Dual-detector hybrid sampling and adaptive discrimination for high-energy neutron signal processing

Abstract

The application discloses a kind of spatial high-energy neutron signal processing methods based on double probe mixed sampling and adaptive discrimination, belong to spatial high-energy neutron detection signal processing technical field.The method obtains He-3 gas detector pulse signal and plastic scintillator detector pulse signal, and is mixed sampling under the control of global unified clock respectively not less than 40Msps and not less than 500Msps;Waveform is input into FPGA after sampling, and is cooperatively triggered, waveform delay processing, digital filtering, peak seeking, adaptive baseline drift correction, monte carlo optimization long-short gate integration PSD discrimination and PHA analysis are carried out.The method can improve the discrimination accuracy of spatial high-energy neutron and gamma ray, reduce the influence of baseline drift and noise, meet the demand of spaceborne real-time detection.

Description

Technical Field

[0001] This invention belongs to the field of space high-energy neutron detection signal processing technology, specifically relating to a high-energy neutron signal processing method with dual-detector hybrid sampling and adaptive discrimination. Background Technology

[0002] Real-time measurement of high-energy neutrons in space is of fundamental scientific value for assessing space radiation risks, designing spacecraft protective structures, and conducting radiation risk assessments for commercial space COTS (Coal-to-Shipment) systems. Currently, ground-based high-energy neutron detection electronics are relatively mature, but their weight, power consumption, and size make them difficult to directly apply to space exploration platforms with strict payload weight and power consumption limitations.

[0003] Current space neutron detection signal processing methods mostly employ analog discrimination or fixed-gated-width digital discrimination. Analog discrimination is easily affected by temperature, power supply, and count rate variations in the long-term space operating environment; fixed-gated-width digital discrimination is difficult to adapt to the waveform differences of neutron pulses and gamma-ray pulses of different energies, and suffers from problems such as poor noise immunity, low-energy discrimination failure, and severe baseline drift, which cannot meet the requirements for long-term reliable operation on spacecraft.

[0004] Furthermore, space-based high-energy neutron detection systems typically need to simultaneously address high-energy neutron and gamma-ray resolution, weak signal amplitude extraction, high-count-rate real-time processing, and sampling synchronization. A single detector or sampling strategy is insufficient to simultaneously handle the weak, slow pulse signals output by the He-3 gas detector and the nanosecond-level fast pulse signals output by the plastic scintillator detector. Therefore, a space-based high-energy neutron signal processing method based on dual-detector hybrid sampling and adaptive discrimination within an FPGA (real-time hardware processing core) is required to improve neutron-gamma ray discrimination accuracy, baseline stability, real-time processing capability, and high-count-rate data integrity. Summary of the Invention

[0005] The technical problem to be solved by this invention is: to address the problems of low neutron-gamma discrimination accuracy, weak real-time processing capability, baseline drift affecting measurement accuracy, poor discrimination effect in the low energy range, insufficient high-speed sampling synchronization, and insufficient data integrity under high count rate in existing spaceborne space neutron detection signal processing methods, and to provide a high-energy neutron signal processing method with dual detector hybrid sampling and adaptive discrimination. Technical solution

[0006] To address the aforementioned technical problems, this invention provides a high-energy neutron signal processing method using dual-detector hybrid sampling and adaptive discrimination, applicable to a space high-energy neutron detection system including a He-3 gas detector, a plastic scintillator detector, an analog-to-digital converter, and an FPGA. The method includes the following steps: Step S1: Acquire the gas detector pulse signal output from the He-3 gas detector and the plastic scintillator detector pulse signal output from the plastic scintillator detector. (See attached document) Figure 2 In a preferred embodiment, the gas detector pulse signal is formed by the weak charge signal output from the He-3 gas detector after charge-sensitive preamplification, pole-zero cancellation, and quasi-Gaussian shaping filtering; the plastic scintillator pulse signal is formed by the nanosecond-level fast pulse signal output from the plastic scintillator detector via a PMT (photomultiplier tube) after voltage following and impedance matching processing.

