Method and device for preparing radionuclide in soil

Abstract

The invention relates to the technical field of nuclear radiation detection, and discloses a method and a device for preparing radionuclides in soil, and the device is characterized in that layered detection modules separated by gradient cascade shielding assemblies are arranged in a probe unit. The method is executed based on a device comprising a probe unit, a data acquisition and processing unit and an upper computer terminal, and comprises the following steps: extracting a characteristic secondary radiation signal generated by excitation of a shielding assembly, and quantitatively stripping interlayer crosstalk by using the signal to obtain a net energy spectrum of the layer; calculating a Compton scattering ratio and inverting soil in-situ equivalent density; calculating a transmission efficiency correction factor, and adjusting a pre-stored standard response matrix to reconstruct a dynamic response matrix; and constructing a linear analysis model based on the net energy spectrum of the layer and the dynamic response matrix, and resolving the nuclide activity. According to the invention, crosstalk is removed through physical tracing and density-based adaptive calibration is carried out, so that the problem of errors caused by large layered measurement interference and environmental differences is effectively solved.

Description

Technical Field

[0001] This invention relates to the field of power system supply and demand interaction technology, specifically to a method and apparatus for preparing radioactive nuclides from soil. Background Technology

[0002] Monitoring the vertical distribution of radionuclides in soil is crucial for environmental assessment and pollution remediation. Existing in-situ monitoring technologies typically employ highly integrated portable devices, such as stratified measurement equipment that includes a host computer system, a digital multichannel system, and a probe system. The core of these devices lies in the probe system, which contains multiple detection modules spaced along its length (e.g., cadmium zinc telluride detectors). These modules interact with gamma rays to generate pulse signals, which are then digitally processed to acquire energy spectrum data. For data analysis, current techniques often employ full-spectrum least squares analysis algorithms. Based on a pre-acquired detector response function matrix across the entire energy range, interpolation is used to calculate the response functions at different energies. Then, inversion calculations are used to separate overlapping peaks to identify the types and concentrations of radionuclides.

[0003] While the aforementioned layered detection and full-spectrum analysis techniques have improved the convenience and resolution of measurements to some extent, they still have significant limitations in practical applications in complex soil environments. Firstly, gamma rays exhibit isotropic emission characteristics and strong penetrating power; high-energy rays in deep soil can easily penetrate the physical spacing between detection modules and enter adjacent detection areas. Although existing spacing settings can provide some physical shielding, they lack an effective marking and stripping mechanism for these interlayer penetrating rays. This results in the energy spectrum data at a certain depth often being mixed with signals from adjacent soil layers, causing aliasing and distortion of the vertical distribution information of nuclides.

[0004] Secondly, existing full-spectrum analysis algorithms mainly rely on response matrices established under standard laboratory conditions, or simply perform mathematical interpolation based on the relationship between energy and full width at half maximum (FWHM). This approach ignores the differences in the physical properties of the soil medium itself. The in-situ density of soil at the actual measurement site is significantly affected by compaction, moisture content, and soil type, often deviating significantly from the standard density calibrated in the laboratory. Changes in medium density directly affect the transmission efficiency and self-absorption effect of X-rays in the soil. If a response matrix with fixed parameters or only energy interpolation is directly used, it is impossible to dynamically correct the systematic errors introduced by changes in medium density, thus affecting the accuracy of quantitative analysis.

[0005] Furthermore, when processing soil energy spectrum data with low activity or high background, conventional full-spectrum least squares algorithms typically assume a uniform distribution of data errors and lack a weighting mechanism for the statistical fluctuation characteristics of different energy channels. Simultaneously, purely mathematical inversion calculations can sometimes output physically meaningless negative results, leading to unstable results in trace nuclide analysis and failing to meet the requirements of high-precision in-situ measurements. Summary of the Invention

[0006] To address the shortcomings of existing technologies, this invention provides a method and apparatus for preparing radionuclides in soil. It aims to solve the problems in existing technologies, such as the difficulty in effectively and quantitatively removing interlayer X-ray crosstalk during layered measurements, the inability of fixed-parameter response matrices to adapt to the differences in transmission efficiency caused by in-situ soil density changes, and the lack of statistical weight allocation and poor solution stability of traditional full-spectrum inversion algorithms under low activity conditions.

[0007] To achieve the above objectives, the first aspect of the present invention provides a method for preparing radioactive nuclides in soil. This method is executed using a device comprising a probe unit, a data acquisition and processing unit, and a host computer terminal. The probe unit contains multiple detection modules distributed along the depth direction, and adjacent detection modules are separated by a gradient-cascaded shielding assembly. The method includes the following steps: The system controls each detection module to acquire raw energy spectrum data and obtains characteristic secondary radiation signals generated by the gradient cascaded shielding components. Crosstalk components are removed from the original energy spectrum data using characteristic secondary radiation signals, thereby obtaining the net energy spectrum data of this layer; Calculate the Compton scattering ratio of the net energy spectrum data of this layer, and inversely determine the in-situ equivalent density of the soil at the current detection layer based on this ratio; The transmission efficiency correction factor is calculated based on the in-situ equivalent density, and the pre-stored laboratory standard response matrix is ​​adjusted using the transmission efficiency correction factor to reconstruct the dynamic response matrix. A linear analytical model was constructed based on the net energy spectrum data and dynamic response matrix of this layer to calculate the types and activities of radionuclides in the soil.

[0008] Preferably, the aforementioned gradient cascaded shielding assembly is constructed using layered composite materials to block straight rays propagating along the probe unit axis and to generate characteristic secondary radiation signals with characteristic energy when excited by large-angle oblique rays penetrating the edge of the assembly.

[0009] In one specific embodiment, the process of removing crosstalk components from the original energy spectrum data includes: extracting the region of interest (ROI) corresponding to the characteristic secondary radiation signal in the original energy spectrum data; calculating the net count rate of the ROI using the tracer signal extraction formula; multiplying the net count rate by a preset crosstalk energy spectrum shape factor to reconstruct the crosstalk energy spectrum characterizing the crosstalk information; and finally subtracting the crosstalk energy spectrum from the original energy spectrum data to obtain the net energy spectrum data of this layer.

[0010] Preferably, the process of calculating the Compton scattering ratio of the net energy spectrum data of this layer includes defining the integration interval, specifically: searching for the local maximum point of the high-energy reference peak in the net energy spectrum data of this layer, fitting the local maximum point and its neighborhood data with a Gaussian function to determine the center address of the reference peak and the standard deviation of the characteristic peak; using the center address of the reference peak as a benchmark, expanding the standard deviation of the characteristic peak to the left and right by a preset multiple to establish the starting and ending addresses of the full-energy peak integration interval; selecting a continuous spectrum region located to the left of the high-energy reference peak and having a flat feature as the Compton scattering feature window, and using the current energy gain coefficient and the zero-point energy intercept of the system, converting the preset physical energy range into the corresponding address interval to establish the starting and ending addresses of the Compton scattering feature window.

