of the device as the target energy spectrum. The target energy spectrum data format is a one-dimensional array with a length of 4096, 2048, or 1024. Steps 1-2: Perform peak finding in the target energy spectrum and determine the area of ​​the corresponding energy peak; Steps 1-3: Determine the position of the 1460 keV characteristic peak of the natural nuclide K40 based on the information of the found energy peak; Steps 1-4: Determine the drift ratio of the target energy spectrum relative to the energy spectrum at the time of manufacture based on the position of the characteristic peak of 1460 keV of the natural nuclide K40. Steps 1-5: Linearly stretch the target energy spectrum according to the drift ratio; Steps 1-6: Convert the target energy spectrum into an energy spectrum using a pre-calibrated channel-energy curve.

3. The rapid nuclide identification method according to claim 2, characterized in that: the conversion step curve formula is: y = a*x^3 + b*x^2 + c*x^+d; Where x represents the channel address of the energy spectrum, and if it is an array of length 4096, then x takes the value from 0 to 4095; y represents the energy value corresponding to the channel address, in keV; and a, b, c, and d are pre-calibrated fixed coefficient values.

4. The rapid nuclide identification method according to claim 2, characterized in that step 2 includes the following: Step 2-1: Background energy spectrum subtraction: The background energy spectrum is the measured energy spectrum of the device in a passive environment when it is powered on. At the end of the passive measurement, the accumulated energy spectrum needs to be normalized to the energy spectrum and then saved for later use. Background energy spectrum subtraction only requires subtracting the background energy spectrum from the target energy spectrum. Step 2-2: Use the snip method to eliminate the influence of Compton plateau on the energy spectrum.

5. The rapid nuclide identification method according to claim 2, characterized in that step 4 includes the following: Step 4-1. Calculate the final fitted superimposed energy spectrum data based on X1*C1+X2*C2+.....+Xn*Cn; Step 4-2. Determine whether a nuclide is trustworthy by calculating the 90% characteristic area region of the energy spectrum of each nuclide and comparing the similarity between the synthesized energy spectrum and the target energy spectrum in this region.

6. A non-volatile storage medium, characterized in that: The non-volatile storage medium includes a stored program, wherein the program, when executed, controls the device where the non-volatile storage medium is located to perform the method described in any one of claims 1 to 5.

7. A nuclear radiation monitoring signal processing device, characterized in that: It includes a processor and a memory; the memory stores computer-readable instructions, and the processor is configured to execute the computer-readable instructions, wherein the computer-readable instructions, when executed, perform the method according to any one of claims 1 to 5.

Images