A plastic scintillator gamma spectrum system calibration method
By constructing a global optimization model and comparing algorithms, the problem that plastic scintillators cannot detect full-energy peaks was solved, high-precision system calibration was achieved, and its application in precision energy spectrum analysis was expanded.
Patent Information
- Application Number
- CN202610193545.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-02-10
- Publication Date
- 2026-06-12
AI Technical Summary
In existing technologies, plastic scintillators cannot detect the full-energy peak, making traditional calibration methods unsuitable. Furthermore, calibration methods based on Compton features fail to fully utilize the full-spectrum information, resulting in insufficient calibration accuracy and reliability.
The system calibration problem is transformed into a multi-parameter global optimization problem. A simulated energy spectrum is generated by constructing an accurate detector response model and comparing it with the experimental energy spectrum over the entire spectrum. The global optimization algorithm is used to find the optimal calibration parameters, bypassing the dependence on a single full-energy peak or Compton edge.
This invention enables calibration that does not rely on a single characteristic peak, improves the accuracy of calibration parameters, reduces error propagation, and broadens the application potential of plastic scintillators in the field of precision energy dispersive spectroscopy.
Smart Images
Figure CN122194230A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of gamma-ray spectrum measurement and analysis technology, and specifically relates to a calibration method for a plastic scintillator gamma-ray spectrum system. Background Technology
[0002] For nuclear radiation detector systems, accurate system calibration is fundamental to obtaining reliable data. Whether it's energy calibration or peak calibration, the core idea is to use existing information to calibrate unknown situations, thereby ensuring the accuracy and traceability of measurement results. Any rigorous nuclear radiation detection work must be based on accurate calibration.
[0003] Plastic scintillators are a type of organic scintillator widely used in nuclear radiation detection due to their advantages such as high detection efficiency, fast response speed, low cost, ease of processing, radiation resistance, and stable performance. However, because the luminescence mechanism of plastic scintillators involves multiple energy transfer processes, and Compton scattering is dominant, it is almost impossible to detect the full-energy peak, making traditional system calibration methods based on the full-energy peak difficult to apply.
[0004] Therefore, researchers began exploring systematic calibration methods based on Compton scattering. Early methods typically used the Compton peak formed by Compton edge broadening to systematically calibrate plastic scintillators. However, due to the poor energy resolution of plastic scintillators, the Compton peak formed by Compton edge broadening does not correspond to the energy of the Compton edge itself. To determine the energy corresponding to the Compton peak, the right half-height value of the Compton peak can be simply used as the Compton edge energy, but this method has significant errors. To further reduce calibration errors, some researchers have iteratively optimized the energy broadening parameters and compared the Compton peak in the simulated spectrum with the measured spectrum to determine the actual position of the Compton edge; others have established deep autoencoder models to reconstruct the Compton edge in the γ-ray spectrum of plastic scintillators.
[0005] The above method shifts the research object from photoelectric effect to Compton scattering. In essence, it is the same as the traditional system calibration method based on full-energy peak. Both methods seek the mapping relationship between energy and channel address, i.e., energy calibration, by obtaining the actual channel addresses of multiple characteristic energy points. However, they do not make full use of the full spectrum information, and the resulting system calibration is still not accurate enough.
[0006] To broaden the application scenarios of plastic scintillators and overcome the limitations of plastic scintillators being almost undetectable at their full-energy peaks and difficult to calibrate systematically, it is necessary to study a calibration method to achieve systematic calibration of plastic scintillators. Summary of the Invention
[0007] (a) Technical problems to be solved The present invention aims to solve the technical problem that in the prior art, the traditional calibration method based on characteristic energy points is not applicable because plastic scintillators cannot detect the full-energy peak, while the existing calibration method based on Compton features fails to make full use of the full spectrum information, resulting in insufficient calibration accuracy and reliability.
