Portable nuclide identification system based on arrayed CZT detector

Abstract

The present application relates to radionuclide detection technical field, especially portable nuclide identification system based on array CZT detector, including detector bias high voltage establishment module, raw energy spectrum acquisition module, high voltage power-off emergency protection module, analysis thread module and energy spectrum UI display module, it is through PID control to realize high voltage uniform speed rise and fall and accurate voltage stabilizing, provide reliable bias for CZT detector, complete high-precision conversion of gamma ray energy and data verification transmission in combination with array CZT detector and special ASIC readout circuit, and real-time cumulative storage to energy spectrum data, power seamless switching and soft voltage drop shutdown can also be realized, guarantee software and hardware and data security, through analysis thread module, through preprocessing, peak fitting, overlapping peak intelligent disassembly and branch ratio check, complete nuclide qualitative identification and activity concentration quantitative analysis, and through LTTB algorithm downsampling, realize energy spectrum smooth display, support touch screen interaction and nuclide information real-time labeling.

Description

Technical Field

[0001] This invention relates to the field of radionuclide detection technology, and in particular to a portable nuclide identification system based on an array-type CZT detector. Background Technology

[0002] In online monitoring scenarios of nuclear facilities, such as primary loop coolant monitoring, liquid effluent detection, and aerosol online monitoring, portable radionuclide analysis terminals are required to have long-term stability, high resolution, and strong anti-interference capabilities. CZT (cadmium zinc telluride) detectors, as core detection components, are widely used in portable radionuclide identification devices due to their advantages such as high energy resolution, room temperature operation, and small size. Their normal operation requires a high voltage bias of about 2000V and they are extremely sensitive to power supply stability and voltage rise and fall timing. Once a power supply abnormality occurs, it can easily cause permanent damage to the detector.

[0003] Currently, portable radionuclide identification devices generally suffer from the following technical shortcomings, making it difficult to meet the high-precision and high-reliability requirements of online monitoring of nuclear facilities:

[0004] 1. The high-voltage power supply system lacks power failure safety protection: When the external power supply is suddenly interrupted, the 2000V high-voltage bias of the CZT detector will drop instantly, forming a violent reverse electric field and thermal stress impact inside the crystal, which can easily cause irreversible damage to the detector crystal, seriously reduce the reliability and service life of the equipment, and may also lead to the failure to detect nuclear radiation hazards due to the interruption of detection.

[0005] 2. Difficulty in balancing high-resolution energy spectrum analysis and real-time interface performance: Traditional analytical algorithms are not effective at removing noise and background from the energy spectrum, have weak ability to identify and decompose overlapping peaks, and are prone to missing characteristic peaks masked by the main peak. Furthermore, they rely solely on simple peak position matching for nuclide identification without combining branch ratio verification and quantitative calculation, resulting in low accuracy of qualitative nuclide identification and large errors in quantitative analysis results. This makes it difficult to meet the requirements of high-precision detection. In addition, the energy spectrum data volume is huge, and conventional systems do not have efficient downsampling processing algorithms, which easily leads to problems such as stuttering and unsmooth display of energy spectrum curves.

[0006] Therefore, there is an urgent need for an integrated portable nuclide identification solution that can simultaneously achieve high-voltage safety protection of CZT detectors, high real-time performance spectrum analysis, and intelligent disassembly of overlapping peaks, in order to solve the traditional technical pain points and meet the online monitoring needs of nuclear facilities. Summary of the Invention

[0007] The purpose of this invention is to provide a portable nuclide identification system based on an array-type CZT detector. This system achieves high-voltage, precise, and stable control of the detector, efficient and high-fidelity energy spectrum acquisition and transmission, dual protection of hardware and software data in case of power failure, intelligent, precise qualitative and quantitative identification of nuclides, and human-computer interaction adapted to portable scenarios by optimizing the design and collaborative working logic of each module, thereby solving the aforementioned technical defects.

[0008] The objective of this invention can be achieved through the following technical solution: a portable nuclide identification system based on an array-type CZT detector, comprising a detector bias high voltage establishment module, a raw energy spectrum acquisition module, a high voltage power failure emergency protection module, an analysis thread module, and an energy spectrum UI display module;

[0009] The detector bias high voltage establishment module is used to receive the high voltage application command sent by the main processor through the coprocessor, and to perform the detector bias high voltage establishment operation in response to the high voltage application command, and output the high voltage ready status frame.

[0010] The raw energy spectrum acquisition module uses an array-type CZT detector and a built-in dedicated ASIC readout circuit to convert gamma-ray energy into electrical pulse signals and digitize them into energy spectrum data.