[0007] Step S2: Under the control of a globally unified clock, the gas detector pulse signal and the plastic scintillator pulse signal are sampled by analog-to-digital conversion to obtain the gas detector sampling waveform and the plastic scintillator sampling waveform, respectively. The plastic scintillator pulse signal is acquired at a sampling rate of not less than 500 Msps, and the gas detector pulse signal is acquired at a sampling rate of not less than 40 Msps. In a preferred embodiment, the plastic scintillator pulse signal is acquired at a sampling specification of not less than 500 Msps and 14 bits, and the gas detector pulse signal is acquired at a sampling specification of not less than 40 Msps and 14 bits; and before sampling, the single-ended signal is converted into a differential signal and then input into the corresponding analog-to-digital converter.

[0008] Step S3: Input the sampled waveforms of the gas detector and the plastic scintillator into the FPGA, and perform coordinated triggering and waveform delay processing within the FPGA to record the sampled waveforms before and after triggering. In a preferred embodiment, the coordinated triggering includes the FPGA generating a trigger signal based on the sampled waveforms of the gas detector and / or the plastic scintillator; the waveform delay processing includes retaining sampled waveforms of preset lengths before and after the generation of the trigger signal, and writing the sampled waveforms before and after triggering, event data, or intermediate processing data into a FIFO (First-In-First-Out buffer) and a RAM (Random Access Memory).

[0009] Step S4: In the FPGA, digital filtering, peak finding, and adaptive baseline drift correction are performed on the sampled waveforms before and after triggering; wherein, the adaptive baseline drift correction includes dynamically extracting baseline data during the pulse gap, calculating the baseline drift amount through moving average filtering, and compensating the sampled waveform in real time according to the baseline drift amount.

[0010] In a preferred embodiment, the digital filtering employs FIR (Finite Impulse Response) digital filtering, and the peak finding employs adaptive threshold peak finding; the adaptive threshold is adjusted based on the baseline level or noise state in the sampled waveform.

[0011] Step S5: Within the FPGA, based on the long and short gate integration parameters obtained through Monte Carlo optimization, the total integrated charge and tail integrated charge of the sampled waveform to be discriminated are calculated in real time for the waveform after adaptive baseline drift correction. Pulse shape discrimination is then performed based on the ratio of the total integrated charge to the tail integrated charge to distinguish between high-energy neutron events and gamma-ray events. (See attached reference.) Figure 4 Before step S5, a step of optimizing the long and short gate integration parameters may also be included: based on the waveform differences between neutron pulses and gamma-ray pulses of different energies, a waveform model is established through Monte Carlo simulation, and the parameters of the short integration gate, long integration gate, tail integration gate and delay gate are optimized, and the optimized long and short gate integration parameters are configured to the FPGA.

[0012] In a preferred embodiment, the optimization step of the long and short gate integration parameters includes: calculating the total integrated charge, tail integrated charge, and charge ratio of neutron pulse samples and gamma-ray pulse samples for different combinations of candidate short integration gate, candidate long integration gate, candidate tail integration gate, and candidate delay gate parameters, and determining the long and short gate integration parameters for real-time discrimination based on the degree of separation between neutron pulse samples and gamma-ray pulse samples in the charge ratio.

[0013] In a preferred embodiment, the pulse shape discrimination includes: calculating the total integrated charge of the sampled waveform to be discriminated within a long integration gate, calculating the tail integrated charge of the sampled waveform to be discriminated within a tail integration gate, calculating the ratio of the tail integrated charge to the total integrated charge, comparing the ratio with a preset discrimination threshold, and determining the corresponding event as a high-energy neutron event or a gamma-ray event based on the comparison result. Let the total integrated charge be Q_total, the tail integrated charge be Q_tail, and the charge ratio be R, then: R = Q_tail / Q_total; Step S6 involves extracting the pulse amplitude of the sampled waveforms identified as high-energy neutron events by pulse shape discrimination, and counting them by amplitude channel to obtain the statistical results of space high-energy neutron detection. In a preferred embodiment, the pulse amplitude extraction includes: extracting the pulse peak value or pulse amplitude from the sampled waveforms after adaptive baseline drift correction and digital filtering, mapping the pulse peak value or pulse amplitude to the corresponding amplitude channel address, and independently counting the identified high-energy neutron events.

