Method and device for predicting leakage hazard range of nuclear facility

Abstract

The invention discloses a method and a device for predicting a leakage hazard range of a nuclear facility, and relates to the technical field of nuclear response emergency. The method comprises the following steps: acquiring reconnaissance data corresponding to a nuclear facility leakage area; performing energy spectrum preprocessing on the gamma energy spectrum original pulse signal to obtain processed energy spectrum data, and performing background deduction processing on the processed energy spectrum data based on an adaptive SNIP algorithm to obtain net energy spectrum data; performing nuclide identification processing according to the net energy spectrum data to obtain a nuclide identification result, and performing nuclide activity concentration calculation based on the nuclide identification result and the nuclear radiation reconnaissance data to obtain nuclide activity concentration; performing source item inversion based on geographic data, meteorological data and nuclide activity concentration to obtain leakage source item information; and obtaining a hazard range prediction result based on the meteorological data, the leakage source item information and the nuclide activity concentration. According to the invention, the sampling precision of the system can be improved, the pulse accumulation can be reduced, the accumulation discarding problem can be effectively solved, and the adaptability is higher.

Description

Technical Field

[0001] This application relates to the field of nuclear response emergency technology, and in particular to a method and apparatus for predicting the hazard range of a nuclear facility leak. Background Technology

[0002] In the fields of nuclear emergency monitoring and radiation protection, the rapid and accurate identification of radionuclides and the determination of the extent of nuclear facility leak hazards are crucial for accident assessment and emergency response.

[0003] Currently, the identification of the hazard range of nuclear facility leaks mainly relies on nuclide identification technology based on gamma-ray spectroscopy analysis. However, this technology faces three major challenges: First, traditional peak-finding methods lack sensitivity for detecting weak peaks and are difficult to effectively decompose overlapping peaks, potentially leading to missed or misidentified nuclides. Second, it is difficult to distinguish between different nuclides with the same characteristic peaks (such as the 511 keV peaks of F-18 and Na-22). Third, existing methods rely on human intervention, which is insufficient to meet timeliness requirements in scenarios like nuclear facility leaks where there is a radiation hazard to humans. Furthermore, although existing technologies have been improved through neural networks or peak area weighted matching, they still suffer from poor generalization and insufficient anti-interference capabilities. Summary of the Invention

[0004] The purpose of this application is to address at least one of the aforementioned technical deficiencies.

[0005] On the one hand, embodiments of this application provide a method for predicting the hazard range of a nuclear facility leak, the method comprising: Acquire reconnaissance data corresponding to the area of ​​nuclear facility leakage. The reconnaissance data includes raw gamma-ray energy spectrum pulse signals, meteorological data, geographical data, and nuclear radiation reconnaissance data. The original γ-energy spectrum pulse signal is preprocessed to obtain the processed energy spectrum data. The background subtraction is then performed on the processed energy spectrum data based on the adaptive SNIP algorithm to obtain the net energy spectrum data. The energy spectrum preprocessing includes digital trapezoidal forming and symmetrical zero-area forming. Nuclide identification is performed based on net energy spectrum data to obtain nuclide identification results. Nuclide activity concentration is then calculated based on the nuclide identification results and nuclear radiation detection data to obtain the nuclide activity concentration. Source term inversion is performed based on geographical data, meteorological data, and radionuclide activity concentration to obtain leakage source term information, which includes leakage source strength and leakage location. Based on meteorological data, leakage source term information, and radionuclide activity concentration, the hazard range prediction results are obtained, which include the predicted radionuclide activity concentration field, dose rate field, and pollution boundary.

[0006] Optionally, the raw γ-ray energy spectrum pulse signal is preprocessed to obtain processed energy spectrum data, including: Digital signal processing is performed on the original energy spectrum pulse signal to obtain the trapezoidal pulse corresponding to the original γ energy spectrum pulse signal. The trapezoidal pulse is then stacked based on the degree of overlap of each pulse in the trapezoidal pulse to obtain the processed trapezoidal pulse. Obtain the convolution kernel, and perform convolution processing on the processed trapezoidal pulses according to the convolution kernel to obtain the convolved trapezoidal pulses. The convolution kernel has a symmetrical shape and an area of ​​zero. The pulse amplitudes of the convolved trapezoidal pulses are statistically analyzed to obtain the processed energy spectrum data.

[0007] Optionally, the processed energy spectrum data is subjected to background subtraction based on the adaptive SNIP algorithm to obtain net energy spectrum data, including: The half-width at half-height (WHM) of each energy point is calculated using the energy scale formula, and the number of iterations is determined based on the WHM of each energy point. The processed energy spectrum data is iterated from the high energy end to the low energy end based on the adaptive SNIP algorithm until the number of iterations is reached, so as to obtain the estimated background energy spectrum data. Based on the processed energy spectrum data and the background energy spectrum data, the net energy spectrum data is obtained.

[0008] Optionally, nuclide identification processing can be performed based on the net energy spectrum data to obtain nuclide identification results, including: The net energy spectrum data is convolved using a Gaussian symmetric zero-area convolution kernel, and the position of each candidate peak is determined based on the convolved net energy spectrum data. The nuclide matching is performed on each candidate peak position based on the Bayesian posterior probability algorithm to obtain the nuclide identification result, which includes the nuclide type and the characteristic peak information corresponding to each nuclide type.

[0009] Optionally, the position of each candidate peak is determined based on the net energy spectrum data after convolution, including: The position of each candidate peak is determined based on the position of the zero-crossing point in the net energy spectrum data after convolution; Multiple Gaussian fitting is used to perform peak decomposition based on the position of each candidate peak to determine each candidate peak position.

[0010] Optionally, nuclide matching is performed on each candidate peak position based on a Bayesian posterior probability algorithm to obtain nuclide identification results, including: Determine the energy value corresponding to each candidate peak position, and match the energy value corresponding to each candidate peak position with the theoretical characteristic energy of nuclides in the preset nuclide library to obtain each candidate nuclide; For each candidate nuclide, the posterior probability of each candidate nuclide is determined based on the Bayesian posterior probability algorithm, and the nuclide identification result is obtained based on the posterior probability of each candidate nuclide.

[0011] Optionally, the nuclear radiation reconnaissance data includes alpha aerosol concentration and surface contamination measurements; the nuclide activity concentration includes gamma nuclide activity concentration, alpha aerosol activity concentration, and surface contamination activity concentration. Based on the nuclide identification results and the nuclear radiation reconnaissance data, the nuclide activity concentration is calculated to obtain the nuclide activity concentration, including: The detection efficiency and sampling time corresponding to the acquisition of the raw pulse signal of the γ energy spectrum are obtained, and the activity concentration of the nuclide is calculated based on the characteristic peak information, detection efficiency and sampling time corresponding to each type of nuclide to obtain the γ nuclide activity concentration; Obtain the sampling flow rate and filtration efficiency corresponding to the α aerosol concentration, and obtain the α aerosol activity concentration based on the α aerosol concentration, sampling flow rate and filtration efficiency; The detector efficiency and detection area corresponding to the surface contamination measurement value are obtained, and the surface contamination activity concentration is obtained based on the surface contamination measurement value, detector efficiency, and detection area.