[0011] Furthermore, the process of calculating the Compton scattering ratio of the net energy spectrum data of this layer is carried out using the Compton scattering ratio calculation formula, specifically including: numerically integrating the counts within the Compton scattering characteristic window and the counts within the full-energy peak integration interval respectively; calculating the total background count within the full-energy peak integration interval, and subtracting the total background count from the count within the full-energy peak integration interval to obtain the net count of the full-energy peak; calculating the ratio of the sum of the counts within the Compton scattering characteristic window to the net count of the full-energy peak, and using this ratio as the Compton scattering ratio.

[0012] In one specific embodiment, the process of inverting the in-situ equivalent density of the soil at the current detection level includes: pre-establishing a correlation model between scattering ratio and soil density, which is obtained through calibration experiments in multiple sets of standard soil simulation modules with known densities; substituting the Compton scattering ratio into the correlation model, and using the density response model formula to calculate the in-situ equivalent density of the soil at the current detection level.

[0013] Preferably, the process of calculating the transmission efficiency correction factor includes: retrieving the corresponding mass attenuation coefficient from a preset mass attenuation coefficient database based on the center energy value corresponding to each channel in the response matrix to be reconstructed; calculating the difference between the in-situ equivalent density and the standard density used when generating the laboratory standard response matrix; and calculating the transmission efficiency correction factor for each energy point based on the Beer-Lambert law using the transmission efficiency correction factor calculation formula. The value of this factor is equal to the value of an exponential function with the natural constant as the base and the product of the negative mass attenuation coefficient, the effective detection radius of the probe unit, and the difference as the exponent.

[0014] Furthermore, the process of reconstructing the dynamic response matrix is ​​implemented using a matrix transformation formula, including: retrieving and parsing the characteristic main emission energy corresponding to each column vector in the laboratory standard response matrix; substituting the characteristic main emission energy into the calculation result of the transmission efficiency correction factor to obtain the corresponding scalar correction value; and performing the operation defined by the matrix transformation formula to multiply all elements of each column vector in the laboratory standard response matrix by the scalar correction value corresponding to that column vector to generate the dynamic response matrix.

[0015] Preferably, the process of constructing a linear analytical model includes: establishing a linear response model formula expressed by a system of linear equations, wherein the linear response model formula includes the net energy spectrum vector of the current layer, the dynamic response matrix, the nuclide activity vector to be determined, and the measurement error vector; the net energy spectrum vector of the current layer is equal to the matrix product of the dynamic response matrix and the nuclide activity vector plus the measurement error vector.

[0016] Furthermore, the process of calculating the types and activities of radionuclides in the soil is performed using a weighted least squares inversion formula, including: constructing a diagonal weight matrix, where the diagonal elements are configured as the reciprocals of the variances of the corresponding energy address counts; performing matrix operations of the weighted least squares inversion formula, i.e., calculating the transpose of the dynamic response matrix, the inverse of the product of the diagonal weight matrix and the dynamic response matrix, and then multiplying it sequentially by the transpose of the dynamic response matrix, the diagonal weight matrix, and the net energy spectrum data of the current layer to obtain the nuclide activity estimation vector; and using a non-negative least squares constrained algorithm to iteratively correct the nuclide activity estimation vector to obtain a non-negative solution vector.

[0017] A second aspect of the present invention provides an apparatus for preparing radionuclides in soil, configured to perform the method described in the first aspect above. The apparatus includes a probe unit, a data acquisition and processing unit, and a host computer terminal. The probe unit is constructed as a rod-shaped structure suitable for vertical insertion into the soil medium to be tested. Inside the probe unit, a detector array is spaced apart along its axial length. The detector array consists of multiple independent detection modules, and a gradient cascaded shielding component is arranged between adjacent detection modules. The gradient cascaded shielding component is configured to generate characteristic secondary radiation signals. The data acquisition and processing unit establishes an electrical connection with each detection module to convert analog pulse signals from the detection modules into digitized raw energy spectrum data. The host computer terminal is connected to the data acquisition and processing unit. This invention provides a method and apparatus for preparing radioactive nuclides from soil. It has the following beneficial effects: 1. This invention sets up a gradient cascaded shielding component between adjacent detection modules, and uses the characteristic secondary radiation signal generated by this component when excited by large-angle oblique rays as a physical tracer. By identifying this tracer signal, crosstalk components from adjacent soil layers can be quantitatively calculated and subtracted from the original energy spectrum data, effectively solving the problem of distinguishing between the signal of the current layer and the interference of adjacent layers in traditional layered measurements, and ensuring that the net energy spectrum data obtained for each layer truly reflects the nuclide distribution at that depth level.

[0018] 2. This invention inverts the in-situ equivalent density of soil by calculating the Compton scattering ratio of the net energy spectrum of the current layer, and calculates the transmission efficiency correction factor accordingly. It then adjusts the pre-stored laboratory standard response matrix element by element, establishing a dynamic mapping relationship between the laboratory environment and the field environment. This enables the measurement system to reconstruct the response matrix in real time according to the changes in the actual density of the soil to be measured, effectively overcoming the influence of density differences caused by soil compaction, moisture content, etc., on the self-absorption effect, and improving the measurement accuracy in complex soil environments.

[0019] 3. This invention constructs a linear analytical model based on the net energy spectrum and dynamic response matrix of the current layer, and uses a weighted least squares algorithm combined with non-negativity constraints for inversion solution. By introducing a weight matrix with elements of the reciprocal of the count variance, this method automatically reduces the weight of data with large statistical fluctuations during the solution process. At the same time, the non-negativity constraints avoid the occurrence of physically meaningless negative values ​​in the mathematical solution. This allows the algorithm to maintain high solution stability and accuracy even under conditions of low count rate or complex nuclide types. Attached Figure Description

[0020] Figure 1 This is a system framework diagram of the present invention; Figure 2 This is a schematic diagram of the method flow of the present invention.

[0021] 10. Probe unit; 110. Detection module; 120. Gradient cascaded shielding assembly; 121. Physical shielding substrate; 122. Energy relay layer; 123. Fluorescent tracer layer; 20. Data acquisition and processing unit; 210. High voltage power supply module; 220. Charge-sensitive preamplifier unit; 230. Shaping amplification unit; 240. Multichannel pulse amplitude analyzer; 30. Host computer terminal. Detailed Implementation

[0022] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0023] See attached document Figure 1 The present invention provides a device for preparing radionuclides in soil, which mainly includes a probe unit 10, a data acquisition and processing unit 20 and a host computer terminal 30.