[0014] In a preferred embodiment, the method further includes a data output step: outputting the space high-energy neutron detection statistics, event data, and / or sampled waveform data via an optical fiber network. Beneficial effects

[0015] Compared with existing technologies, this invention provides a high-energy neutron signal processing method with dual-detector hybrid sampling and adaptive discrimination, which has the following beneficial effects: 1. This invention addresses the different signal characteristics of He-3 gas detectors and plastic scintillator detectors by employing gas detector channel sampling of no less than 40 Msps and plastic scintillator channel sampling of no less than 500 Msps, which can simultaneously capture slow pulse waveforms and capture nanosecond-level fast pulse details.

[0016] 2. This invention uses a unified clock to control the sampling timing of two types of ADCs and supports multi-channel collaborative triggering within the FPGA, which can improve high-speed sampling synchronization and sampling integrity.

[0017] 3. This invention records the complete sampling waveform before and after triggering through waveform delay processing, and combines FIFO and RAM caching mechanisms to reduce the risk of data loss at high count rates.

[0018] 4. This invention dynamically extracts baseline data during pulse gaps, calculates baseline drift through moving average filtering, and compensates for it in real time. This can suppress baseline shifts caused by changes in temperature, power supply, and count rate, and improve the accuracy of weak signal amplitude extraction.

[0019] 5. This invention optimizes the parameters of the short integration gate, long integration gate, tail integration gate, and delay gate through Monte Carlo simulation, enabling the long and short gate integration PSD discrimination to adapt to the waveform differences between neutron pulses and gamma-ray pulses of different energies, thereby improving the neutron-gamma discrimination accuracy across the entire energy range.

[0020] 6. This invention calculates the ratio of total integrated charge to tail integrated charge in real time within an FPGA, and distinguishes between high-energy neutron events and gamma-ray events by threshold comparison. It does not rely on complex software post-processing, has low processing latency, and is suitable for spaceborne real-time detection missions.

[0021] 7. This invention integrates adaptive baseline drift correction, digital filtering, adaptive threshold peak finding, PSD discrimination, and PHA analysis into a fully digital real-time FPGA processing flow, which can meet the requirements of real-time performance, stability, and reliability for long-term on-orbit operation of micro-nano satellite platforms. Attached Figure Description

[0022] Figure 1 This is a flowchart of the overall process of the space high-energy neutron signal processing method based on dual-detector hybrid sampling and adaptive discrimination of the present invention.

[0023] Figure 2 This is a flowchart of the dual-detector signal acquisition and mixed sampling process in this invention.

[0024] Figure 3 This is a flowchart of the adaptive baseline drift correction and pulse amplitude analysis in this invention.

[0025] Figure 4This is a flowchart of the Monte Carlo optimization long and short gate integral PSD discrimination process in this invention. Detailed Implementation

[0026] The present invention will be further described below with reference to the accompanying drawings. It should be understood that the following embodiments are used to illustrate the present invention. This is not intended to limit the scope of protection of the present invention; without departing from the concept of the present invention, those skilled in the art can make conventional adjustments to the sampling rate, integration gate parameters, threshold parameters, buffer length and data output format according to actual engineering needs.

[0027] Example 1, see attached document Figure 1 This embodiment provides a space high-energy neutron signal processing method based on dual-detector hybrid sampling and adaptive discrimination. This method is applied to a space high-energy neutron detection system including a He-3 gas detector, a plastic scintillator detector, an ADC, and an FPGA.