[0012] Optionally, source term inversion can be performed based on geographic data, meteorological data, and radionuclide activity concentration to obtain leakage source term information, including: The diffusion coefficient value was calculated based on geographical and meteorological data, and a Gaussian multiplying smoke model was constructed based on the diffusion coefficient value. The initial leakage source term information is obtained, and the predicted nuclide activity concentration is obtained based on the initial leakage source term information and the Gaussian multipuff model. The initial leakage source term information includes the assumed leakage source strength and the assumed leakage location. Based on the predicted nuclide activity concentration and the nuclide activity concentration processed by Kalman filtering, leakage source term information is obtained.

[0013] Optionally, based on meteorological data, leakage source term information, and radionuclide activity concentration, the hazard range prediction results are obtained, including: The leak source term information and meteorological data are input into a preset Gaussian diffusion model to obtain the predicted activity concentration field of the nuclide; Obtain the radionuclide dose conversion factor, and calculate the dose rate field based on the radionuclide dose conversion factor and the predicted activity concentration field of the radionuclide; Obtain the pollution level threshold, and based on the dose rate field and the pollution level threshold, obtain the pollution boundary.

[0014] On the other hand, embodiments of this application provide a device for predicting the hazard range of a nuclear facility leak, comprising: The reconnaissance data acquisition module is used to acquire reconnaissance data corresponding to the nuclear facility leakage area. The reconnaissance data includes raw gamma-ray energy spectrum pulse signals, meteorological data, geographical data, and nuclear radiation reconnaissance data. The energy spectrum preprocessing module is used to preprocess the original γ energy spectrum pulse signal to obtain the processed energy spectrum data, and to perform background subtraction processing on the processed energy spectrum data based on the adaptive SNIP algorithm to obtain the net energy spectrum data. The energy spectrum preprocessing includes digital trapezoidal shaping processing and symmetrical zero-area shaping processing. The nuclide processing module is used to perform nuclide identification processing based on net energy spectrum data to obtain nuclide identification results, and to calculate nuclide activity concentration based on nuclide identification results and nuclear radiation detection data to obtain nuclide activity concentration. The source term inversion module is used to perform source term inversion based on geographical data, meteorological data, and radionuclide activity concentration to obtain leakage source term information, which includes leakage source strength and leakage location. The hazard range prediction module is used to obtain hazard range prediction results based on meteorological data, leakage source term information and radionuclide activity concentration. The hazard range prediction results include the predicted activity concentration field, dose rate field and contamination boundary of the radionuclide.

[0015] In another aspect, embodiments of this application provide an electronic device, including a processor and a memory: The memory is configured to store machine-readable instructions that, when executed by the processor, cause the processor to perform any of the methods for predicting the extent of a nuclear facility leak hazard.

[0016] The beneficial effects of the technical solutions provided in this application include at least the following: In this application, the raw γ-ray energy spectrum pulse signal can be acquired, and then digital trapezoidal shaping and symmetrical zero-area shaping processing are performed on the acquired raw γ-ray energy spectrum pulse signal. This converts the exponentially decaying pulse signal into a trapezoidal pulse, reducing errors caused by pulse accumulation and improving the signal-to-noise ratio. Symmetrical zero-area shaping processing is also performed on the data after digital trapezoidal shaping to suppress baseline drift and improve energy spectrum resolution. Furthermore, when identifying nuclide activity concentration, nuclide identification processing is based on net energy spectrum data. In determining the net energy spectrum data, overlapping peaks under different conditions are fully considered, and peak finding is performed in a corresponding manner according to the actual situation, which can effectively decompose overlapping peaks and reduce the possibility of nuclide missed detection or misjudgment. At the same time, compared with traditional processing methods, it does not produce nonlinear errors introduced by analog circuits, can improve the sampling accuracy of the system, improve the system resolution, effectively shorten the pulse width, reduce pulse accumulation, improve the system counting throughput, and can more effectively handle the problem of accumulation rejection. It is not affected by changes in external temperature and humidity, making it more flexible and adaptable. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 A flowchart illustrating a method for predicting the hazard range of a nuclear facility leak, provided as an embodiment of this application; Figure 2 A schematic diagram of the digital pulse shaping and energy spectrum acquisition system provided in the embodiments of this application; Figure 3 A schematic diagram of signal changes provided in an embodiment of this application; Figure 4 A schematic diagram illustrating the pulse portion accumulation situation provided in an embodiment of this application; Figure 5 A schematic diagram illustrating a severe pulse accumulation situation provided in an embodiment of this application; Figure 6 A schematic diagram illustrating the complete pulse stacking situation provided in an embodiment of this application; Figure 7 A schematic diagram of the trapezoidal pulse forming process provided in an embodiment of this application; Figure 8 A schematic diagram of the structure of a nuclear facility leakage hazard range prediction device provided in an embodiment of this application; Figure 9 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0019] The embodiments of this application are described in detail below. Examples of these embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain this application, and should not be construed as limiting the invention.

[0020] Those skilled in the art will understand that, unless specifically stated otherwise, the singular forms “a,” “an,” “the,” and “the” used herein may also include the plural forms. It should be further understood that the term “comprising” as used in this application means the presence of the stated features, integers, steps, operations, elements, and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or groups thereof. It should be understood that when we say an element is “connected” or “coupled” to another element, it can be directly connected or coupled to the other element, or there may be intermediate elements. Furthermore, “connected” or “coupled” as used herein can include wireless connections or wireless coupling. The term “and / or” as used herein includes all or any units and all combinations of one or more associated listed items.

[0021] To make the objectives, technical solutions, and advantages of this application clearer, the embodiments of this application will be described in further detail below with reference to the accompanying drawings.

[0022] The technical solution of this application and how the technical solution of this application solves the above-mentioned technical problems are described in detail below with specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments. The embodiments of this application will now be described with reference to the accompanying drawings.

[0023] Specifically, such as Figure 1 As shown, the method may include: Step S101: Obtain reconnaissance data corresponding to the nuclear facility leakage area. The reconnaissance data includes raw γ-ray energy spectrum pulse signals, meteorological data, geographical data, and nuclear radiation reconnaissance data.

[0024] Optionally, the raw gamma-ray spectrum pulse signal can be acquired using a LaBr3 detector via an H3GSM01 energy dispersive spectrometer. The H3GSM01 energy dispersive spectrometer consists of a detector body, detector end cap, detector core, and core buffer pad. The core buffer pad is divided into upper, middle, and lower sections, encasing the detector core. All core buffer pads are made of silicone rubber, exhibiting excellent axial elasticity and excellent radial (petal-shaped) contractility, ensuring excellent shock absorption. When the detector is subjected to impact or vibration, the core buffer pad contracts under external pressure, absorbing most of the impact energy, thus buffering and damping the detector core.