[0024] The probe unit 10 is constructed as a rod-shaped structure suitable for vertical insertion into the soil medium to be tested, and its outer shell is made of a high-strength material with low radiation attenuation. Inside the probe unit 10, a detector array is spaced apart along its axial length. This detector array consists of multiple independent detection modules 110, distributed along the depth direction, corresponding to different soil detection layers at different depths. The detection modules 110 use cadmium zinc telluride crystals as radiation-sensitive elements, configured to respond to gamma rays and generate pulsed electrical signals proportional to the ray energy.

[0025] A gradient cascaded shielding assembly 120 is provided between adjacent detection modules 110. The gradient cascaded shielding assembly 120 is arranged in a ring structure inside the probe unit 10 to physically isolate the field of view of adjacent detection modules 110. The gradient cascaded shielding assembly 120 is constructed of layered composite material and is configured to block straight rays propagating along the axial direction of the probe unit 10, and to generate a characteristic secondary radiation signal of specific energy for large-angle oblique rays penetrating the edge of the assembly. This characteristic secondary radiation signal is used to mark interlayer crosstalk.

[0026] The data acquisition and processing unit 20 establishes an electrical connection with each detection module 110 and is configured to synchronously receive analog pulse signals from each detection module 110. The data acquisition and processing unit 20 integrates a preamplifier circuit, a shaping circuit, and a multichannel pulse amplitude analyzer to convert the analog pulse signals into digitized multichannel energy spectrum data. The host computer terminal 30 connects to the data acquisition and processing unit 20 via a communication interface and is configured to receive the multichannel energy spectrum data and run data processing algorithms to perform nuclide identification, density inversion, and quantitative calculation processes.

[0027] See attached document Figure 2 This invention provides a method for preparing radionuclides from soil. The method is based on the aforementioned apparatus architecture and specifically includes the following steps: Step S1: Acquire multilayer raw energy spectrum data: Place the probe unit 10 into the soil to be tested to a predetermined depth, and control each detection module 110 to operate under uniform high-pressure bias conditions. Each detection module 110 acquires gamma-ray signals from the corresponding soil layer in parallel. After the signals are amplified, shaped, and digitized by the data acquisition and processing unit 20, the raw energy spectrum data of each layer, including environmental background radiation and interlayer crosstalk radiation, is output.

[0028] Step S2: Stripping Interlayer Crosstalk Components: The host computer terminal 30 analyzes the raw energy spectrum data obtained in step S1, and extracts the characteristic secondary radiation signal generated by the gradient cascaded shielding component 120 under oblique radiation excitation in the low-energy region. Based on the intensity of this characteristic secondary radiation signal, the interlayer crosstalk component is quantized, and the interlayer crosstalk component is subtracted from the raw energy spectrum data to obtain the net energy spectrum data corresponding to each depth level.

[0029] Step S3: Inverting Soil In-situ Density: In the net energy spectrum data of this layer obtained in step S2, the ratio of the high-energy characteristic peak count to the Compton continuous spectrum count is extracted using the physical correlation between the Compton effect and the medium density. This ratio is substituted into a preset scattering ratio and density correlation model to calculate the in-situ equivalent soil density corresponding to this detection layer.

[0030] Step S4: Reconstruct the dynamic response matrix: Based on the in-situ equivalent soil density obtained in Step S3, and combined with the attenuation law of gamma rays in the medium, calculate the transmission efficiency correction factor. Use this transmission efficiency correction factor to correct the pre-stored laboratory standard response matrix energy-point by energy to generate a dynamic response matrix adapted to the current field soil environment.

[0031] Step S5: Calculate the types and contents of radionuclides: Establish a linear analytical model between the net energy spectrum data of this layer and the dynamic response matrix. Solve this model using a full-spectrum least squares numerical inversion algorithm to analyze the types of radionuclides in the soil at each depth level, and calculate the quantitative activity data of each radionuclide based on the inversion results.

[0032] See attached document Figure 2 Regarding the process of acquiring multi-layer raw energy spectrum data in step S1, relying on the hardware architecture of the aforementioned probe unit 10 and data acquisition and processing unit 20, this process is specifically implemented through the following sub-steps S110 to S130 to achieve parallel data acquisition and standardized processing.

[0033] In step S110, bias establishment and gain stability calibration of the detector array are performed. The data acquisition and processing unit 20 is internally equipped with a high-voltage power supply module 210, which is independently connected to each detector module 110 within the probe unit 10 via shielded cables. During the measurement startup phase, the high-voltage power supply module 210 applies a preset DC operating bias voltage to the electrodes at both ends of the cadmium zinc telluride crystal of each detector module 110, establishing a charge collection electric field within the crystal. Considering the manufacturing process variations of different detector modules 110 and the influence of ambient temperature on the baseline drift of the electronic system, the host computer terminal 30 utilizes potassium-40 (a naturally occurring radioactive nuclide widely found in the soil environment), a naturally occurring radioactive nuclide in the soil. 40K) is used as a reference source for in-situ energy calibration. The system automatically searches for the center address of the full-energy peak at 1460 keV in the energy spectrum of each detection channel. Based on the linear response assumption, a mapping relationship is established between the channel address values ​​of the multichannel analyzer and the physical energy values. This energy calibration process follows a linear calibration formula: ; in, This represents the physical energy value of the corresponding channel; This represents the discretized channel address value output by the multichannel analyzer; This represents the energy gain coefficient of the system, characterizing the energy width represented by each channel; It represents the zero-point energy intercept of the system and characterizes the bias level of the electronics baseline.

[0034] By solving for the gain coefficient and zero intercept in the above relationship, the system uniformly calibrates the raw data output by each physical level detection module 110 to the standard energy coordinate axis, eliminating the response differences between hardware channels.

[0035] In step S120, the photoelectric conversion and charge collection process of gamma rays is performed. When gamma rays in the soil penetrate the probe shell and enter the sensitive volume of the detection module 110, they interact with the cadmium zinc telluride crystal through the photoelectric effect or Compton scattering, generating electron-hole pairs proportional to the energy of the incident rays. Under the influence of the internal electric field established in step S110, these charge carriers drift towards the electrodes, inducing weak current pulse signals on the electrodes. This process occurs simultaneously in all detection modules 110 distributed along the depth direction, with the detection physical processes at each level being independent and time-synchronized.

[0036] In step S130, front-end conditioning and digital quantization of the analog signal are performed. The data acquisition and processing unit 20 is equipped with an independent analog front-end channel for each detection module 110, including a charge-sensitive preamplifier unit 220 and a shaping amplifier unit 230. The charge-sensitive preamplifier unit 220 is configured to convert the charge pulse output by the detector into a voltage step signal, the output amplitude of which is proportional to the total amount of charge collected. Subsequently, the shaping amplifier unit 230 performs differentiation and integration shaping on the voltage step signal, converting it into a quasi-Gaussian waveform with a specific peak time to optimize the signal-to-noise ratio and meet the timing requirements of subsequent analog-to-digital conversion.