[0028] First, the output signals of the dual detectors are acquired. The weak charge signal output by the He-3 gas detector is converted into a voltage signal by a charge-sensitive preamplifier, then the long-tailed pole of the preamplifier is canceled by pole-zero cancellation, and finally filtered by a quasi-Gaussian shaping filter to form a gas detector pulse signal that meets the ADC range. The plastic scintillator detector outputs a nanosecond-level fast pulse signal through a PMT, and this fast pulse signal is processed by voltage following and impedance matching to form the plastic scintillator detector pulse signal.

[0029] Then, refer to the appendix. Figure 2 Hybrid sampling is employed. The plastic scintillator pulse signal, after single-ended to differential conversion, is acquired using a high-speed ADC with a speed of at least 500 Msps and a 14-bit interface; the gas detector pulse signal, after single-ended to differential conversion, is acquired using an ADC with a speed of at least 40 Msps and a 14-bit interface. The sampling timing of both types of ADCs is controlled by a globally unified clock to ensure phase consistency across multiple ADCs and reduce aperture jitter.

[0030] The sampled waveforms from the gas detector and the plastic scintillator are fed into the FPGA. The FPGA generates a coordinated trigger signal based on the sampled waveforms and records the complete waveforms before and after triggering through waveform delay processing. The waveform data before and after triggering, event data, and intermediate processing data are written to the FIFO and RAM for subsequent processing. This process improves data integrity in high count rate environments.

[0031] Example 2, see attached document Figure 3 Based on Example 1, this example illustrates the adaptive baseline drift correction, digital filtering, and peak finding process.

[0032] The FPGA dynamically extracts baseline data during the pulse gaps between adjacent pulses. This baseline data is then filtered using a moving average to obtain the baseline drift. The FPGA uses this baseline drift to compensate for subsequent sampled waveforms in real time, thereby suppressing baseline shifts caused by temperature, power supply, and count rate fluctuations.

[0033] Specifically, during the pulse gap where no valid pulse is detected, the FPGA selects multiple consecutive sampling points as baseline sampling points and performs a moving average calculation on these baseline sampling points to obtain the current baseline estimate. When processing subsequent sampled waveforms, the FPGA subtracts the current baseline estimate from the corresponding sampled values ​​to obtain the baseline-compensated sampled values. The current baseline estimate is dynamically updated with the baseline sampling points in the pulse gap, thereby achieving adaptive baseline drift correction.

[0034] After baseline drift correction, the FPGA performs FIR digital filtering on the sampled waveform to reduce noise interference. Subsequently, the FPGA employs an adaptive threshold peak-finding method to extract the effective pulse position and pulse amplitude. The adaptive threshold can be adjusted based on the baseline level or noise condition to reduce noise-induced false triggering and improve the accuracy of weak signal amplitude extraction.

[0035] Example 3, see attached document Figure 4 Based on Example 1 or Example 2, this example illustrates the Monte Carlo optimization process for the long and short gate integral PSD discrimination.

[0036] During the parameter optimization phase, based on the waveform differences between neutron pulses and gamma-ray pulses of different energies, a waveform model was established through Monte Carlo simulation. The parameters of the short integration gate, long integration gate, tail integration gate, and delay gate were optimized to obtain long and short gate integration parameters suitable for real-time discrimination by the FPGA. These long and short gate integration parameters were then configured into the FPGA.

[0037] Specifically, multiple candidate short integration gate, candidate long integration gate, candidate tail integration gate, and candidate delay gate parameter combinations are defined. For each set of candidate parameters, neutron pulse samples and gamma-ray pulse samples are integrated separately to obtain the total integrated charge and the tail integrated charge, and the ratio of the tail integrated charge to the total integrated charge is calculated. The degree of separation between neutron pulse samples and gamma-ray pulse samples in the ratio under different candidate parameter combinations is compared, and the parameter combination with the higher degree of separation is used as the optimized long and short gate integration parameters.