[0011] The high-voltage power failure emergency protection module is used to monitor the status of the external power supply in real time and execute emergency power supply switching and soft voltage reduction shutdown procedures when the external power supply is interrupted.

[0012] The parsing thread module is used to obtain the currently accumulated raw energy spectrum data from shared memory, and to perform background subtraction, peak shape correction and overlapping peak decomposition on the collected raw energy spectrum data;

[0013] The energy spectrum UI display module uses a preset LTTB algorithm to downsample the energy spectrum data and achieves graphical display and interaction.

[0014] Preferably, the analysis process of the high-voltage ready state frame is as follows:

[0015] Based on the high voltage application command sent by the main processor, the coprocessor outputs a linearly increasing control voltage to the high voltage module through the DAC pin. The high voltage module generates a corresponding high voltage output according to the control voltage and applies it to the array-type CZT detector. The voltage boosting process increases at a set slope at a constant speed.

[0016] The output voltage feedback value of the high-voltage module is sampled in real time, and a PID control algorithm is used to compare the feedback value with the target value and dynamically adjust the DAC output.

[0017] Once the target voltage value is reached and stabilized, the coprocessor sends a high-voltage ready status frame to the main processor.

[0018] Preferably, the process for acquiring and analyzing the energy spectrum data is as follows:

[0019] The array-type CZT detector operates under the bias of the target voltage value. After the gamma rays are incident, they interact with the crystal to generate electron-hole pairs. The electrons drift to the anode under the action of the electric field, forming a charge pulse signal.

[0020] The preamplifier of the ASIC readout circuit converts the charge pulse into a voltage pulse, and the shaping circuit shapes the voltage pulse into a quasi-Gaussian waveform, which is compared with a set energy threshold to eliminate noise pulses.

[0021] The output voltage of the peak hold circuit is converted into a digital value by the ADC, and the digital value is converted into the corresponding channel address value. The energy spectrum accumulator inside the ASIC circuit increments the counter corresponding to the channel address value by 1 to form energy spectrum data.

[0022] Preferably, the ASIC circuit encapsulates the data frame according to a preset data frame format and performs integrity verification on the encapsulated data frame: checking whether the frame header and frame tail match, calculating the CRC check value and comparing it with the CRC field in the frame, checking whether the detection unit number is within the valid range, parsing the data frame that passes the verification, and extracting the energy spectrum data block; discarding the data frame that fails the verification, recording the frame loss count, and reporting the error log.

[0023] The main processor accumulates the energy spectrum data by channel address to form a cumulative energy spectrum, which is stored in shared memory as a 4096-channel floating-point array.

[0024] Preferably, the analysis process of the high-voltage power outage emergency protection module is as follows:

[0025] The voltage signal of the external power supply after step-down rectification is continuously sampled at a preset period. The sampled value is compared with a preset threshold to determine whether the external power supply is stable. When the sampled value is lower than the preset threshold for three consecutive times, the external power supply is determined to be interrupted, triggering a seamless switching operation.

[0026] After the seamless switching operation, the user can click the "Continue Data Collection" button to cancel the automatic shutdown and further monitor the battery voltage. If the voltage is higher than the preset voltage value, the operation will continue. If the voltage is lower than the preset voltage value, the countdown will be continuously checked to see if it has reached zero and if the battery voltage is lower than the preset danger threshold. If either condition is met, a forced safety shutdown process will be initiated.

[0027] Preferably, the soft voltage reduction shutdown procedure is as follows:

[0028] Sa: Stop data acquisition, parsing and writing; Sb: Perform soft voltage reduction; Sc: System shutdown.

[0029] Preferably, the background subtraction, peak shape correction, and overlapping peak decomposition process of the original energy spectrum data is as follows:

[0030] The raw energy spectrum data is preprocessed to obtain the net energy spectrum data. The second derivative peak finding algorithm is executed on the net energy spectrum data to identify candidate characteristic peaks. For each candidate characteristic peak, the peak energy, peak area, half width at half maximum (FWHM), and net count are calculated. The peak energy is preliminarily compared with the nuclide database to screen the candidate nuclide list.

[0031] For the selected candidate characteristic peaks, a preset KTW asymmetric Gaussian model is introduced for fine fitting. The half-width at half maximum (WHM) of the corrected candidate peaks is compared with the calibrated half-width at half maximum (WHM) of the single-energy peaks. The output is then transferred to the intelligent decomposition analysis step of overlapping peaks or to perform nuclide qualitative identification.