[0037] The shaped pulse signal enters the multichannel pulse amplitude analyzer 240. The multichannel pulse amplitude analyzer 240 is equipped with a high-speed analog-to-digital converter and a peak hold circuit to detect the peak amplitude of the quasi-Gaussian waveform and map this amplitude to the corresponding discrete memory unit for counting and accumulation. After a preset measurement time... Subsequently, the multichannel pulse amplitude analyzers 240 at each level synchronously output histogram arrays composed of the counts of each channel, thus forming the aforementioned raw energy spectrum data vectors for each level. .

[0038] In order to physically mark oblique crosstalk rays from non-detection depth levels, the gradient cascaded shielding component 120 inside the probe unit 10 adopts a three-layer composite structure design based on atomic number gradient, and enhances the signal-to-noise ratio of the tracer signal through a cascaded excitation mechanism.

[0039] The gradient cascaded shielding assembly 120 includes a physical shielding substrate 121, an energy relay layer 122, and a fluorescent tracer layer 123. The physical shielding substrate 121 forms the skeleton of the assembly, located at the outermost or central support position, and is made of a high-density metallic material with an atomic number greater than 70, preferably a tungsten alloy. The physical shielding substrate 121 is configured to block most of the gamma rays that propagate in a straight line along the axial direction using its high linear attenuation coefficient, thus establishing the physical field-of-view boundary between adjacent detection modules 110.

[0040] The energy relay layer 122 is tightly attached to the surface of the physical shielding substrate 121, located between the physical shielding substrate 121 and the fluorescent tracer layer 123. The metal material used for the energy relay layer 122 must have a K-layer characteristic X-ray energy between the high-energy ambient gamma rays and the K-layer absorption edge of the fluorescent tracer layer 123; in this embodiment, lead (Pb) is used. The fluorescent tracer layer 123 is located on the innermost side of the component, directly facing the sensitive surface of the detection module 110. The metal material used for the fluorescent tracer layer 123 must have a K-layer characteristic X-ray energy within the low-energy, high-resolution response range of the detection module 110, and this energy point must avoid the characteristic peaks of common radioactive nuclides in soil; in this embodiment, tin (Sn) is used. Based on the above hardware structure, the gradient cascade excitation process in step S2 is specifically implemented through the following physical action sub-steps S210 to S230: In step S210, high-energy decay and primary conversion of the obliquely incident gamma rays are performed. When interfering gamma rays from adjacent soil layers pass through the edge region of the gradient cascaded shielding assembly 120 at a large angle, the incident high-energy gamma rays (typically with energies greater than 500 keV) first interact with the physical shielding substrate 121 and the energy relay layer 122. Since the cross-section for direct excitation of low atomic number materials by high-energy rays to produce fluorescence is small, the energy relay layer 122, acting as an intermediate medium, absorbs part of the incident ray energy using the photoelectric effect and de-excites to produce lead K-series characteristic X-rays with energies in the range of approximately 75 keV to 85 keV. We define the lead characteristic X-rays generated in this process as the precursor radiation signal.

[0041] In step S220, secondary energy coupling of the precursor radiation is performed. The precursor radiation signal generated in step S210 is emitted isotropically within the component, with the inward component bombarding the fluorescent tracer layer 123. Since the energy of the precursor radiation signal (approximately 75 keV) is slightly higher than and adjacent to the K-shell electron binding energy (approximately 29.2 keV) of the tin material in the fluorescent tracer layer 123, the photoelectric absorption probability reaches its maximum value according to the photoelectric effect cross-section law. This energy matching characteristic allows the precursor radiation signal to excite atoms in the fluorescent tracer layer 123 with extremely high efficiency.

[0042] In step S230, the final emission and labeling of the characteristic tracer signal is performed. The excited atoms of the fluorescent tracer layer 123 undergo electronic transitions, emitting a characteristic secondary radiation signal with an energy strictly corresponding to the Kαα line of tin, with a central energy of 25.2 keV. This characteristic secondary radiation signal directly enters the detector module 110 of this layer, forming an independent counting peak at the low energy end of the energy spectrum.

[0043] The above cascade excitation process follows the following energy transfer inequality: ; in, This indicates the energy of the incident interfering gamma rays; This indicates that the K layer of the energy relay layer 122 material absorbs edge energy; This indicates the characteristic X-ray energy emitted by the stimulated emission of energy relay layer 122; This indicates the K-layer absorption edge energy of the fluorescent tracer layer 123 material; This represents the energy of the characteristic secondary radiation signal emitted by the stimulated emission of the fluorescent tracer layer 123.

[0044] Through this physical mechanism, only obliquely incident interference rays that penetrate the edge of the shielding component will trigger the aforementioned cascade process and generate a 25.2 keV signal, while rays incident perpendicularly to the same layer do not pass through the edge of the shielding component and cannot trigger this signal. Therefore, a clear physical correspondence is established between the peak area at 25.2 keV in the energy spectrum and the flux of interlayer crosstalk rays, realizing the physical labeling of invisible interference components.

[0045] Following the physical process of converting crosstalk in the spatial dimension into characteristic secondary radiation signals in the energy dimension by the aforementioned gradient cascaded shielding components, the host computer terminal 30 performs digital extraction and quantization of characteristic tracer signals for the original energy spectrum data of each layer obtained in step S1 through the following sub-steps S240 to S250.

[0046] In step S240, the region of interest in the energy spectrum is dynamically defined. Since the detector array may be affected by temperature drift during in-situ measurements, the host computer terminal 30 first locates the theoretical center address of the tin (Sn)Kαα characteristic peak at the low-energy end of the original energy spectrum data vector based on the energy calibration coefficients established in step S110. Subsequently, according to the energy resolution characteristics of the cadmium zinc telluride detector, a peak region window covering the full width of the characteristic peak and background estimation windows located on either side of the peak region window are set. The width of the peak region window is set to 1.2 to 1.5 times the full width at half maximum (FWHM) of the characteristic peak to ensure complete inclusion of the characteristic secondary radiation signal count while minimizing the introduction of the Compton continuum background. The background estimation window selects a flat continuum region on either side of the characteristic peak to fit the baseline level in that energy range.