[0038] During the real-time discrimination phase, the FPGA performs integration processing on the sampled waveform to be discriminated based on the long and short gate integration parameters. The total integrated charge Q_total is calculated within the long integration gate, and the tail integrated charge Q_tail is calculated within the tail integration gate. The charge ratio R = Q_tail / Q_total is then calculated. The FPGA compares R with a preset discrimination threshold, and based on the comparison result, classifies the corresponding event as either a high-energy neutron event or a gamma-ray event.

[0039] Through the above processing, this embodiment can overcome the problem that traditional fixed-gate width digital PSDs cannot adapt to neutron waveforms of different energies, improve the neutron-gamma discrimination accuracy across the entire energy range, and improve the discrimination effect in the low-energy range.

[0040] Example 4, see attached document Figure 3 Based on Example 3, this example illustrates the PHA analysis and neutron event statistics process.

[0041] For sampled waveforms identified as high-energy neutron events by the PSD, the FPGA extracts the pulse peak value or pulse amplitude and counts them according to a preset amplitude channeling rule to form the space high-energy neutron detection statistics. Data identified as gamma-ray events by the PSD are not included in the neutron event statistics, or are removed from the neutron statistics results.

[0042] Specifically, the FPGA determines the pulse peak value or pulse amplitude based on the sampled waveform after baseline compensation and digital filtering, and assigns the pulse peak value or pulse amplitude to the corresponding amplitude channel; when the corresponding event is determined to be a high-energy neutron event, the count of the corresponding amplitude channel is incremented by one. During continuous detection, the counts of each amplitude channel accumulate to form the statistical results of space high-energy neutron detection.

[0043] Finally, the FPGA outputs the event data, sampled waveform data, and / or space high-energy neutron detection statistics via a fiber optic network. Therefore, this embodiment enables real-time adaptive discrimination and neutron event statistics based on dual-detector hybrid sampling in space high-energy neutron detection scenarios.

[0044] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A high-energy neutron signal processing method using dual-detector hybrid sampling and adaptive discrimination, applied to a space high-energy neutron detection system including a He-3 gas detector, a plastic scintillator detector, an analog-to-digital converter, and an FPGA, characterized in that, Includes the following steps: S1, acquire the gas detector pulse signal output by the He-3 gas detector and the plastic scintillator detector pulse signal output by the plastic scintillator detector; S2, under the control of a globally unified clock, the gas detector pulse signal and the plastic scintillator pulse signal are sampled by analog-to-digital conversion to obtain the gas detector sampling waveform and the plastic scintillator sampling waveform; wherein, the plastic scintillator pulse signal is acquired at a sampling rate of not less than 500Msps, and the gas detector pulse signal is acquired at a sampling rate of not less than 40Msps. S3, input the sampled waveforms of the gas detector and the plastic flash detector into the FPGA, and perform coordinated triggering and waveform delay processing in the FPGA to record the sampled waveforms before and after triggering; S4, within the FPGA, digital filtering, peak finding, and adaptive baseline drift correction are performed on the sampled waveforms before and after triggering; wherein, the adaptive baseline drift correction includes dynamically extracting baseline data during pulse gaps, calculating the baseline drift amount through moving average filtering, and compensating the sampled waveform in real time based on the baseline drift amount; S5. Based on the long and short gate integration parameters obtained by Monte Carlo optimization within the FPGA, the total integrated charge and tail integrated charge of the sampled waveform to be identified after adaptive baseline drift correction are calculated in real time, and the pulse shape is identified according to the ratio of the total integrated charge to the tail integrated charge to distinguish between high-energy neutron events and gamma-ray events. S6 extracts the pulse amplitude of the sampled waveforms that are identified as high-energy neutron events by pulse shape discrimination, and counts them by amplitude channel to obtain the statistical results of space high-energy neutron detection.