[0032] Preferably, the method also includes the following steps for intelligent disassembly and analysis of overlapping peaks:

[0033] T1: Peak Fitting and Subtraction: In the overlapping peak group, the characteristic peak with the highest energy is identified as the main peak. If the energies are similar, the peak with the largest net count is selected as the main peak. The KTW asymmetric Gaussian model is used to accurately fit the main peak, and its peak position, peak area and tailing coefficient are determined. The fitted main peak curve is subtracted from the original net energy spectrum one channel at a time to generate the residual energy spectrum.

[0034] T2: Secondary peak finding of residual spectrum: For the residual energy spectrum after removing the main peak, the second derivative peak finding algorithm is re-executed to extract the secondary peak information that is covered by the main peak. If there are still abnormal peak widths in the residual spectrum, T1-T2 is repeated for multiple rounds of iterative stripping.

[0035] T3: Branch ratio verification and nuclide confirmation: Match the energy of all extracted peak positions with the nuclide database to obtain a list of candidate nuclides for each peak position. For multiple characteristic peaks of the same candidate nuclide, calculate the peak area ratio as the measured branch ratio. Compare the measured branch ratio with the standard branch ratio in the database to calculate the matching degree. When the matching degree is greater than the preset matching degree threshold, the existence of the nuclide is confirmed.

[0036] For the confirmed nuclide, the identification confidence level is calculated based on the branch ratio matching degree and peak area statistical error, and the nuclide identification result is output.

[0037] T4: Quantitative analysis: For confirmed nuclides, calculate the nuclide activity concentration A based on its peak area, detector detection efficiency, measurement time, and decay correction factor.

[0038] Preferably, the analysis process of the energy spectrum UI display module is as follows:

[0039] The raw energy spectrum data in the preset energy spectrum format is divided into M data buckets (M>0). Each bucket contains k = preset energy spectrum format / M raw data points. The first data point is selected for the first bucket, and the last data point is selected for the last bucket. For each intermediate bucket, the representative point of the previous bucket and the first point of the next bucket are used as the two vertices of a triangle. The area of ​​the triangle formed by each point in the current bucket and these two vertices is calculated, and the point with the largest area is selected as the representative point of the bucket.

[0040] Combine the representative points of all buckets in order to form downsampled energy spectrum data. Use the downsampled energy spectrum data as the data source of the QCPGraph object and set the curve color, line width, and fill style.

[0041] Simultaneously, it listens to the user's two-finger zoom gesture on the touchscreen to obtain the zoom ratio and center point, recalculates the energy range of the energy spectrum display, and listens to the user's single-finger drag gesture to obtain the offset, translates the current display window, and refreshes the energy spectrum curve in real time.

[0042] When a user clicks or long-presses the energy spectrum curve, the system calculates the energy address corresponding to the clicked location, queries the nuclide information near that energy in the current analysis results, and displays the nuclide name, activity, and confidence level in a pop-up label box on the interface.

[0043] The beneficial effects of this invention are as follows:

[0044] (1) This invention uses a PID control algorithm to achieve high voltage uniform speed rise and fall and precise voltage stabilization, avoiding voltage sudden change damage to CZT crystal. At the same time, it provides real-time feedback on the working status, ensuring that the detector operates under the best working conditions and extending the service life of the hardware. Furthermore, by using an array-type CZT detector with a dedicated ASIC readout circuit, it achieves high-precision and high-resolution conversion of γ-ray energy. Meanwhile, it eliminates erroneous frames through multiple data verifications and combines real-time accumulation and shared memory storage to provide complete and accurate raw data for analysis.

[0045] (2) This invention can also monitor the power supply status in real time and achieve seamless switching. It supports user-intervention shutdown design. The soft voltage reduction shutdown process first completes the data writing to disk and then gradually reduces the voltage to shut down, which protects the detector hardware and prevents the loss of key energy spectrum data and operation logs. Finally, after multiple preprocessing steps to remove noise, combined with fine peak shape fitting and multi-round iterative overlapping peak decomposition algorithm, it effectively identifies the masked feature peaks and solves the problem that traditional methods cannot accurately analyze overlapping peaks. Combined with the branch ratio verification of the nuclide database, the existence of nuclides is determined by the matching degree between the measured value and the standard value. At the same time, the identification confidence and activity concentration are calculated to achieve the dual goals of accurate qualitative identification and accurate quantitative analysis of nuclides. Furthermore, it downsamples the massive energy spectrum data, which greatly reduces the amount of data while ensuring the complete preservation of energy spectrum features, and achieves smooth graphical display of energy spectrum curves. Attached Figure Description

[0046] The invention will now be further described with reference to the accompanying drawings;

[0047] Figure 1 This is a flowchart of the system of the present invention;

[0048] Figure 2 This is a flowchart analyzing the soft voltage reduction shutdown process of the present invention;

[0049] Figure 3 This is a reference diagram for partial analysis of the thread module. Detailed Implementation

[0050] The technical solutions of 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.