[0047] In step S250, the net count rate of the characteristic peaks is calculated. The original energy spectrum in the low-energy region mainly consists of the Compton continuous spectrum background formed by high-energy ray scattering and the characteristic secondary radiation signal peaks superimposed on it. To obtain physical quantities that represent only the intensity of interlayer crosstalk, the host computer terminal 30 uses a linear background subtraction algorithm to remove the continuous spectrum background beneath the characteristic peaks. This calculation process is based on the tracer signal extraction formula: ; in, The net count rate of the extracted characteristic secondary radiation signal is represented by this parameter, which serves as a scalar indicator for quantifying the intensity of interlayer crosstalk. Indicates the effective live time of energy spectrum acquisition; Indicates the starting address index of the peak area window; Indicates the end address index of the peak area window; This indicates that the raw energy spectrum data is in the 1st... The count value of the path; This represents the set of road addresses covered by the baseline estimation window on the left. This represents the set of road addresses covered by the background estimation window on the right. This indicates the total number of channels included in the baseline estimation window on the left; This indicates the total number of channels included in the background estimation window on the right.

[0048] Through the above calculations, the system eliminated background components unrelated to interlayer crosstalk, obtaining a pure scalar value that is positively correlated with the flux of the oblique-sun rays. Net count rate This value not only represents the intensity of tin fluorescence, but is also defined as a scaling factor for the crosstalk component in subsequent processing, used to drive the reconstruction of full-spectrum crosstalk.

[0049] Based on the net count rate of the characteristic secondary radiation signal calculated in step S250, the host terminal 30 maps the scalar form of crosstalk intensity back to the vector space of the full spectrum through the following sub-steps S260 to S270, and completes the final signal purification.

[0050] In step S260, parameterized reconstruction of the crosstalk energy spectrum is performed. Since the geometry and material properties of the gradient cascaded shielding component 120 are fixed, the shape of the scattered energy spectrum (i.e., the count distribution probability of different energy points) generated when the interlayer oblique rays penetrate the edge of the component exhibits physical stability and does not change significantly with variations in incident intensity. Based on this physical characteristic, the host computer terminal 30 pre-stores a crosstalk energy spectrum shape factor vector obtained through Monte Carlo particle transport simulations (e.g., using the MCNP program) or laboratory standard source oblique projection calibration experiments. The host computer terminal 30 calls this shape factor vector and uses the previously extracted net count rate to perform amplitude modulation on it, reconstructing an interlayer crosstalk component energy spectrum that matches the current measurement state. This reconstruction process is based on the crosstalk spectrum construction formula: ; in, The energy spectrum of the interlayer crosstalk component obtained from the reconstruction is shown in the first... The count value of the channel, which characterizes the background noise contributed only by oblique rays from adjacent levels; This represents the net count rate of the characteristic secondary radiation signals extracted in the preceding steps; This represents the preset crosstalk energy spectrum shape factor vector at the th... The normalized value of the vector describes the probability density of the energy distribution of the oblique rays after they penetrate the shielding components within the full range of the detector. This indicates the effective live time of energy spectrum acquisition, used to convert the count rate back to an absolute count; This represents the channel address index of the multichannel analyzer.

[0051] In step S270, differential acquisition of the net energy spectrum is performed. After obtaining the quantized interlayer crosstalk component energy spectrum, the host computer terminal 30 subtracts it channel by channel from the raw energy spectrum data obtained in step S1. This operation is mathematically equivalent to eliminating radiation components contributed by soil layers outside the current energy spectrum across the entire energy spectrum. This process is based on the spectral stripping formula: ; in, This indicates that the net energy spectrum data obtained after processing is in the [number]th layer. The count value of the channel reflects the true radiation information contributed only by radionuclides in the soil at this depth level; This indicates that the raw energy spectrum data obtained in step S1 is in the [number]th [year]. The count value of the path; The energy spectrum of the interlayer crosstalk component calculated in step S260 is shown in the first... The count value of the path.

[0052] After the above processing, the output net energy spectrum data for this layer is... The signal aliasing interference in the depth direction was eliminated, and the spatial resolution capability of each detection module 110 to the local soil micro-elements was restored, providing an accurate spectral basis for subsequent physical parameter inversion based on single-layer data.

[0053] For negative counts that may be generated due to statistical fluctuations during subtraction, the system uses a non-negativity constraint algorithm to set them to zero or retain them for subsequent least squares fitting residual analysis. The specific numerical processing strategy can be set according to the actual algorithm convergence requirements, which falls within the scope of conventional numerical calculation.

[0054] After obtaining the net energy spectrum data of this layer with crosstalk eliminated using step S2, in order to further eliminate the influence of soil medium density differences on the measurement results, the host computer terminal 30 first needs to accurately define each integration interval used to calculate physical characteristic parameters in the energy spectrum. This process achieves automatic locking of high-energy reference peaks and dynamic delineation of Compton scattering characteristic windows through the following sub-steps S310 to S320.

[0055] In step S310, automatic search and fitting of the high-energy reference peak is performed. This is because the natural radionuclide potassium-40 is commonly found in soil. 40 K), whose emitted 1460 keV gamma rays have high energy, are less affected by low-energy scattering, and have significant characteristics, was therefore selected as the reference peak for density inversion. The host computer terminal 30, using the energy calibration coefficient obtained in step S110, first calculates the theoretical address position corresponding to the 1460 keV energy and sets an estimated search range around this position. Then, within this search range, peak-finding operations are performed on the net energy spectrum data of this layer to identify the local maxima with the highest count rate. To eliminate positioning errors caused by statistical fluctuations, the system uses a Gaussian function to perform nonlinear least-squares fitting on the local maxima and its neighborhood data to determine the precise mathematical center address of this characteristic peak. This fitting process is based on the Gaussian peak shape model formula: ; in, The fitted result represents the first... The count value of the path; The parameter representing the peak amplitude of the characteristic peak; This indicates the center address of the characteristic peak to be determined. This parameter precisely corresponds to the response position of 1460 keV gamma rays in the current detection system. The standard deviation of the characteristic peak is directly related to the energy resolution (FWHM) of the detector. The basis function representing the characteristic peak is usually described by a linear or quadratic polynomial.

[0056] Through the above fitting process, the host computer terminal 30 obtained a center address that could achieve sub-address accuracy. and the standard deviation of the peak width. This allows for precise locking of the reference peak, overcoming the slight spectral shifts caused by temperature drift in the electronic system.

[0057] In step S320, the dynamic correlation delineation of the integration interval is performed. This is based on the center address determined in step S310. with standard deviation The system automatically defines the full-energy peak integration interval and Compton scattering characteristic window for subsequent calculations. For the full-energy peak integration interval, the system uses the center channel address... Based on the standard deviation, expand to the left and right by a preset multiple (usually taken as the standard deviation). This establishes the starting and ending channel addresses of the full-energy peaks, thereby covering more than 99.7% of the full-energy peak counts. For the Compton scattering characteristic window, the system selects a continuous spectral region located to the left of the reference peak and exhibiting a flat characteristic.

