N / gamma discrimination method based on pulse front and back edge combined gradient

Abstract

The invention provides an n / gamma discrimination method based on a pulse front and back edge combined gradient. The n / gamma discrimination method comprises the following steps: carrying out preprocessing including smooth filtering, baseline deduction, accumulation discarding and maximum value amplitude normalization on a mixed radiation field pulse signal; searching a peak point of the preprocessed n / gamma pulse signal, respectively selecting sample points at a time interval before and after the peak point, and respectively making gradients between the two sample points and the peak point; multiplying the obtained front and back edge gradients of the pulse to obtain a combined gradient value; and discriminating the n / gamma pulse signal by taking the combined gradient value as a pulse waveform discriminating parameter. Compared with the existing pulse gradient method, the method has the advantage that the screening effect is obviously improved on the premise of not increasing excessive calculation amount.

Description

Technical Field

[0001] The present invention belongs to the technical field of digital pulse processing in nuclear electronics, and particularly relates to an n / γ discrimination method based on the combined gradient of the leading and trailing edges of a pulse. Background Art

[0002] Neutron detection is very useful in fields such as security inspection, nuclear security, and nuclear physics experiments. Neutrons usually exist accompanied by γ-rays. In order to detect neutrons, neutron-gamma (n / γ) discrimination is very necessary. The pulse signals measured by a detector have differences in waveform, and the pulse shape discrimination (PSD) technology can be used to distinguish the pulse signals in a mixed radiation field.

[0003] The most commonly used discrimination methods include the charge comparison method, the pulse gradient method, the zero-crossing time method, etc. In 2007, D’Mellow et al. proposed the pulse gradient method for n / γ discrimination research of liquid scintillator detectors. The pulse gradient method uses the difference between the neutron and γ-ray pulses in the falling edge stage after maximum normalization to perform discrimination. After maximum normalization, the neutron pulse in the falling edge stage is slower than the γ pulse. Calculate the relative gradient of the peak point and the point after the peak, and achieve the purpose of n / γ discrimination through the relative gradient. The pulse gradient method proposed by D’Mellow et al. has been widely used in n / γ discrimination of various detectors. Due to its simple calculation and good pulse passing rate, it can be used for real-time discrimination. However, due to the overly simple value-taking points, the discrimination effect of this method is poor, not as good as other PSD methods. The widely used charge comparison method has a good discrimination effect, but since it realizes discrimination by calculating the integral of the fast and slow components of the pulse, the integral calculation takes a certain amount of time and will reduce the pulse passing rate. Therefore, it is necessary to design a discrimination method that can balance the discrimination effect and the pulse passing rate.

[0004] Among many neutron detectors, the CLYC scintillator can interact with γ-rays to produce the special core-valence luminescence (CVL) characteristic, while the CLYC scintillator does not produce this luminescence characteristic with neutrons. Therefore, the CLYC scintillator has a good n / γ discrimination effect. At the same time, D’Olympia et al. found in their research that there are differences between the normalized neutron and γ-ray pulses measured by the CLYC detector in both the rising edge and falling edge stages, which makes the CLYC detector an excellent n / γ discrimination detector. Summary of the Invention

[0005] (1) Technical Problems to be Solved

[0006] The present invention proposes an n / γ discrimination method based on the combined gradient of the leading and trailing edges of a pulse, and solves the technical problem of how to balance the discrimination effect and a high calculation speed to achieve n / γ pulse discrimination by calculating the bilateral gradient of the leading and trailing edges of the pulse.