[0008] (II) Technical Solution To address the aforementioned problems, this invention proposes a calibration method for a plastic scintillator gamma-ray spectrum system. The core idea of this invention is to transform the system calibration problem into a multi-parameter global optimization problem: by constructing an accurate detector response model to generate a simulated energy spectrum, and comparing it with the actual measured experimental energy spectrum across the entire spectrum, an optimization algorithm directly finds a set of optimal calibration parameters that minimize the overall difference between the simulated and experimental energy spectra, thereby bypassing the dependence on precise localization of a single full-energy peak or Compton edge. The specific steps are as follows: S1. Construct a plastic scintillator detector system to obtain energy spectrum data from sources of different scales; S2. Preprocess the obtained energy spectrum data to obtain the energy spectrum data required by the optimization algorithm. Assume the total number of channel addresses is... ; S3. Select commonly used energy scale and peak shape scale formulas, as follows, to prepare for the subsequent global search optimization to obtain the simulated energy spectrum and experimental energy spectrum.
[0009]
[0010]
[0011] In the formula m, n, q, a, b, c The desired scale parameter is denoted as .
[0012] S4. Build a simulation model based on the experimental conditions, constructing a model with a size of [size missing]. Unexpanded response matrix D Set the energy range to (keV), energy step size is (keV):
[0013] Based on the peak shape scaling formula, construct a scale of... Gaussian broadening matrix ,in for The Line number The value of the column, for D The Column corresponding If the value is , then:
[0014] and These represent the center and standard deviation of the Gaussian function, respectively. The normalization coefficient is calculated as follows:
[0015] matrix sum matrix Multiplying them together yields a value of 1. System response matrix .
[0016]
[0017] Theoretically, any energy range can be obtained based on the characteristic energies of different radionuclides and the weights of their emission probabilities. The energy step size is Simulated energy spectrum:
[0018] in, This represents the simulated energy spectrum. Representation matrix The List, This represents the weight assigned to that column.
[0019] S5. According to the energy calibration formula, convert the energy spectrum data (Counts-Channel) into a calibrated experimental energy spectrum (Counts-Energy). The energy range of the experimental energy spectrum is then determined. (keV).
[0020]
[0021] Note that the experimental energy spectrum differs in energy range from the simulated energy spectrum obtained above. Therefore, when At that time, Perform interpolation; when At that time, By merging the two spectra to make their energy ranges the same, the experimental energy spectrum is finally obtained.
[0022] S6. Based on the experimental and simulated energy spectra obtained above, model the system calibration problem and transform it into a multi-parameter optimization problem. The optimization parameters are the parameters in the energy calibration formula. 、 、 In the peak shape scale formula 、 、 The optimization objectives are as follows:
[0023] In the formula, The first part represents the simulated energy spectrum. Dao data; The first part of the experimental energy spectrum represents the... Dao data, for , used in total The energy spectrum of each experiment was used for optimization.
[0024] S7. Use global search optimization algorithms such as genetics or particle swarm optimization to find the optimal solution that satisfies the optimization objective and obtain the energy scale and peak shape scale parameters of the plastic scintillator.
[0025] S8. The above method addresses the problem of the inability to detect the full-energy peak in plastic scintillators, making it difficult to perform system calibration. It directly obtains calibration parameters through full-spectrum optimization, providing a technical approach for the calibration of plastic scintillator systems.
[0026] Furthermore, in step S2, the preprocessing method is as follows: S2-1. Background subtraction is performed on the obtained raw energy spectrum data; S2-2. The smoothed energy spectrum data is calculated by using the window averaging method on the obtained raw energy spectrum data.
[0027] Furthermore, in step S7, the global search strategy is as follows: S7-1. To shorten the optimization time, a wider search range can be estimated based on the energy spectrum data of several calibration sources; S7-2. Before optimization, the simulated energy spectrum and the experimental energy spectrum need to be normalized. S7-3. If parameters encounter boundary conditions during optimization, the range can be appropriately widened.
[0028] (III) Beneficial Effects Compared with the prior art, the present invention has the following beneficial effects: 1. This invention constructs a complete system response model that includes both the unbroadened response and the energy broadening effect, and utilizes a global optimization algorithm to match the simulated and experimental energy spectra across the entire spectrum, achieving calibration independent of any single characteristic peak (such as the full-energy peak or the Compton edge). This method fully leverages the shape information of the entire spectrum, improving the accuracy of the calibration parameters.