2. The high-energy neutron signal processing method with dual-detector hybrid sampling and adaptive discrimination according to claim 1, characterized in that, In step S1, the gas detector pulse signal is formed by the weak charge signal output by the He-3 gas detector after charge-sensitive preamplification, pole-zero cancellation and quasi-Gaussian shaping filtering. The plastic scintillator pulse signal is formed by voltage following and impedance matching processing of the nanosecond-level fast pulse signal output by the plastic scintillator detector through the photomultiplier tube.

3. The high-energy neutron signal processing method with dual-detector hybrid sampling and adaptive discrimination according to claim 1, characterized in that, In step S2, the plastic scintillator pulse signal is acquired using a sampling specification of not less than 500 Msps and 14 bits, and the gas detector pulse signal is acquired using a sampling specification of not less than 40 Msps and 14 bits; and before sampling, the single-ended signal is converted into a differential signal and then input into the corresponding analog-to-digital converter.

4. The high-energy neutron signal processing method with dual-detector hybrid sampling and adaptive discrimination according to claim 1, characterized in that, In step S3, the coordinated triggering includes: the FPGA generating a trigger signal based on the sampled waveform of the gas detector and / or the sampled waveform of the plastic scintillator; the waveform delay processing includes: retaining sampled waveforms of a preset length before and after the generation of the trigger signal, and writing the sampled waveforms before and after the triggering, event data or intermediate processing data into a first-in-first-out buffer and a random access memory.

5. The high-energy neutron signal processing method with dual-detector hybrid sampling and adaptive discrimination according to claim 1, characterized in that, In step S4, the digital filtering adopts finite impulse response digital filtering, and the peak finding adopts adaptive threshold peak finding; the adaptive threshold is adjusted according to the baseline level or noise state in the sampled waveform.

6. The high-energy neutron signal processing method with dual-detector hybrid sampling and adaptive discrimination according to claim 1, characterized in that, Before step S5, there is also a step to optimize the long and short gate integration parameters: based on the waveform differences between neutron pulses and gamma-ray pulses of different energies, a waveform model is established through Monte Carlo simulation, and the parameters of the short integration gate, long integration gate, tail integration gate and delay gate are optimized, and the optimized long and short gate integration parameters are configured into the FPGA.

7. The high-energy neutron signal processing method with dual-detector hybrid sampling and adaptive discrimination according to claim 6, characterized in that, The optimization steps for the long and short gate integration parameters include: calculating the total integrated charge, tail integrated charge, and charge ratio of neutron pulse samples and gamma-ray pulse samples for different combinations of candidate short integration gate, candidate long integration gate, candidate tail integration gate, and candidate delay gate parameters, and determining the long and short gate integration parameters for real-time discrimination based on the degree of separation between neutron pulse samples and gamma-ray pulse samples in the charge ratio.

8. The high-energy neutron signal processing method with dual-detector hybrid sampling and adaptive discrimination according to claim 1, characterized in that, In step S5, the pulse shape discrimination includes: calculating the total integrated charge of the sampled waveform to be discriminated within a long integration gate, calculating the tail integrated charge of the sampled waveform to be discriminated within a tail integration gate, calculating the ratio of the tail integrated charge to the total integrated charge, comparing the ratio with a preset discrimination threshold, and determining the corresponding event as a high-energy neutron event or a gamma-ray event based on the comparison result.

9. The high-energy neutron signal processing method with dual-detector hybrid sampling and adaptive discrimination according to claim 1, characterized in that, In step S6, the pulse amplitude extraction includes: extracting the pulse peak or pulse amplitude from the sampled waveform after adaptive baseline drift correction and digital filtering, mapping the pulse peak or pulse amplitude to the corresponding amplitude channel address, and performing independent statistics on the identified high-energy neutron events.

10. The high-energy neutron signal processing method with dual-detector hybrid sampling and adaptive discrimination according to claim 1, characterized in that, The method further includes a data output step: outputting the statistical results, event data, and / or sampled waveform data of the space high-energy neutron detection through an optical fiber network.

Images