[0007] (2) Technical solution

[0008] To solve the above technical problems, the present invention proposes an n / γ discrimination method based on the combined gradient of the leading and trailing edges of a pulse. The n / γ discrimination method includes the following steps:

[0009] S1. Preprocess the n / γ mixed radiation field pulse signal obtained by the detector to obtain the n / γ pulse signal;

[0010] S2. Find the peak point of the preprocessed n / γ mixed radiation field pulse signal, take sample points within a certain time interval before and after the peak point, and calculate the pulse gradients of the two sample points with respect to the peak point respectively. The rising-edge pulse gradient and the falling-edge pulse gradient are respectively expressed as:

[0011]

[0012] In the formula, m leading and m trailing respectively represent the gradients of the sampling points of the pulse signal with respect to the peak point in the rising-edge and falling-edge stages; y p represents the pulse amplitude at the peak point, and takes the value of 1 after amplitude normalization; y d1 and y d2 respectively represent the pulse amplitudes of the sampling points of the pulse signal in the rising-edge and falling-edge stages; Δt1 and Δt2 respectively represent the time intervals between the two sampling points and the peak point in the rising-edge and falling-edge stages;

[0013] S3. Multiply the rising-edge pulse gradient and the falling-edge pulse gradient to obtain the combined gradient m:

[0014] m = m leading × m trailing

[0015] S4. Use the combined gradient m as the pulse shape discrimination parameter, and discriminate the n / γ pulse data by fitting the distribution with a double Gaussian function.

[0016] Furthermore, a detector with differences in output signals at both the leading and trailing edges is used.

[0017] Furthermore, in step S1, the preprocessing method is as follows:

[0018] S1-1. Smooth the n / γ mixed radiation field pulse signal obtained by the detector;

[0019] S1-2. Calculate the pulse baseline by the arithmetic mean method for the baseline part before triggering and subtract it;

[0020] S1-3. Judge and discard the pile-up of the pulse after subtracting the baseline, and discard the invalid pulses with pile-up and false triggering;

[0021] S1-4. Normalize the amplitude of the signal after pile-up rejection to its maximum value;

[0022] Further, in step S1-1, ten-point smoothing is adopted to filter the n / γ mixed radiation field pulse signal.

[0023] Further, in step S2, by means of statistical methods, the peak points of the preprocessed n / γ mixed radiation field pulse signal are found.

[0024] (III) Beneficial effects

[0025] The present invention proposes an n / γ discrimination method based on the joint gradient of the leading and trailing edges of a pulse, including performing preprocessing on the mixed radiation field pulse signal, such as smoothing filtering, baseline subtraction, pile-up rejection, and maximum amplitude normalization; finding the peak points of the preprocessed n / γ pulse signal, respectively selecting sample points at a certain time interval before and after the peak points, and calculating the gradients of the two sample points with respect to the peak point; multiplying the obtained leading and trailing edge gradients of the pulse to obtain the joint gradient value; using the joint gradient value as the pulse waveform discrimination parameter to discriminate the n / γ pulse signal. Compared with the existing pulse gradient method, this method significantly improves the discrimination effect without increasing too much computational complexity. Description of the drawings

[0026] Figure 1 is the flowchart of the n / γ discrimination method based on the joint gradient of the leading and trailing edges of a pulse according to the present invention;

[0027] Figure 2 is the schematic diagram of the typical shapes of neutron and γ pulses after maximum normalization of the output signal of the CLYC detector in the present invention;

[0028] Figure 3 is the schematic diagram of the statistical result of the peak time position of the output signal of the CLYC detector in the present invention;

[0029] Figure 4 is the schematic diagram of the principle of the n / γ discrimination method based on the joint gradient of the leading and trailing edges of a pulse according to the present invention;

[0030] Figure 5a is the two-dimensional histogram of the PSD parameter and the amplitude channel address of the result of the pulse gradient method, Figure 5b is the result of the charge comparison method, Figure 5c is the result of the joint gradient method of the leading and trailing edges of a pulse;

[0031] Figure 6a is the PSD histogram of the result of the pulse gradient method and the Gaussian fitting result, Figure 6b is the result of the charge comparison method, Figure 6c is the result of the joint gradient method of the leading and trailing edges of a pulse. Detailed implementation manners

[0032] To make the objectives, contents and advantages of the present invention clearer, the following further describes in detail the specific implementation manners of the present invention with reference to the accompanying drawings and embodiments.