[0029] 2. This invention transforms the complex calibration process into a clear mathematical optimization problem, which is solved using a mature global optimization algorithm. The process is clear, highly automated, and reduces reliance on operator experience.
[0030] 3. This invention performs cross-optimization through full-spectrum comparison of simulation and experiment, so that the energy scale parameters and peak shape scale parameters are mutually constrained and determined collaboratively during the optimization process, thus overcoming the error propagation problem that may be introduced by traditional step-by-step scaling.
[0031] 4. The method described in this invention has strong universality and is particularly suitable for detectors such as plastic scintillators, which are difficult to observe full-energy peaks. It provides a reliable and systematic calibration solution for them, and broadens their application potential in the field of precision energy spectrum analysis. Attached Figure Description
[0032] Figure 1 This is a flowchart of the system calibration method based on a plastic scintillator according to the present invention; Figure 2 A schematic diagram of the simulation model established for the experimental conditions; Among them: 1. Plastic scintillator, 2. MgO, 3. Al, 4. Photomultiplier tube, 5. Ceramic tabletop, 6. Radiation source, 7. Foam block; Figure 3a Plotting the 3D model of the unstretched response matrix. Figure 3b Plot a 3D graph of the Gaussian broadening matrix. Figure 3c Plot the system response matrix in 3D; Figure 4 This is a graph showing how the optimal and average loss function values change with the number of iterations during optimization. Figure 5 a is a graph showing the energy calibration results of the plastic scintillator detector. Figure 5 b is a graph showing the peak shape calibration results of the plastic scintillator detector; Figure 6 a is 137 The coincidence diagram between the experimental and simulated energy spectra of Cs nuclides. Figure 6 b is 60 The coincidence diagram between the experimental and simulated energy spectra of Co nuclide. Figure 6 c is 22 The coincidence diagram between the experimental and simulated energy spectra of sodium nuclide. Figure 6 d is 54 The coincidence diagram of the experimental and simulated energy spectra of Mn nuclides. Detailed Implementation
[0033] To make the objectives, contents, and advantages of the present invention clearer, the specific embodiments of the present invention will be described in further detail below with reference to the accompanying drawings and examples.
[0034] This embodiment proposes a calibration method for a plastic scintillator gamma-ray spectroscopy system, the process of which is as follows: Figure 1 As shown, the specific steps include the following: S1. Construct a plastic scintillator detector system to obtain energy spectrum data from sources of different scales; S2. Preprocess the obtained energy spectrum data to obtain the energy spectrum data required by the optimization algorithm. Assume the total number of channel addresses is... ; The preprocessing method is as follows: S2-1. Background subtraction is performed on the obtained raw energy spectrum data; S2-2. The smoothed energy spectrum data is calculated by using the window averaging method on the obtained raw energy spectrum data.
[0035] S3. Select commonly used energy scale and peak shape scale formulas, as follows, to prepare for the subsequent global search optimization to obtain the simulated energy spectrum and experimental energy spectrum.
[0036]
[0037]
[0038] S4. Build a simulation model based on the experimental conditions, constructing a model with a size of [size missing]. Unexpanded response matrix D Set the energy range to (keV), energy step size is (keV):
[0039] Based on the peak shape scaling formula, construct a scale of... The Gaussian broadening matrix, where for The Line number The value of the column, for D The Column corresponding If the value is , then:
[0040] and These represent the center and standard deviation of the Gaussian function, respectively. The normalization coefficient is calculated as follows:
[0041] matrix sum matrix Multiplying them together yields a value of 1. System response matrix .
[0042]
[0043] Theoretically, any energy range can be obtained based on the characteristic energies of different radionuclides and the weights of their emission probabilities. The energy step size is Simulated energy spectrum:
[0044] in, This represents the simulated energy spectrum. Representation matrix The List, This represents the weight assigned to that column.
[0045] S5. According to the energy calibration formula, convert the energy spectrum data (Counts-Channel) into a calibrated experimental energy spectrum (Counts-Energy). The energy range of the experimental energy spectrum is then determined. (keV).
[0046]
[0047] Note that the experimental energy spectrum differs in energy range from the simulated energy spectrum obtained above. Therefore, when At that time, Perform interpolation; when At that time, The data was then merged. The final experimental energy spectrum was obtained.