[0025] Meteorological data, including wind speed, wind direction, temperature, and humidity, can be acquired through vehicle-mounted meteorological sensors; geographic data, including three-dimensional terrain and building distribution, can be obtained through vehicle-mounted GIS (Geographic Information System); and nuclear radiation reconnaissance data includes alpha aerosol concentration and surface contamination measurements. Alpha aerosol concentration data is obtained from the H2PTM01M mobile aerosol monitor, while surface contamination measurements are obtained from the H3ABM01 surface contamination meter.

[0026] Step S102: Perform energy spectrum preprocessing on the original γ energy spectrum pulse signal to obtain processed energy spectrum data. Then, perform background subtraction processing on the processed energy spectrum data based on the adaptive SNIP (Statistics-sensitive Nonlinear Iterative Peak-clipping) algorithm to obtain net energy spectrum data. The energy spectrum preprocessing includes digital trapezoidal shaping processing and symmetrical zero-area shaping processing.

[0027] Optionally, to improve the resolution and counting throughput of the measurement system, the acquired raw γ-ray energy spectrum pulse signals can be digitally trapezoidal shaped to convert the exponentially decaying pulse signals into trapezoidal pulses, reducing errors caused by pulse stacking and improving the signal-to-noise ratio. Furthermore, symmetrical zero-area shaping processing is applied to the digitally trapezoidal shaped data to suppress baseline drift and improve energy spectrum resolution. Then, based on adaptive SNIP, background subtraction, a crucial task, is performed on the processed energy spectrum data to obtain the net energy spectrum data.

[0028] In an optional embodiment of this application, the raw γ-ray energy spectrum pulse signal is subjected to energy spectrum preprocessing to obtain processed energy spectrum data, including: Digital signal processing is performed on the original γ-energy spectrum pulse signal to obtain the trapezoidal pulse corresponding to the original γ-energy spectrum pulse signal. The trapezoidal pulse is then stacked based on the degree of overlap of each pulse in the trapezoidal pulse to obtain the processed trapezoidal pulse. Obtain the convolution kernel, and perform convolution processing on the processed trapezoidal pulses according to the convolution kernel to obtain the convolved trapezoidal pulses. The convolution kernel has a symmetrical shape and an area of ​​zero. The pulse amplitudes of the convolved trapezoidal pulses are statistically analyzed to obtain the processed energy spectrum data.

[0029] Optional, based on such Figure 2The digital pulse shaping and energy spectrum acquisition system shown can digitize (e.g., ADC (Analog-to-Digital Converter) conversion) the raw γ-ray energy spectrum pulse signal after obtaining it. The sampling is performed in two separate channels: pulse shaping (i.e., trapezoidal pulse) and pulse triggering. The pulse triggering module locates the pulse arrival time and provides a pulse positioning signal to the amplitude extraction unit after a delay. Then, the trapezoidal pulses are stacked based on the overlap of each pulse. The pulse shaping module performs trapezoidal shaping on the pulse sequence and simultaneously estimates and subtracts the baseline before the pulse arrival from the trapezoidal pulse to reduce amplitude fluctuations caused by baseline shift. Then, the processed trapezoidal pulse is convolved using a symmetrical zero-convolution kernel with an area of ​​0, resulting in a convolved trapezoidal pulse. The effective trapezoidal pulse's flat-top value is extracted based on the pulse control signal to obtain the pulse amplitude value. Finally, the pulse amplitude of the convolved trapezoidal pulse is accumulated in memory to form the processed energy spectrum data. The energy spectrum preprocessing process of the raw γ-ray energy spectrum pulse signal is described in detail below: The pulse triggering module in the system uses a second-order differential filter to process the input signal (such as...) Figure 3 The second-order difference process is performed on (as shown in a), and its output is as follows. Figure 3 As shown in b in the diagram. The output signal is compared with a specific threshold by a comparator. When its amplitude is greater than the threshold, the zero-crossing detection module is enabled. When the zero-crossing detection module is enabled, it performs zero-crossing detection on the output signal of the second-order differential filter and generates a trigger signal at its zero-crossing point, as shown in the diagram. Figure 3 As shown in c in the diagram. The trapezoidal filter converts the input exponential pulse signal into a trapezoidal signal, specifically as shown in... Figure 3 As shown in 'd', the rise time, peak time, and fall time of the trapezoidal signal can all be set by software. The baseline estimation module estimates the baseline of the trapezoidal signal output from the trapezoidal filter and calculates the baseline value of the signal. The baseline subtraction module subtracts the baseline value from the trapezoidal signal to achieve baseline subtraction and eliminate the influence of baseline drift. The peak extraction module performs peak sampling on the baseline-filtered trapezoidal signal. The sampling point for peak sampling is the trigger output from the zero-crossing detection module delayed. This delay time can be set by software to be located at the midpoint of the peak region of the trapezoidal signal, as shown in the figure. Figure 3 As shown in e, the sampled peak energy information is stored in the internal memory of the FPGA (Field-Programmable Gate Array). The memory is connected to the processor via an external bus interface, enabling the processor to access the internal memory to read the pulse peak information.

[0030] In practical applications, during the peak sampling of pulses, there is a certain probability that two pulses will overlap and accumulate. To overcome the impact of pulse accumulation on measurement accuracy, it is necessary to discard the overlapping pulses during the peak sampling process. Specifically, the following methods can be used to handle the overlap situation: When two consecutive pulses only partially overlap (i.e., the right falling edge of the first trapezoid overlaps with the left rising edge of the second trapezoid, but the rising and falling edges of the trapezoids do not overlap with the top of the other trapezoid), specifically as follows: Figure 4 As shown, if a flat and stable peak region is still present after passing through the pulse shaping circuit and the peak hold circuit, then both pulse peaks will still be sampled, and the acquisition process will not introduce any error. When two consecutive pulses severely overlap (i.e., when the right falling edge of the first trapezoid overlaps with the flat top of the second trapezoid), specifically as shown... Figure 5 As shown, if the pulse shaping circuit and peak hold circuit cannot separate the two pulses, and the peak hold result is severely distorted, then these two pulses are discarded, no sampling trigger signal is generated, and no sampling is performed. When two consecutive pulses completely overlap (i.e., when the first trapezoidal peak overlaps with the second trapezoidal peak), specifically as shown... Figure 6 As shown, the pulse acquisition module will process these two pulses as a single pulse, and the sampled peak value will be the sum of the peak values ​​of the two pulses.

[0031] In this application, the acquired raw γ-energy spectrum pulse signal is simulated using a digital trapezoidal forming method. Compared with traditional analog multichannel processing methods, this method does not introduce nonlinear errors from analog circuits, improves the sampling accuracy and resolution of the system, effectively shortens the pulse width, reduces pulse accumulation, and increases the system's counting throughput. It also exhibits better performance consistency across the entire operating range, more effectively handles the accumulation rejection problem, is unaffected by changes in external temperature and humidity, and is more flexible and adaptable.