[0058] To ensure the consistency of density inversion, the energy range of this feature window needs to remain physically constant (e.g., fixed at 200). (Energy range up to 500keV). The system uses the current energy calibration coefficient to convert this physical energy range into the corresponding address interval in real time. The above address conversion and interval definition process is based on the interval mapping formula: ; in, Indicates the starting address of the full-energy peak integration interval; Indicates the terminal address of the full-energy peak integration interval; Indicates the starting address of the Compton scattering feature window; This indicates the terminal address of the Compton scattering feature window; This indicates the reference peak center address obtained from the fitting in step S310; This represents the energy gain coefficient of the system; This represents the full-energy peak coverage factor, which is usually taken as 3; This indicates the initial physical energy of the preset Compton scattering characteristic window (e.g., 200 keV); This indicates the preset Compton scattering characteristic window termination physical energy (e.g., 500 keV); This represents the current energy gain coefficient; Represents the zero-point energy intercept of the system; This indicates the floor function; This indicates the rounding up operation; This indicates the rounding operation.

[0059] Through this dynamic delineation mechanism, the system can always lock onto the energy spectrum region with the same physical meaning, regardless of whether the detector has gain drift. This ensures that the Compton scattering ratio calculated subsequently only reflects the changes in the physical properties of the soil medium, rather than the fluctuations in the instrument's state.

[0060] After accurately defining the channel address range of the full-energy peak integration interval and the Compton scattering feature window using the aforementioned step S320, the host computer terminal 30 continues to execute the following sub-steps S330 to S340 to complete the extraction of physical feature quantities and their mapping and conversion to medium density parameters.

[0061] In step S330, the energy spectrum characteristic integral and scattering ratio are quantitatively calculated. The host computer terminal 30 calls the net energy spectrum data of this layer obtained in step S2 above, and performs numerical integration on the counts within the full-energy peak integration interval and the Compton scattering characteristic window, respectively. Since the net energy spectrum data of this layer has eliminated crosstalk components from adjacent layers, the count distribution within the window at this time only reflects the scattering and absorption characteristics of gamma rays by the soil medium in this layer. The system derives the dimensionless characteristic value characterizing the physical properties of the medium according to the following Compton scattering ratio calculation formula: ; in, This represents the calculated Compton scattering ratio, which is positively correlated with the electron density of the soil medium. This indicates that the net energy spectrum data of this layer is in the [number]th [epoch]. The count value of the path; This indicates the starting address of the Compton scattering feature window determined in the preceding steps; This indicates the termination address of the Compton scattering feature window determined by the preceding steps; This indicates the starting address of the full-energy peak integration interval determined by the preceding steps; This indicates the termination address of the full-energy peak integration interval determined by the preceding steps; This represents the total base background count within the integration interval of the full-energy peak. This value is estimated by multiplying the average count of the edge channels on both sides of the full-energy peak by the interval width, and is used to ensure that the denominator only represents the direct radiation intensity that has not been scattered.

[0062] In step S340, in-situ density inversion based on a nonlinear response model is performed. Changes in soil medium density significantly alter the transport process of gamma rays within the medium: as density increases, the probability of Compton scattering of high-energy gamma rays increases, leading to a relative decrease in the total energy peak count, while the number of scattered photons in specific energy ranges relatively increases or maintains a specific ratio. This physical mechanism makes Compton scattering more efficient than... There is a monotonic nonlinear mapping relationship between it and the density of the medium.

[0063] To quantify this relationship, the system includes a pre-defined model relating scattering ratio to soil density. This model was developed in a laboratory environment by placing the probe unit into multiple standard soil simulation modules with known densities (density coverage of 1.0 g / cm³). 3 Up to 2.5g / cm 3 The calibration experiment was conducted within the specified range. The host computer terminal 30 then calculates the real-time data obtained in step S330. Substituting the values ​​into the correlation model, the in-situ equivalent density of the soil at the current detection layer is calculated. This inversion process is based on the density response model formula: ; in, This represents the in-situ equivalent density of the soil at the current detection level obtained through inversion, typically expressed in units of... ; This represents the currently calculated Compton scattering ratio; The coefficients of the quadratic terms in the correlation model represent the nonlinear curvature of the density response to the scattering ratio. The coefficient of the first-order term of the correlation model represents the linear sensitivity of the density to the scattering ratio response. The constant term intercept of the correlation model represents the system's response bias under baseline conditions.

[0064] Through the above calculation process, this invention achieves real-time measurement of in-situ soil density by utilizing the scattering characteristics of the energy spectrum itself, without the need for sampling and drying or the use of an additional radiation source. Not only do they serve as output parameters of soil physical properties, but they also act as crucial intermediate variables, being passed to the subsequent response matrix reconstruction step to correct for quantitative errors caused by self-absorption effects. Regarding the fitting coefficients... , , The determination was obtained by performing least squares polynomial regression analysis on laboratory calibration data points, and stored in the database of the host computer terminal in the form of a configuration file.

[0065] After using the aforementioned step S340 to invert the in-situ equivalent density of the soil reflecting the actual physical state of the current detection layer, in order to overcome the quantitative deviation caused by the density mismatch between the laboratory calibration environment and the actual field environment, the host computer terminal 30 quantifies the impact of the medium self-absorption effect on the detection efficiency from the perspective of physical mechanism through the following sub-steps S410 to S420.

[0066] In step S410, an energy correlation retrieval of the attenuation parameters is performed. Since the attenuation characteristics of gamma rays in matter mainly depend on the incident energy and the atomic composition of the medium, the host terminal 30 pre-stores a database of mass attenuation coefficients for standard soil components. This database covers a continuous energy range from low to high energy (e.g., 30 keV to 3000 keV). The system uses the central energy values ​​corresponding to each energy channel in the response matrix to be reconstructed. The corresponding mass attenuation coefficient is retrieved from the database using an interpolation algorithm. This parameter characterizes the macroscopic removal capacity of a unit mass thickness of soil medium for gamma rays of a specific energy, and its value is independent of the physical density of the medium.

[0067] In step S420, the analytical calculation of the transmission efficiency correction factor is performed. Based on the exponential decay law of rays described by Beer-Lambert's law, changes in soil medium density directly cause nonlinear changes in the effective ray flux within the detector's sensitive volume. The host terminal 30 then converts the aforementioned inverted in-situ equivalent density... Standard density compared to laboratory calibration By comparing the results and considering the geometric response characteristics of probe unit 10, calculations were performed for each energy point. The transmission efficiency correction factor. This calculation is based on the efficiency correction formula: ; in, Indicates that for energy The transmission efficiency correction factor for gamma rays, a dimensionless coefficient that characterizes the detection efficiency gain or loss ratio of the field environment relative to the standard environment. Indicates energy as The mass attenuation coefficient of gamma rays in soil medium; This represents the in-situ equivalent density of the soil at the current detection level obtained in step S3. This represents the physical density of the calibration phantom used to generate the initial standard response matrix, i.e., the standard density during laboratory calibration. This value is a known system constant. The effective detection radius of the probe unit is represented by this geometric parameter, which is determined by the crystal size of the detection module 110 and the geometric angle between the soil and the detector, and is determined by Monte Carlo simulation integration.