[0033] This embodiment proposes an n / γ discrimination method based on the combined gradient of the leading and trailing edges of a pulse, and its process is as Figure 1 shown, and specifically includes the following steps:

[0034] S1. Preprocess the n / γ mixed radiation field pulse signal obtained by the CLYC detector to obtain the n / γ pulse signal; the preprocessing method is:

[0035] S1-1. Smooth the n / γ mixed radiation field pulse signal obtained by the CLYC detector. To retain the pulse shape information, perform a ten-point smoothing to filter the signal.

[0036] S1-2. According to the characteristics of the output signal of the CLYC detector, calculate the pulse baseline by using the arithmetic mean method for the baseline part before triggering and subtract it.

[0037] S1-3. Perform pile-up rejection on the pulses after subtracting the baseline, and discard the invalid pulses with pile-up and false triggering.

[0038] S1-4. Perform maximum amplitude normalization on the signal after pile-up rejection, as Figure 2 shown, to further amplify the difference in shape between neutron and γ pulses.

[0039] S2. By using statistical methods, find the peak points of the preprocessed n / γ mixed radiation field pulse signal, and the statistical results are as Figure 3 shown. Take sample points at a time interval before and after each peak point, and calculate the pulse gradients between the two sample points and the peak point respectively; the rising-edge pulse gradient and the falling-edge pulse gradient are respectively expressed as:

[0040]

[0041] In the formula, m leading and m trailing respectively represent the gradients between the sampling points of the pulse signal at the rising edge and the falling edge and the peak point; y p represents the pulse amplitude at the peak point, and takes the value of 1 after the amplitude normalization operation; y d1 and y d2 respectively represent the pulse amplitudes of the sampling points of the pulse signal at the rising edge and the falling edge; Δt1 and Δt2 respectively represent the time intervals between the two sampling points at the rising edge and the falling edge and the peak point.

[0042] Each value-taking point and the relative gradient are asFigure 4 as shown

[0043] S3. Multiply the rising-edge pulse gradient and the falling-edge pulse gradient to obtain the combined gradient m:

[0044] m = m leading × m trailing

[0045] S4. Use the combined gradient m as the pulse shape discrimination parameter, and perform discrimination on the n / γ pulse data by fitting the distribution with a double Gaussian function.

[0046] For the evaluation of the n / γ pulse discrimination result, the evaluation criteria involve the figure of merit and the calculation consumption time. The larger the FoM value, the better the discrimination performance.

[0047] The following verifies the discrimination result and makes a comparison through experiments:

[0048] The experimental equipment and parameter settings are as follows: Select 241 the Am-Be source as the radiation source, and the average energy of the neutron fluence generated is 4.17 MeV. In this experiment, a CLYC scintillation detector containing 6 Li is used, and a digitizer and a high-performance computer are connected for the acquisition and calculation of the n / γ discrimination experiment. The model of the digitizer is DT-5730, the sampling rate is 500 MS / s, the sampling depth is 14 bit, and the acquisition length of a single pulse is 50 μs. The collected neutron and γ pulses are calculated on a high-performance computer, and the calculation software is MATLAB. The main parameters of the combined gradient method of the pulse front and rear edges are that the rising-edge time interval Δt1 = 60 ns and the falling-edge time interval Δt2 = 1120 ns.

[0049] The following uses two discrimination methods to make a comparison with the combined gradient method of the pulse front and rear edges PJGM, namely the charge comparison method CCM and the pulse gradient method PGA. In the subsequent experiments, the adjustable parameters of these discrimination methods have been optimized, and the displayed results are the best discrimination effects achieved by these methods under the experimental conditions.

[0050] Figures 5a - 5c They are respectively the two-dimensional histograms of the PSD parameter - amplitude channel address for realizing n / γ discrimination by using the above three discrimination methods. Figure 5a is the result of the charge comparison method. Figure 5b is the result of the pulse gradient method. Figure 5c is the discrimination result of the combined gradient method of the pulse front and rear edges. In each two-dimensional histogram, the side with a larger PSD parameter is the neutron event, and the smaller side is the γ event. The separation degree of the neutron and γ events represents a better discrimination effect.

[0051] Figure 6 is the result obtained by making a histogram and Gaussian fitting for the events within the red frame in Figure 5, and the figure of merit FoM of the discrimination is also obtained from this.Figure 6a is the result of the charge comparison method, Figure 6b is the result of the pulse gradient method, and Fig. 6 is the fitting and calculation results of the combined gradient method of the leading and trailing edges of the pulse.