[0032] Optionally, the trapezoidal shaping algorithm described above requires real-time baseline estimation. However, baseline estimation pauses after pulse detection and during pulse shaping, and low-frequency jitter and noise during this process affect the accuracy of peak extraction. To address this issue, the processed trapezoidal pulse can be convolved using a symmetrically shaped, zero-area convolution kernel (i.e., a symmetrical zero-area shaping algorithm) to obtain the convolved trapezoidal pulse. The shaping process is shown in Figure 7. Due to the symmetrical zero-area characteristic of the shaping, it has zero response to DC signals and first-order linear signals (which are usually components of low-frequency jitter). Therefore, it can effectively suppress low-frequency jitter. Regardless of how the DC baseline of the input signal changes, the baseline of the output signal is always 0, eliminating the need for baseline estimation. Baseline subtraction will no longer pause due to the pulse shaping process, thus exhibiting excellent system resolution characteristics.

[0033] In an optional embodiment of this application, background subtraction is performed on the processed energy spectrum data based on the adaptive SNIP algorithm to obtain net energy spectrum data, including: The half-width at half-height (WHM) of each energy point is calculated using the energy scale formula, and the number of iterations is determined based on the WHM of each energy point. The processed energy spectrum data is iterated from the high energy end to the low energy end based on the adaptive SNIP algorithm until the number of iterations is reached, so as to obtain the estimated background energy spectrum data. Based on the processed energy spectrum data and the background energy spectrum data, the net energy spectrum data is obtained.

[0034] Optionally, background subtraction is one of the important tasks in energy spectrum processing and interpretation. Based on this, for the obtained net energy spectrum data, the full width at half maximum (FWHM) of each energy point can be calculated using the energy scale formula shown below.

[0035]

[0036] Where FWHM is the full width at half maximum (FWHM), E is the energy, and a and b are peak width calibration coefficients.

[0037] Accordingly, the number of iterations of the SNIP algorithm can be determined based on the full width at half maximum (FWHM) of each energy point. Then, based on the adaptive SNIP algorithm, the algorithm iterates from the high-energy end to the low-energy end of the preprocessed energy spectrum data. In each iteration, the value of each channel is replaced with the lower value of the current channel and its left and right adjacent channels (the window size changes with the number of iterations). Smoothing is then performed until the determined number of iterations is reached to obtain the estimated background energy spectrum data. Finally, the processed energy spectrum data is subtracted from the background energy spectrum data to obtain the net energy spectrum data.

[0038] Step S103: Perform nuclide identification processing based on net energy spectrum data to obtain nuclide identification results, and calculate nuclide activity concentration based on nuclide identification results and nuclear radiation detection data to obtain nuclide activity concentration.

[0039] Optionally, in order to effectively distinguish overlapping peaks, since peak finding has limited ability to find weak peaks, this application performs nuclide identification processing based on the peak finding method of the nuclide library after obtaining the net energy spectrum data, obtains the nuclide identification result, and then calculates the nuclide activity concentration based on the obtained nuclide identification result and nuclear radiation detection data to obtain the nuclide activity concentration.

[0040] In an optional embodiment of this application, nuclide identification processing is performed based on net energy spectrum data to obtain nuclide identification results, including: The net energy spectrum data is convolved using a Gaussian symmetric zero-area convolution kernel, and the position of each candidate peak is determined based on the convolved net energy spectrum data. The nuclide matching is performed on each candidate peak position based on the Bayesian posterior probability algorithm to obtain the nuclide identification result, which includes the nuclide type and the characteristic peak information corresponding to each nuclide type.

[0041] Optionally, the obtained net energy spectrum data is convolved using a symmetric zero-area Gaussian convolution kernel, and then the position of each candidate peak is determined based on the convolved net energy spectrum data. The new convolution kernel function is shown below:

[0042] Furthermore, based on the Bayesian posterior probability algorithm, nuclide matching is performed on each candidate peak position to obtain the nuclide type and the characteristic peak information corresponding to each nuclide type.

[0043] In an optional embodiment of this application, determining the position of each candidate peak based on the convolved net energy spectrum data includes: The position of each candidate peak is determined based on the position of the zero-crossing point in the net energy spectrum data after convolution; Multiple Gaussian fitting is used to perform peak decomposition based on the position of each candidate peak to determine each candidate peak position.

[0044] Optionally, the symmetric zero-area Gaussian convolution kernel function has symmetry and zero-area characteristics, and has zero response (i.e. zero crossing point) to DC signals and first-order linear signals (usually the main components of the background). Since the symmetric zero-area Gaussian convolution kernel function has a Gaussian form, it has an enhancement effect on nuclide peaks (which are also Gaussian in form). It can continuously adjust the variance p with the constantly changing FWHM (Full Width at Half Maximum) in the energy spectrum to achieve optimal peak detection and obtain the position of each candidate peak.

[0045] In practical applications, overlapping peaks may occur. For overlapping peaks, a peak-finding algorithm based on nuclides can be used to traverse the peaks of each nuclide. Based on the position of each candidate peak, the peak boundary is searched near that position, and the total peak area, background area, and net area are calculated. Then, the uncertainty of the total area and background area is calculated according to the Poisson statistical model. Finally, the uncertainty of the net area is calculated according to the error transfer function. If the ratio of the uncertainty of the net area to the net area is less than a preset threshold, it is determined that a nuclide peak exists at that location.

[0046] The peak area can be calculated using Gaussian fitting. If there are no other peaks around a peak, a single Gaussian fit is used; otherwise, multiple neighboring peaks are fitted simultaneously using a multi-Gaussian fit. The fitting formula is as follows:

[0047]

[0048] in, M To simultaneously fit the number of peaks, The amplitude of each peak, The background for fitting the quadratic term, For each peak position, the fitting problem is transformed into The problem is to minimize the function, and when optimizing the global function, the gradient descent method based on the partial derivative model is used for optimization.

[0049] A typical problem with gradient descent is that the algorithm may converge to a local minimum rather than the global minimum. To ensure that the algorithm converges to the global minimum to the greatest extent possible, the most effective method is to find a suitable initial value for the parameters. That is, to find a set of base quadratic coefficients that are closest to the optimal value. b 0 , b 1 b2 , The initial value is used to start the gradient descent algorithm. b 0 , b 1 The initial value of b2 can be obtained from the previous baseline estimate. , The initial values ​​can be obtained from the peak finding algorithm, thus providing a good set of initial values, which ensures the convergence of the algorithm.

[0050] In an optional embodiment of this application, nuclide matching is performed on each candidate peak position based on a Bayesian posterior probability algorithm to obtain nuclide identification results, including: Determine the energy value corresponding to each candidate peak position, and match the energy value corresponding to each candidate peak position with the theoretical characteristic energy of nuclides in the preset nuclide library to obtain each candidate nuclide; For each candidate nuclide, the posterior probability of each candidate nuclide is determined based on the Bayesian posterior probability algorithm, and the nuclide identification result is obtained based on the posterior probability of each candidate nuclide.