[0068] Through the above calculations, the system generates corresponding correction weights for each energy slice across the full energy spectrum. When hour, This indicates that the high density of the soil at the site leads to enhanced self-absorption, necessitating a reduction in the expected detection efficiency; conversely, a lower density soil... .

[0069] This process transforms macroscopic density differences into microscopic energy-dependent corrections, achieving precise decoupling of geometric and physical coupling effects and ensuring that the subsequently reconstructed response matrix can accurately reflect the radiative transfer characteristics of the soil in the field.

[0070] Following the aforementioned calculation process of the transmission efficiency correction factor, the host terminal 30 applies the calculated physical correction coefficient to the data structure of the standard response matrix by executing sub-steps S430 to S440, thereby completing the dynamic reconstruction of the system response characteristics.

[0071] In step S430, the standard response matrix is ​​retrieved and the column vectors are parsed. The host computer terminal 30 reads the laboratory standard response matrix from non-volatile memory. This matrix is ​​a matrix with dimension 1. A two-dimensional numerical array, in which The total number of channel addresses corresponding to the multi-channel energy spectrum. This corresponds to the number of radionuclide species to be analyzed or the discrete incident gamma ray energy levels. Each column vector in the matrix... Characterized the first Nuclide or the first Each characteristic energy point at standard density Under these conditions, the normalized probability density distribution generated across the full energy spectrum of the detector. Based on a pre-set nuclide library index file, the system identifies the characteristic main emission energy corresponding to each column of the matrix. (For example, for the cesium-137 nuclide series, corresponding to) keV).

[0072] In step S440, an element-wise weighted transformation of the response matrix is ​​performed. The host terminal 30 uses the transmission efficiency correction factor function calculated in step S420. A specific gain correction is applied to each column vector of the standard response matrix. Specifically, the system extracts the first... eigenenergy corresponding to the column Substituting the values ​​into the correction factor function yields the corresponding scalar correction value, which is then multiplied by all elements of the column vector. This process essentially transforms the difference in detection efficiency caused by changes in macroscopic medium density into a linear scaling of the response matrix on the amplitude axis, thereby numerically simulating the in-situ equivalent density of the soil at the current detection layer. The physical effects of recalibration are then described. This matrix reconstruction process is based on the matrix transformation formula: ; in, The generated dynamic response matrix is ​​represented in the first... line, number The column contains numerical elements, representing the value of the [number]th [unit] under the current site density conditions. Nuclide or energy in the first The expected count response generated by the channel; This represents the pre-stored laboratory standard response matrix at the th... line, number The numerical elements of the column; Indicates the corresponding to the first Column characteristic energy The transmission efficiency correction factor is calculated by the aforementioned step S420; The channel address index represents the multichannel energy spectrum, with values ​​ranging from 1 to... ; Indicates the nuclide type or energy index, with values ​​ranging from 1 to... .

[0073] By examining all of the matrix After the column vectors have completed the above traversal calculations, the system will recombine the updated column vectors to generate a complete dynamic response matrix. This matrix not only incorporates the detector's inherent energy resolution and Gaussian broadening characteristics, but also integrates the differentiated attenuation characteristics of the soil medium at different energy levels. This establishes a precise linear mapping relationship between the detected net energy spectrum of this layer and the activity of unknown nuclides, providing a mathematical operator that conforms to the objective physical scenario for subsequent quantitative analysis.

[0074] The dynamic response matrix is ​​completed using the aforementioned step S4. The reconstruction, combined with the net energy spectrum data of this layer obtained in step S2. Then, the host computer terminal 30 executes sub-steps S510 to S530 in step S5 to transform the physical measurement problem into a mathematical inverse problem in order to analyze the specific types and activity values ​​of each radionuclide in the soil.

[0075] In step S510, a linear analytical model is constructed. Based on the principle of linear superposition in gamma-ray spectroscopy, i.e., the total energy spectrum recorded by the detector is equal to the linear weighted sum of the energy spectra contributed by each individual nuclide, the host computer terminal 30 establishes a system of linear equations describing the relationship between the net energy spectrum vector, the dynamic response matrix, and the activity vector of the nuclide to be determined. This model abstracts the physical process into matrix operations, and the system constructs the following linear response model formula: ; in, This represents the column vector of the net energy spectrum of this layer obtained in step S2, with dimension . ,Include The count measurement value of each road address; This represents the dynamic response matrix generated in step S4, with dimensions of... ,Include Nuclide to be identified Unit activity response probability at each address; This represents the column vector of nuclide activities to be determined, with dimension . Each element in the vector corresponds to the activity concentration (e.g., Bq / kg) of a specific nuclide; This represents a column vector of measurement errors, with dimensions of... It includes random biases caused by statistical fluctuations and electronic noise.

[0076] In this model, The column space forms the signal space of the detector. For the observation projection in this space, and That is, the coordinate components of the projection on each basis vector.

[0077] In step S520, the statistical weighting matrix is ​​configured. Since nuclear radiation measurements follow a Poisson distribution, the count values ​​at different energy addresses have different statistical uncertainties. The relative error is larger for lower count channels and smaller for higher count channels. To improve the accuracy of the inversion results, the host computer terminal 30 constructs a diagonal weighting matrix. This is used to assign greater weight to data points with high confidence levels in subsequent least squares solutions. The element arrangement of this weight matrix follows the statistical weighting formula: ; in, Represents the weight matrix The Middle The diagonal elements of the column; Indicates the first The variance of the count values; This represents the first element in the net energy spectrum vector of this layer. The count value of the path.

[0078] By introducing a weight matrix The system transforms the ordinary least squares problem into a maximum likelihood estimation problem, making the solution statistically optimal.

[0079] In step S530, a weighted least squares (WLS) inversion solution is performed. This is for the overdetermined linear equation system (i.e., the number of equations) constructed in step S510. Far greater than the unknown The host computer terminal 30 uses a weighted least squares algorithm to solve the objective function. The system directly analyzes the optimal estimate of the nuclide activity vector through matrix operations. This solution process is based on the weighted least squares inversion formula: ; in, This represents the calculated nuclide activity estimation vector; The transpose of the dynamic response matrix; This represents the inverse operation of a matrix.

[0080] The vector obtained by solving The non-zero element index indicates the types of radionuclides present in the soil, and the corresponding element value is the quantitative activity of that nuclide. For negative solutions that may occur during the calculation (which are physically meaningless), the system iteratively corrects them based on the above analytical solutions by combining a non-negative least squares constraint algorithm, forcing the negative components to converge to zero, thereby obtaining the final activity data that conforms to physical reality.