[0048] S6. Based on the experimental and simulated energy spectra obtained above, model the system calibration problem and transform it into a multi-parameter optimization problem. The optimization parameters are the parameters in the energy calibration formula. 、 、 In the peak shape scale formula 、 、 The optimization objectives are as follows:
[0049] In the formula, The first part represents the simulated energy spectrum. Dao data; The first part of the experimental energy spectrum represents the... Dao data, for , used in total The energy spectrum of each experiment was used for optimization.
[0050] S7. Use global search optimization algorithms such as genetics or particle swarm optimization to find the optimal solution that satisfies the optimization objective and obtain the energy scale and peak shape scale parameters of the plastic scintillator.
[0051] The global search strategy is as follows: S7-1. To shorten the optimization time, a wider search range can be estimated based on the energy spectrum data of several calibration sources; S7-2. Before optimization, the simulated energy spectrum and the experimental energy spectrum need to be normalized. S7-3. If parameters encounter boundary conditions during optimization, the range can be appropriately widened.
[0052] S8. The above method addresses the problem of the inability to detect the full-energy peak in plastic scintillators, making it difficult to perform system calibration. It directly obtains calibration parameters through full-spectrum optimization, providing a technical approach for the calibration of plastic scintillator systems.
[0053] The system calibration effect was verified by comparing experimental and simulated energy spectra. The experimental equipment and parameters were set as follows: 22 Na、 54 Mn, 60 Co、 137 Using four radionuclides of Cs as calibration sources, the characteristic gamma-ray energies and emission probabilities of the above calibration sources are shown in Table 1.
[0054] Table 1 Characteristic energies and emission probabilities of different nuclides
[0055] This experiment uses an ORTEC 572A amplifier to provide DC power and an ORTEC 556 high-voltage power supply to provide high voltage. A Φ7.5... A 7.5 cm BC408 scintillator detector, connected to a digitizer and a high-performance computer, was used to acquire and calculate the aforementioned nuclide energy spectrum data. The digitizer was a DT-5730 model with a sampling rate of 500 MS / s, a sampling depth of 14 bits, and a single pulse acquisition length of 100 μs. The acquired nuclear pulse signals were transmitted to the high-performance computer for calculation using CoMPASS software, and data processing software was Matlab.
[0056] Furthermore, based on the above experimental conditions, a simulation environment was built, and the unstretched response matrix was constructed; the Gaussian stretched matrix was initialized; and the two were multiplied to obtain the system response matrix. Figure 2 This is a schematic diagram of the simulation model. Figure 3a ~ c represent the 3D plots of the three types of matrices mentioned above.
[0057] After completing the above data preparation, an evolutionary strategy was used for a global search. After multiple parameter adjustments and comparisons, the population size was set to 3000, the mutation probability to 0.25, the crossover probability to 0.9, the maximum number of iterations to 30, and the maximum number of stalled iterations to 5. Figure 4 The changes in the optimal and average loss function values with the number of iterations when using the above data and parameters for optimization are described. It can be seen that the optimal loss function value decreases rapidly in the first 5 iterations, then tends to level off, and converges in the 27th iteration. The average loss function value fluctuates greatly in the first 10 iterations, then tends to level off, and reaches its minimum in the 28th iteration. This indicates that the algorithm converges effectively and has a strong global optimization ability.
[0058] Figure 5 a ~ b are the results of energy and peak shape calibration of the plastic scintillator detector obtained by direct calculation using the optimization method, respectively.
[0059] Figure 6 a ~ d represent the values when using the optimization results as the system scale, respectively. 22 Na、 54 Mn, 60 Co and 137 The experimental and simulated energy spectra of the four Cs nuclides showed good agreement with coefficients of determination of 0.9801, 0.9729, 0.9512 and 0.9944, respectively, all above 0.95, indicating accurate calibration.