[0051] Optionally, the energy and area of ​​each candidate peak can be obtained based on the peak finding results, and then the energy value corresponding to each candidate peak position can be determined based on the energy and area of ​​each candidate peak. For each candidate peak position, the energy value corresponding to the candidate peak position can be matched with the theoretical characteristic energy of the nuclides in the preset nuclide library to obtain each candidate nuclide corresponding to the candidate peak position. For each candidate nuclide, the posterior probability of each candidate nuclide is determined based on the Bayesian posterior probability algorithm, and the nuclide identification result is obtained based on the posterior probability of each candidate nuclide, such as taking the candidate nuclide corresponding to the posterior probability as the nuclide identification result corresponding to the candidate peak position.

[0052] The posterior probability of each candidate nuclide, determined by the Bayesian posterior probability algorithm, can be calculated using the following formula:

[0053]

[0054] in, Indicates the first nuclide in the nuclide library i One nuclide, This indicates the collection of energy spectrum data. This indicates that the energy spectrum data contains nuclides. i The probability, Prior probabilities can be determined based on actual prior knowledge. For example, the probability of naturally occurring nuclides such as K-40 and TH-232 being present is relatively high, so their prior probabilities can be appropriately increased. Conversely, the probability of detecting nuclides with short half-lives (such as F-18) is relatively low, so their prior probabilities can be appropriately decreased. This represents the probability that nuclide i can lead to the acquisition of a known energy spectrum. It is mainly composed of two factors: 1. Nuclide i The degree of matching between each full-energy peak and the energy peak in the data energy spectrum is measured by the full-energy peak's half-width at half-maximum. The closer the two peak positions are, the higher the degree of matching and the higher the factor coefficient. 2. Nuclide i The degree of matching between the branch ratio of each full-energy peak and the energy peak area ratio in the corresponding data energy spectrum is considered, while also taking into account the influence of detector efficiency on the energy peak area. The closer the two ratios are, the higher the degree of matching, and the higher the factor coefficient.

[0055] In optional embodiments of this application, the nuclear radiation reconnaissance data includes alpha aerosol concentration and surface contamination measurements; the nuclide activity concentration includes gamma nuclide activity concentration, alpha aerosol activity concentration, and surface contamination activity concentration; and the nuclide activity concentration is calculated based on the nuclide identification results and the nuclear radiation reconnaissance data to obtain the nuclide activity concentration, including: The detection efficiency and sampling time corresponding to the acquisition of the raw pulse signal of the γ energy spectrum are obtained, and the activity concentration of the nuclide is calculated based on the characteristic peak information, detection efficiency and sampling time corresponding to each type of nuclide to obtain the γ nuclide activity concentration; Obtain the sampling flow rate and filtration efficiency corresponding to the α aerosol concentration, and obtain the α aerosol activity concentration based on the α aerosol concentration, sampling flow rate and filtration efficiency; The detector efficiency and detection area corresponding to the surface contamination measurement value are obtained, and the surface contamination activity concentration is obtained based on the surface contamination measurement value, detector efficiency, and detection area.

[0056] Optionally, the determined nuclide activity concentrations include gamma nuclide activity concentration, alpha aerosol activity concentration, and surface contamination activity concentration. For gamma nuclide activity concentration, the detection efficiency and sampling time can be obtained. The detection efficiency is the detection efficiency used when acquiring the raw gamma energy spectrum pulse signal, and the sampling time is the time taken to acquire the raw gamma energy spectrum pulse signal. Further, for the nuclide type, the nuclide activity concentration can be calculated using the following formula based on the characteristic peak information corresponding to that nuclide type, the detection efficiency, and the sampling time, to obtain the gamma nuclide activity concentration.

[0057] The gamma nuclide activity concentration = net peak area / (detection efficiency × sampling time) The gamma nuclide activity concentration is relative to the detector position, representing the activity corresponding to the radiation intensity produced by the nuclide at the detector position. The net peak area is the net peak area calculated from the characteristic peak information corresponding to the nuclide type.

[0058] To obtain the alpha aerosol activity concentration, the sampling flow rate and filtration efficiency used when collecting the alpha aerosol concentration can be obtained. Then, the net alpha count rate can be obtained based on the alpha aerosol concentration. Finally, the alpha aerosol activity concentration can be obtained based on the net alpha count rate, sampling flow rate, and filtration efficiency using the following formula.

[0059] Alpha aerosol activity concentration = net alpha count rate / (sampling flow rate × filtration efficiency) The specific implementation method for obtaining the net α count rate based on the α aerosol concentration can be based on existing technologies, and will not be elaborated further in this application.

[0060] To determine the surface contaminant activity concentration, the detector efficiency and detection area used to measure the surface contaminant can be obtained. Then, based on the surface contaminant measurement, detector efficiency, and detection area, the surface contaminant activity concentration can be calculated using the following formula.

[0061] Surface contaminant activity concentration = Surface contaminant measurement / (Detector efficiency × Detection area) The efficiency of the detector used to measure surface contamination can be directly obtained from the distance measured during detection.

[0062] Step S104: Based on geographical data, meteorological data and radionuclide activity concentration, source term inversion is performed to obtain leakage source term information, which includes leakage source strength and leakage location.

[0063] Optionally, after determining the nuclide activity concentration, source term inversion can be performed based on previously acquired geographical data, meteorological data, and nuclide activity concentration to obtain the leakage source strength and leakage location.

[0064] In optional embodiments of this application, source term inversion is performed based on geographical data, meteorological data, and radionuclide activity concentration to obtain leakage source term information, including: The diffusion coefficient value was calculated based on geographical and meteorological data, and a Gaussian multiplying smoke model was constructed based on the diffusion coefficient value. The initial leakage source term information is obtained, and the predicted nuclide activity concentration is obtained based on the initial leakage source term information and the Gaussian multipuff model. The initial leakage source term information includes the assumed leakage source strength and the assumed leakage location. Based on the predicted nuclide activity concentration and the nuclide activity concentration processed by Kalman filtering, leakage source term information is obtained.

[0065] Optionally, the diffusion coefficient value can be determined based on existing atmospheric stability classification tables using geographic and meteorological data. , and ( , This represents the atmospheric lateral diffusion coefficient. (This represents the atmospheric vertical diffusion coefficient). Then, based on the determined diffusion coefficient value, a Gaussian multiplying smoke model is constructed, as shown below:

[0066] in, These represent the coordinates of the calculation point relative to the leak point. Indicates in t At any point in time and space （ ） The predicted nuclide concentration at the release location (in the source term parameters) is given, with the leak point at (0, 0, 0). Q This represents the source strength in the initial source term parameters. u This represents the average wind speed at the source elevation. This represents the atmospheric lateral diffusion coefficient. Indicates the atmospheric vertical diffusion coefficient. H This indicates the effective height of the source strength in the initial term parameters.f d This is a radioactive decay term. f q For deposition attenuation term, Λ The wet deposition coefficient is... The settling velocity is the dry deposition velocity of aerosol particles.