Claims

1. A method for preparing radioactive nuclides from soil, characterized in that, Includes the following steps: Each detection module (110) is controlled to collect raw energy spectrum data and acquire characteristic secondary radiation signals generated by the gradient cascaded shielding assembly (120); The crosstalk component is removed from the original energy spectrum data using the characteristic secondary radiation signal to obtain the net energy spectrum data of this layer; Calculate the Compton scattering ratio of the net energy spectrum data of this layer, and inversely determine the in-situ equivalent density of the soil at the current detection layer based on this ratio; The transmission efficiency correction factor is calculated based on the in-situ equivalent density, and the pre-stored laboratory standard response matrix is ​​adjusted using the transmission efficiency correction factor to reconstruct the dynamic response matrix. Based on the net energy spectrum data of this layer and the dynamic response matrix, a linear analytical model is constructed to calculate the types and activities of radionuclides in the soil.

2. The method for preparing radioactive nuclides in soil according to claim 1, characterized in that, The gradient cascaded shielding assembly (120) is constructed from layered composite materials to block straight rays propagating along the axial direction of the probe unit (10) and to generate the characteristic secondary radiation signal of specific energy for large-angle oblique rays penetrating the edge of the assembly. The process of removing crosstalk components from the original energy spectrum data includes: extracting the region of interest (ROI) corresponding to the characteristic secondary radiation signal from the original energy spectrum data; calculating the net count rate of the ROI using the tracer signal extraction formula; multiplying the net count rate by a preset crosstalk energy spectrum shape factor to reconstruct a crosstalk energy spectrum containing only crosstalk information; and finally subtracting the crosstalk energy spectrum from the original energy spectrum data to obtain the net energy spectrum data of this layer.

3. The method for preparing radioactive nuclides in soil according to claim 1, characterized in that, The Compton scattering ratio is obtained through the Compton scattering ratio calculation formula, specifically including: Search for local maxima of high-energy reference peaks in the net energy spectrum data of this layer, and fit the local maxima and their neighborhood data using a Gaussian function to determine the center address of the reference peak and the standard deviation of the characteristic peak. Using the reference peak center address as a benchmark, the characteristic peak standard deviation is extended to the left and right by a preset multiple to establish the starting and ending addresses of the full-energy peak integration interval. A continuous spectral region located to the left of the high-energy reference peak and having a flat characteristic is selected as the Compton scattering feature window. Using the current energy gain coefficient and the zero-point energy intercept of the system, the preset physical energy range is converted into the corresponding channel address interval, and the starting and ending channel addresses of the Compton scattering feature window are established.

4. The method for preparing radioactive nuclides in soil according to claim 3, characterized in that, The process of calculating the Compton scattering ratio of the net energy spectrum data of this layer includes: Numerical integration is performed on the counts within the Compton scattering characteristic window and the counts within the full-energy peak integration interval, respectively. Calculate the total base count within the full-energy peak integration interval, and subtract the total base count from the count within the full-energy peak integration interval to obtain the net full-energy peak count; Calculate the ratio of the total count within the Compton scattering characteristic window to the net count of the full-energy peak, and use this ratio as the Compton scattering ratio.

5. The method for preparing radioactive nuclides in soil according to claim 4, characterized in that, The process of retrieving the in-situ equivalent density of the soil at the current detection level includes: A correlation model between scattering ratio and soil density is pre-established, and the correlation model is obtained through calibration experiments in multiple sets of standard soil simulation modules with known densities. Substituting the Compton scattering ratio into the correlation model, the in-situ equivalent density of the soil at the current detection layer is calculated using the density response model formula.

6. The method for preparing radioactive nuclides in soil according to claim 1, characterized in that, The process of calculating the transmission efficiency correction factor includes: Based on the center energy value corresponding to each channel in the response matrix to be reconstructed, the corresponding mass attenuation coefficient is retrieved from the preset mass attenuation coefficient database. Calculate the difference between the in-situ equivalent density and the standard density used when generating the laboratory standard response matrix; The transmission efficiency correction factor for each energy point is calculated based on the efficiency correction formula. The value of the factor is equal to the value of an exponential function with the natural constant as the base and the product of the negative mass attenuation coefficient, the effective detection radius of the probe unit (10), and the difference as the exponent.

7. The method for preparing radioactive nuclides in soil according to claim 1, characterized in that, The process of reconstructing the dynamic response matrix is ​​performed according to the matrix transformation formula, including: Retrieve and analyze the characteristic primary emission energy corresponding to each column vector in the laboratory standard response matrix; Substitute the characteristic main emission energy into the calculation result of the transmission efficiency correction factor to obtain the corresponding scalar correction value; The operation defined by the matrix transformation formula is performed to multiply all elements of each column vector in the laboratory standard response matrix by the scalar correction value corresponding to each column vector to generate the dynamic response matrix.

8. The method for preparing radioactive nuclides in soil according to claim 1, characterized in that, The process of constructing the linear analytical model includes: A linear response model formula is established, wherein the linear response model formula includes a system of linear equations for the net energy spectrum vector of this layer, the dynamic response matrix, the nuclide activity vector to be determined, and the measurement error vector; The net energy spectrum vector of this layer is equal to the matrix product of the dynamic response matrix and the nuclide activity vector plus the measurement error vector.

9. A method for preparing radioactive nuclides in soil according to claim 8, characterized in that, The process of calculating the types and activities of radionuclides in the soil is performed based on the weighted least squares inversion formula, including: Construct a diagonal weight matrix, where the diagonal elements are configured as the reciprocal of the variance of the corresponding energy address count; Perform matrix operations on the weighted least squares inversion formula, namely, calculate the transpose of the dynamic response matrix, the inverse of the product of the diagonal weight matrix and the dynamic response matrix, and then multiply by the transpose of the dynamic response matrix, the diagonal weight matrix and the net energy spectrum data of this layer in sequence to obtain the nuclide activity estimation vector. The nuclide activity estimation vector is iteratively corrected using a non-negative least squares constraint algorithm to obtain a non-negative solution vector.

10. An apparatus for preparing radioactive nuclides in soil, applied to the method for preparing radioactive nuclides in soil according to any one of claims 1-9, characterized in that, include: The probe unit (10) is constructed as a rod-shaped structure suitable for vertical insertion into the soil medium to be tested. Inside the probe unit (10), there are detector arrays spaced apart along the axial length direction. The detector array consists of multiple independent detector modules (110), and a gradient cascaded shielding component (120) is provided between adjacent detector modules (110). The gradient cascaded shielding component (120) generates characteristic secondary radiation signals. The data acquisition and processing unit (20) establishes an electrical connection with each of the detection modules (110) and converts the analog pulse signals from the detection modules (110) into digitized raw energy spectrum data; The host computer terminal (30) is connected to the data acquisition and processing unit (20).

Images