[0052] Table 1 Discrimination results and calculation times of different PSD methods (processing 30,000 pulses)

[0053] PSD method CCM PGA PJGM FoM value 3.65 2.06 5.20 Computation time / s 3.88 1.94 3.56

[0054] As shown in Table 1, an objective evaluation is made for the above three discrimination methods. Among them, the discrimination effect of the pulse gradient method is the worst but the calculation time is the shortest, the FoM value is 2.06, and the calculation time is 1.94 s; the discrimination effect of the charge comparison method is 3.65, and the calculation time is 3.88 s; the discrimination effect of the combined gradient method of the leading and trailing edges of the pulse is 5.20, and the calculation time is 3.56 s. It can be seen that the discrimination effect of the combined gradient method of the leading and trailing edges of the pulse is better than that of the charge comparison method and the pulse gradient method. In terms of calculation time, although it is longer than that of the pulse gradient method, it is comparable to the calculation time of the charge comparison method. The combined gradient method of the leading and trailing edges of the pulse has an ideal discrimination effect and calculation time, and the calculation complexity of this method is relatively low, and it can be used for real-time discrimination in subsequent research.

[0055] The above are only the preferred embodiments of the present invention. It should be pointed out that for those of ordinary skill in the art, without departing from the technical principle of the present invention, several improvements and deformations can still be made, and these improvements and deformations should also be regarded as the protection scope of the present invention.

Claims

1. A method for n / γ discrimination based on the joint gradient of the leading and trailing edges of pulses, characterized in that The n / γ discrimination method includes the following steps: S1. Preprocess the pulse signals of the n / γ mixed radiation field obtained by the detector to obtain n / γ pulse signals; S2. Find the peak points of the preprocessed n / γ mixed radiation field pulse signals, take sample points within a certain time interval before and after the peak points, and calculate the pulse gradients of the two sample points with respect to the peak point respectively. The rising-edge pulse gradient and the falling-edge pulse gradient are respectively expressed as: where m leading and m trailing represent the gradients made by the sampling points and the peak points in the rising edge and falling edge stages of the pulse signal respectively; y p represents the pulse amplitude at the peak point, and the value is 1 after amplitude normalization; y d1 and y d2 represent the pulse amplitudes of the sampling points in the rising edge and falling edge stages of the pulse signal respectively; Δt1 and Δt2 represent the time intervals between the two sampling points and the peak point in the rising edge and falling edge stages respectively; S3. Multiply the rising-edge pulse gradient and the falling-edge pulse gradient to obtain the combined gradient m: m = m leading × m trailing S4. Use the combined gradient m as the pulse shape discrimination parameter, and discriminate the n / γ pulse data by fitting the distribution with a double Gaussian function.

2. The n / γ discrimination method based on the combined gradient of the leading and trailing edges of the pulse according to claim 1, wherein The detector uses a detector with output signals that have differences in both the front and rear edges.

3. The n / γ discrimination method based on the combined gradient of the leading and trailing edges of the pulse according to claim 1, characterized in that, In step S1, the preprocessing method is as follows: S1-1. Smooth the pulse signals of the n / γ mixed radiation field obtained by the detector; S1-2. Calculate the pulse baseline by the arithmetic mean method for the baseline part before triggering and subtract it; S1-3. Judge and discard the pile-up of the pulses after subtracting the baseline, and discard the invalid pulses with pile-up and false triggering; S1-4. Perform maximum amplitude normalization processing on the signals after pile-up judgment and discard.

4. The n / γ discrimination method based on the combined gradient of the leading and trailing edges of pulses according to claim 3, wherein In step S1-1, ten-point smoothing is adopted to filter the pulse signals of the n / γ mixed radiation field.

5. The n / γ discrimination method based on the combined gradient of the leading and trailing edges of the pulse according to claim 1, wherein In step S2, the peak points of the preprocessed n / γ mixed radiation field pulse signals are found by statistical methods.

Images