[0067] Optional, f d It satisfies the radionuclide decay formula. f q The deposition attenuation term satisfies the following formula:

[0068] Λ The wet deposition coefficient satisfies the following formula:

[0069] in, I The value represents the rainfall intensity (mm / h). a and b are empirical coefficients, assigned values ​​based on whether the released substance contains iodine or not. For example, for iodine-containing substances, a can be taken as a = 8 × 10⁻⁶. -5 b = 0.6, while for substances without iodine, a = 1.2 × 10⁻⁶. -4 b=0.5.

[0070] Furthermore, the assumed leakage source strength and assumed leakage location are input into the Gaussian multi-puff model, at which point the predicted nuclide activity concentration can be obtained. Then, based on the predicted nuclide activity concentration and the nuclide activity concentration, the leakage source term information is obtained through Kalman filtering.

[0071] The Kalman filtering process includes a prediction step and an update step. The prediction step is expressed as follows:

[0072]

[0073] in, Information on the source of the leak at the current moment. This refers to the leak source information from the previous moment. Let be the error covariance matrix. Let be the error covariance matrix of the previous time step. Let Q be the state transition matrix and Q be the process noise covariance matrix.

[0074] The update steps are as follows:

[0075]

[0076]

[0077] in, Here is the Kalman gain matrix. Let be the error covariance matrix. Let R be the observation matrix, and let R be the observation noise covariance matrix. Information on the source of the leak at the current moment. denoted as , where I is the nuclide activity concentration.

[0078] Step S105: Based on meteorological data, leakage source term information and radionuclide activity concentration, obtain the hazard range prediction results. The hazard range prediction results include the predicted activity concentration field, dose rate field and pollution boundary of the radionuclide.

[0079] In optional embodiments of this application, based on meteorological data, leakage source term information, and radionuclide activity concentration, a hazard range prediction result is obtained, including: The leak source information and meteorological data were input into the Gaussian multi-puff model to obtain the predicted activity concentration field of the nuclides; Obtain the radionuclide dose conversion factor, and calculate the dose rate field based on the radionuclide dose conversion factor and the predicted activity concentration field of the radionuclide; Obtain the pollution level threshold, and based on the dose rate field and the pollution level threshold, obtain the pollution boundary.

[0080] Optionally, once the leak source term information is obtained, the specific location of the nuclear leak and the rate of radioactive material release can be known. Then, the contamination range can be predicted and determined based on the specific location and the rate of radioactive material release. In order to better understand the dangerous range caused by the nuclear leak, the obtained leak source term information and meteorological data can be input into a preset Gaussian diffusion model to obtain the predicted activity concentration field of the nuclide. Then, a three-dimensional grid data can be formed based on the predicted activity concentration field, so as to better understand the specific dangerous range.

[0081] Furthermore, the radionuclide dose conversion factor can be obtained, and then the dose rate field can be obtained by multiplying the radionuclide dose conversion factor and the predicted activity concentration field of the radionuclide. This dose rate field can be referenced when analyzing the degree of danger, thereby better reducing the degree of danger.

[0082] Furthermore, to ensure that the adopted treatment methods are more adapted to the actual conditions of the nuclear leak area, contamination level thresholds can be obtained, and then isosurfaces can be determined based on the relationship between the dose rate field and the contamination level thresholds. For example, a dose rate field > 10 mSv / h is considered heavily contaminated, a dose rate field of 1–10 mSv / h is considered moderately contaminated, and a dose rate field < 1 mSv / h is considered lightly contaminated. Further, the maximum impact distance of the nuclear leak can be determined based on the determined isosurfaces, i.e., the distance at which the downwind contaminant concentration drops to the limit (e.g., 0.1 Bq / m³).

[0083] This application provides a device for predicting the extent of a nuclear facility leak hazard, such as... Figure 8 As shown, the device may include: a reconnaissance data acquisition module 801, an energy spectrum preprocessing module 802, a nuclide processing module 803, a source term inversion module 804, and a hazard range prediction module 805, wherein, The reconnaissance data acquisition module is used to acquire reconnaissance data corresponding to the nuclear facility leakage area. The reconnaissance data includes raw gamma-ray energy spectrum pulse signals, meteorological data, geographical data, and nuclear radiation reconnaissance data. The energy spectrum preprocessing module is used to preprocess the original γ energy spectrum pulse signal to obtain the processed energy spectrum data, and to perform background subtraction processing on the processed energy spectrum data based on the adaptive SNIP algorithm to obtain the net energy spectrum data. The energy spectrum preprocessing includes digital trapezoidal shaping processing and symmetrical zero-area shaping processing. The nuclide processing module is used to perform nuclide identification processing based on net energy spectrum data to obtain nuclide identification results, and to calculate nuclide activity concentration based on nuclide identification results and nuclear radiation detection data to obtain nuclide activity concentration. The source term inversion module is used to perform source term inversion based on geographical data, meteorological data, and radionuclide activity concentration to obtain leakage source term information, which includes leakage source strength and leakage location. The hazard range prediction module is used to obtain hazard range prediction results based on meteorological data, leakage source term information and radionuclide activity concentration. The hazard range prediction results include the predicted activity concentration field, dose rate field and contamination boundary of the radionuclide.

[0084] Optionally, when the energy spectrum preprocessing module performs energy spectrum preprocessing on the original γ-energy spectrum pulse signal to obtain processed energy spectrum data, it is specifically used for: Digital signal processing is performed on the original energy spectrum pulse signal to obtain the trapezoidal pulse corresponding to the original γ energy spectrum pulse signal. The trapezoidal pulse is then stacked based on the degree of overlap of each pulse in the trapezoidal pulse to obtain the processed trapezoidal pulse. Obtain the convolution kernel, and perform convolution processing on the processed trapezoidal pulses according to the convolution kernel to obtain the convolved trapezoidal pulses. The convolution kernel has a symmetrical shape and an area of ​​zero. The pulse amplitudes of the convolved trapezoidal pulses are statistically analyzed to obtain the processed energy spectrum data.

[0085] Optionally, when the energy spectrum preprocessing module performs background subtraction on the processed energy spectrum data based on the adaptive SNIP algorithm to obtain the net energy spectrum data, it is specifically used for: The half-width at half-height (WHM) of each energy point is calculated using the energy scale formula, and the number of iterations is determined based on the WHM of each energy point. The processed energy spectrum data is iterated from the high energy end to the low energy end based on the adaptive SNIP algorithm until the number of iterations is reached, so as to obtain the estimated background energy spectrum data. Based on the processed energy spectrum data and the background energy spectrum data, the net energy spectrum data is obtained.