[0060] The above description is only a preferred 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 technical principles 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 system calibration method based on a plastic scintillator, characterized in that, The system calibration method includes the following steps: S1. Build a plastic scintillator detector system and obtain energy spectrum data of different calibration sources; S2. Preprocess the obtained energy spectrum data to obtain the energy spectrum data required by the optimization algorithm. Assume the total number of channel addresses is... ; S3. Select common energy calibration and peak shape calibration formulas as follows, to prepare for subsequent obtaining of simulated energy spectra and experimental energy spectra and global search optimization; S4. Build a simulation model based on the experimental conditions, constructing a model with a size of [size missing]. Unexpanded response matrix D Set the energy range to (keV), energy step size is (keV): Based on the peak shape scaling formula, construct a scale of... The Gaussian broadening matrix, where for The Line number The value of the column, for The Column corresponding If the value is , then: and These represent the center and standard deviation of the Gaussian function, respectively. The normalization coefficient is calculated as follows: matrix sum matrix Multiplying them together yields a value of 1. System response matrix , Theoretically, any energy range can be obtained based on the characteristic energies of different radionuclides and the weights of their emission probabilities. The energy step size is Simulated energy spectrum: in, This represents the simulated energy spectrum. Representation matrix The List, This represents the weight assigned to that column; S5. According to the energy calibration formula, convert the energy spectrum data (Counts-Channel) into a calibrated experimental energy spectrum (Counts-Energy). The energy range of the experimental energy spectrum is then determined. (keV), Note that the experimental energy spectrum differs in energy range from the simulated energy spectrum obtained above. Therefore, when At that time, Perform interpolation; when At that time, By merging the two spectra to make their energy ranges the same, the experimental energy spectrum is finally obtained. S6. Based on the experimental and simulated energy spectra obtained above, model the system calibration problem and transform it into a multi-parameter optimization problem. The optimization parameters are the parameters in the energy calibration formula. 、 、 In the peak shape scale formula 、 、 The optimization objectives are as follows: In the formula, The first part represents the simulated energy spectrum. Dao data; The first part of the experimental energy spectrum represents the... Dao data, for , used in total Each experimental energy spectrum was used for optimization; S7. Use a global search optimization algorithm such as genetic algorithm or particle swarm algorithm to find the optimal solution that meets the optimization goal, and obtain the energy calibration and peak shape calibration parameters of the plastic scintillator; S8. Since the above method is faced with the problem that the plastic scintillator cannot detect the full energy peak and it is difficult to perform systematic calibration, the calibration parameters are directly obtained through full-spectrum optimization, providing a technical approach for the system calibration of the plastic scintillator.
2. The system calibration method based on a plastic scintillator according to claim 1, characterized in that, The method is applicable to plastic scintillators with low photoelectric effect reaction cross-sections that cannot detect the full energy peak.
3. The system calibration method based on a plastic scintillator according to claim 1, characterized in that, In step S2, the preprocessing operation includes: S21. Perform background subtraction on the obtained original energy spectrum data; S22. Use the window averaging method to calculate the smoothed energy spectrum data from the obtained original energy spectrum data.
4. The system calibration method based on a plastic scintillator according to claim 1, characterized in that, In step S4, when constructing the unbroadened response matrix D, the energy step d is取值 in the range of [0.1, 10] keV.
5. The system calibration method based on a plastic scintillator according to claim 1, characterized in that, In step S7, the global search optimization algorithm is a genetic algorithm or a particle swarm algorithm.
6. The system calibration method based on a plastic scintillator according to claim 1, characterized in that, In step S7, to shorten the optimization time, a relatively wide search range is预估 based on the energy spectrum data of several calibration sources.
7. The system calibration method based on a plastic scintillator according to claim 1, characterized in that, Before the optimization in step S7, normalize the simulated energy spectrum and the experimental energy spectrum.
8. The system calibration method based on a plastic scintillator according to claim 1 or 6, characterized in that, During the optimization process of step S7, if a parameter encounters a boundary situation, broaden the parameter search range.
9. The system calibration method based on a plastic scintillator according to claim 1, characterized in that, In step S4, build a simulation model according to the experimental conditions to construct the unbroadened response matrix D.
10. The system calibration method based on a plastic scintillator according to claim 1, characterized in that, In step S5, when M < N, perform interpolation processing on the experimental energy spectrum, and when M > N, perform merging processing on the experimental energy spectrum.