[0086] Optionally, when the nuclide processing module performs nuclide identification processing based on the net energy spectrum data to obtain the nuclide identification result, it is specifically used for: The net energy spectrum data is convolved using a Gaussian symmetric zero-area convolution kernel, and the position of each candidate peak is determined based on the convolved net energy spectrum data. The nuclide matching is performed on each candidate peak position based on the Bayesian posterior probability algorithm to obtain the nuclide identification result, which includes the nuclide type and the characteristic peak information corresponding to each nuclide type.

[0087] Optionally, the nuclide processing module, when determining the position of each candidate peak based on the net energy spectrum data after convolution, is specifically used for: The position of each candidate peak is determined based on the position of the zero-crossing point in the net energy spectrum data after convolution; Multiple Gaussian fitting is used to perform peak decomposition based on the position of each candidate peak to determine each candidate peak position.

[0088] Optionally, when the nuclide processing module performs nuclide matching on each candidate peak position based on the Bayesian posterior probability algorithm to obtain the nuclide identification result, it is specifically used for: Determine the energy value corresponding to each candidate peak position, and match the energy value corresponding to each candidate peak position with the theoretical characteristic energy of nuclides in the preset nuclide library to obtain each candidate nuclide; For each candidate nuclide, the posterior probability of each candidate nuclide is determined based on the Bayesian posterior probability algorithm, and the nuclide identification result is obtained based on the posterior probability of each candidate nuclide.

[0089] Optionally, the nuclear radiation reconnaissance data includes alpha aerosol concentration and surface contamination measurements; the nuclide activity concentration includes gamma nuclide activity concentration, alpha aerosol activity concentration, and surface contamination activity concentration. When the nuclide processing module calculates the nuclide activity concentration based on the nuclide identification results and nuclear radiation reconnaissance data, it is specifically used for: The detection efficiency and sampling time corresponding to the acquisition of the raw pulse signal of the γ energy spectrum are obtained, and the activity concentration of the nuclide is calculated based on the characteristic peak information, detection efficiency and sampling time corresponding to each type of nuclide to obtain the γ nuclide activity concentration; Obtain the sampling flow rate and filtration efficiency corresponding to the α aerosol concentration, and obtain the α aerosol activity concentration based on the α aerosol concentration, sampling flow rate and filtration efficiency; The detector efficiency and detection area corresponding to the surface contamination measurement value are obtained, and the surface contamination activity concentration is obtained based on the surface contamination measurement value, detector efficiency, and detection area.

[0090] Optionally, when the source term inversion module performs source term inversion based on geographical data, meteorological data, and radionuclide activity concentration to obtain leakage source term information, it is specifically used for: The diffusion coefficient value was calculated based on geographical and meteorological data, and a Gaussian multiplying smoke model was constructed based on the diffusion coefficient value. The initial leakage source term information is obtained, and the predicted nuclide activity concentration is obtained based on the initial leakage source term information and the Gaussian multipuff model. The initial leakage source term information includes the assumed leakage source strength and the assumed leakage location. Based on the predicted nuclide activity concentration and the nuclide activity concentration processed by Kalman filtering, leakage source term information is obtained.

[0091] Optionally, when the hazard range prediction module obtains the hazard range prediction result based on meteorological data, leakage source term information, and radionuclide activity concentration, it is specifically used for: The leak source term information and meteorological data are input into a preset Gaussian diffusion model to obtain the predicted activity concentration field of the nuclide; Obtain the radionuclide dose conversion factor, and calculate the dose rate field based on the radionuclide dose conversion factor and the predicted activity concentration field of the radionuclide; Obtain the pollution level threshold, and based on the dose rate field and the pollution level threshold, obtain the pollution boundary.

[0092] The device for predicting the hazard range of a nuclear facility leak in this embodiment can execute the method for predicting the hazard range of a nuclear facility leak as shown in the embodiment of this application. The implementation principle is similar and will not be described again here.

[0093] This application provides an electronic device, which includes: a processor; and a memory configured to store machine-readable instructions that, when executed by the processor, cause the processor to perform a method for predicting the extent of a nuclear facility leak hazard.

[0094] This application provides an electronic device, such as... Figure 9 As shown, Figure 9The illustrated electronic device includes a processor 2001 and a memory 2003. The processor 2001 and the memory 2003 are connected, for example, via a bus 2002. Optionally, the electronic device 2000 may further include a transceiver 2004. It should be noted that in practical applications, the transceiver 2004 is not limited to one type, and the structure of this electronic device 2000 does not constitute a limitation on the embodiments of this application.

[0095] Processor 2001 may be a CPU, a general-purpose processor, a DSP, an ASIC, an FPGA, or other programmable logic device, transistor logic device, hardware component, or any combination thereof. It may implement or execute the various exemplary logic blocks, modules, and circuits described in conjunction with the disclosure of this application. Processor 2001 may also be a combination that implements computational functions, such as including one or more microprocessor combinations, a combination of a DSP and a microprocessor, etc.

[0096] Bus 2002 may include a pathway for transmitting information between the aforementioned components. Bus 2002 may be a PCI bus or an EISA bus, etc. Bus 2002 can be divided into address bus, data bus, control bus, etc. For ease of representation, Figure 9 The bus is represented by a single thick line, but this does not mean that there is only one bus or one type of bus.

[0097] The memory 2003 may be ROM or other type of static storage device capable of storing static information and instructions, RAM or other type of dynamic storage device capable of storing information and instructions, or EEPROM, CD-ROM or other optical disc storage, optical disc storage (including compressed optical discs, laser discs, optical discs, digital universal optical discs, Blu-ray discs, etc.), magnetic disk storage media or other magnetic storage devices, or any other medium capable of carrying or storing desired program code in the form of instructions or data structures and accessible by a computer, but not limited thereto.

[0098] The memory 2003 stores the application code that executes the scheme of this application, and its execution is controlled by the processor 2001. The processor 2001 executes the application code stored in the memory 2003 to implement... Figure 8 The illustrated embodiment provides the operation of a device for predicting the extent of a nuclear facility leak hazard.

[0099] It should be understood that although the steps in the flowcharts of the accompanying figures are shown sequentially as indicated by the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the accompanying figures may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times, and their execution order is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the sub-steps or stages of other steps.

[0100] The above description is only a partial embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for predicting the hazard range of a nuclear facility leak, characterized in that, include: Acquire reconnaissance data corresponding to the leak area of ​​the nuclear facility, including raw gamma-ray energy spectrum pulse signals, meteorological data, geographical data, and nuclear radiation reconnaissance data; The original γ-energy spectrum pulse signal is subjected to energy spectrum preprocessing to obtain processed energy spectrum data. The processed energy spectrum data is then subjected to background subtraction processing based on the adaptive SNIP algorithm to obtain net energy spectrum data. The energy spectrum preprocessing includes digital trapezoidal forming processing and symmetrical zero-area forming processing. The nuclide identification process is performed based on the net energy spectrum data to obtain the nuclide identification result. The nuclide activity concentration is then calculated based on the nuclide identification result and the nuclear radiation detection data to obtain the nuclide activity concentration. Source term inversion is performed based on the geographical data, the meteorological data, and the radionuclide activity concentration to obtain leakage source term information, which includes leakage source strength and leakage location. Based on the meteorological data, the leakage source information, and the radionuclide activity concentration, a hazard range prediction result is obtained, which includes the predicted activity concentration field, dose rate field, and pollution boundary of the radionuclide.

2. The method according to claim 1, characterized in that, The step of performing energy spectrum preprocessing on the original γ-energy spectrum pulse signal to obtain processed energy spectrum data includes: The original energy spectrum pulse signal is subjected to digital signal processing to obtain a trapezoidal pulse corresponding to the original γ energy spectrum pulse signal. The trapezoidal pulse is then stacked based on the degree of overlap of each pulse in the trapezoidal pulse to obtain the processed trapezoidal pulse. Obtain the convolution kernel, and perform convolution processing on the processed trapezoidal pulse according to the convolution kernel to obtain the convolved trapezoidal pulse. The convolution kernel is symmetrical and has an area of ​​zero. The pulse amplitude of the convolved trapezoidal pulse is statistically analyzed to obtain the processed energy spectrum data.

3. The method according to claim 1, characterized in that, The background subtraction process performed on the processed energy spectrum data based on the adaptive SNIP algorithm yields net energy spectrum data, including: The half-width at half-height (WHM) of each energy point is calculated using the energy scale formula, and the number of iterations is determined based on the WHM of each energy point. The processed energy spectrum data is iterated from the high energy end to the low energy end based on the adaptive SNIP algorithm until the number of iterations is reached, so as to obtain the estimated background energy spectrum data. The net energy spectrum data is obtained based on the processed energy spectrum data and the background energy spectrum data.

4. The method according to claim 1, characterized in that, The step of performing nuclide identification processing based on the net energy spectrum data to obtain nuclide identification results includes: The net energy spectrum data is convolved using a Gaussian symmetric zero-area convolution kernel, and the position of each candidate peak is determined based on the convolved net energy spectrum data. Nuclide matching is performed on each candidate peak position based on the Bayesian posterior probability algorithm to obtain the nuclide identification result, which includes the nuclide type and the characteristic peak information corresponding to each nuclide type.

5. The method according to claim 4, characterized in that, The step of determining the position of each candidate peak based on the net energy spectrum data after convolution includes: The position of each candidate peak is determined based on the position of the zero-crossing point in the net energy spectrum data after convolution; Based on the position of each candidate peak, multiple Gaussian fitting is used to perform peak decomposition to determine each candidate peak position.

6. The method according to claim 4, characterized in that, The method of performing nuclide matching on each candidate peak position based on the Bayesian posterior probability algorithm to obtain the nuclide identification result includes: Determine the energy value corresponding to each candidate peak position, and match the energy value corresponding to each candidate peak position with the theoretical characteristic energy of nuclides in a preset nuclide library to obtain each candidate nuclide; For each candidate nuclide, the posterior probability of each candidate nuclide is determined based on a Bayesian posterior probability algorithm, and the nuclide identification result is obtained based on the posterior probability of each candidate nuclide.

7. The method according to claim 4, characterized in that, The nuclear radiation reconnaissance data includes alpha aerosol concentration and surface contamination measurements. The nuclide activity concentration includes gamma nuclide activity concentration, alpha aerosol activity concentration, and surface contamination activity concentration. The calculation of nuclide activity concentration based on the nuclide identification results and the nuclear radiation reconnaissance data to obtain the nuclide activity concentration includes: The detection efficiency and sampling time corresponding to the acquisition of the raw pulse signal of the γ energy spectrum are obtained, and the activity concentration of the nuclide is calculated based on the characteristic peak information corresponding to each type of nuclide, the detection efficiency and the sampling time to obtain the activity concentration of the γ nuclide; The sampling flow rate and filtration efficiency corresponding to the collection of the α aerosol concentration are obtained, and the α aerosol activity concentration is obtained based on the α aerosol concentration, the sampling flow rate and the filtration efficiency. The detector efficiency and detection area corresponding to the surface contamination measurement value are obtained, and the surface contamination activity concentration is obtained based on the surface contamination measurement value, the detector efficiency, and the detection area.

8. The method according to claim 1, characterized in that, The process of source term inversion based on the geographical data, the meteorological data, and the radionuclide activity concentration to obtain leakage source term information includes: The diffusion coefficient value is calculated based on the geographical data and the meteorological data, and a Gaussian multiplying smoke model is constructed based on the diffusion coefficient value. Obtain initial leakage source term information, and based on the initial leakage source term information and the Gaussian multipuff model, obtain the predicted nuclide activity concentration. The initial leakage source term information includes the assumed leakage source strength and the assumed leakage location. The leakage source term information is obtained by processing the predicted nuclide activity concentration and the nuclide activity concentration using Kalman filtering.

9. The method according to claim 8, characterized in that, The prediction of the hazard range based on the meteorological data, the leakage source information, and the radionuclide activity concentration includes: The leakage source term information and the meteorological data are input into a preset Gaussian diffusion model to obtain the predicted activity concentration field of the nuclide; Obtain the radionuclide dose conversion factor, and calculate the dose rate field based on the radionuclide dose conversion factor and the predicted activity concentration field of the radionuclide; Obtain the pollution level threshold, and based on the dose rate field and the pollution level threshold, obtain the pollution boundary.

10. A device for predicting the hazard range of a nuclear facility leak, characterized in that, include: The reconnaissance data acquisition module is used to acquire reconnaissance data corresponding to the leak area of ​​the nuclear facility. The reconnaissance data includes raw γ-ray energy spectrum pulse signals, meteorological data, geographical data, and nuclear radiation reconnaissance data. The energy spectrum preprocessing module is used to preprocess the original γ energy spectrum pulse signal to obtain processed energy spectrum data, and to perform background subtraction processing on the processed energy spectrum data based on the adaptive SNIP algorithm to obtain net energy spectrum data. The energy spectrum preprocessing includes digital trapezoidal shaping processing and symmetrical zero-area shaping processing. The nuclide processing module is used to perform nuclide identification processing based on the net energy spectrum data to obtain nuclide identification results, and to calculate the nuclide activity concentration based on the nuclide identification results and the nuclear radiation detection data to obtain the nuclide activity concentration. The source term inversion module is used to perform source term inversion based on the geographical data, the meteorological data and the radionuclide activity concentration to obtain leakage source term information, which includes leakage source strength and leakage location. The hazard range prediction module is used to obtain a hazard range prediction result based on the meteorological data, the leakage source information and the radionuclide activity concentration. The hazard range prediction result includes the predicted activity concentration field, dose rate field and pollution boundary of the radionuclide.

Images