[0051] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of the invention. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments;

[0052] Example 1: Please refer to Figures 1 to 3 As shown, this invention is a portable nuclide identification system based on an array-type CZT detector, including a detector bias high voltage establishment module, a raw energy spectrum acquisition module, a high voltage power failure emergency protection module, an analysis thread module, and an energy spectrum UI display module. The detector bias high voltage establishment module and the raw energy spectrum acquisition module are connected in a one-way communication. The raw energy spectrum acquisition module is connected in a one-way communication with the high voltage power failure emergency protection module, the analysis thread module, and the energy spectrum UI display module. The high voltage power failure emergency protection module and the analysis thread module are connected in a one-way communication with the energy spectrum UI display module.

[0053] The detector bias high voltage establishment module is used to receive high voltage application commands sent by the main processor through the coprocessor, and to perform detector bias high voltage establishment operations in response to the high voltage application commands, outputting a high voltage ready status frame, specifically including:

[0054] The HC32 coprocessor receives high-voltage application commands sent by the i.MX8MP main processor via an isolated serial port. The command data frame contains parameters such as the target voltage value (e.g., 2000V), the boost slope (e.g., 50V / s), and the voltage stability accuracy (e.g., 2000V±0.5%).

[0055] The coprocessor outputs a linearly increasing control voltage to the high-voltage module through the DAC pin. The high-voltage module generates a corresponding high-voltage output based on the control voltage and applies it to the array-type CZT detector. The voltage boosting process increases at a set slope at a constant speed to avoid voltage sudden changes from impacting the CZT crystal.

[0056] The coprocessor's ADC pin samples the output voltage feedback value of the high-voltage module in real time. Using a PID control algorithm, the feedback value is compared with the target value, and the DAC output is dynamically adjusted to ensure that the output voltage is stable within the range of 2000V±0.5% and the stabilization time does not exceed 5 seconds.

[0057] Once the output voltage reaches the target value and stabilizes, the coprocessor sends a high-voltage ready status frame to the main processor via the isolated serial port. The status frame contains information such as the current actual voltage value, stabilization time, and temperature compensation coefficient.

[0058] The raw energy spectrum acquisition module employs an array-type CZT detector and a built-in dedicated ASIC readout circuit to convert gamma-ray energy into electrical pulse signals and digitize them into energy spectrum data. It interacts with the main control module via a network interface, specifically including:

[0059] The array-type CZT detector operates under the bias of the target voltage value. After the gamma rays are incident, they interact with the crystal to generate electron-hole pairs. The electrons drift to the anode under the action of the electric field, forming a charge pulse signal. The pulse amplitude is proportional to the energy of the gamma rays, and the pulse rise time is 50-100 ns.

[0060] The preamplifier of the ASIC readout circuit converts the charge pulse into a voltage pulse. The shaping circuit shapes the voltage pulse into a quasi-Gaussian waveform with a pulse width of about 2μs. The peak hold circuit captures the peak voltage of the pulse and compares it with the set energy threshold to remove noise pulses.

[0061] The output voltage of the peak hold circuit is converted into a digital value by an ADC with a resolution of 14 bits. The digital value is then converted into the corresponding channel address value (channels 0-4095). The energy spectrum accumulator inside the ASIC circuit increments the counter corresponding to the channel address value by 1 to form the energy spectrum data.

[0062] The ASIC circuit encapsulates the data frame according to the preset data frame format and performs integrity verification on the encapsulated data frame: it checks whether the frame header and frame tail match, calculates the CRC check value and compares it with the CRC field in the frame, checks whether the detection unit number is within the valid range, parses the data frame that passes the verification, and extracts the energy spectrum data block; the data frame that fails the verification is discarded, the frame loss count is recorded, and the error log is reported.

[0063] For example, the frame structure is as follows: frame header (2 bytes, 0xAA55), detector unit number (1 byte, N), timestamp (4 bytes, Unix timestamp), energy spectrum data block (4N bytes, energy spectrum data of N detector units, 4096 channels × 4 bytes per unit), CRC check (2 bytes, CRC-16), frame tail (1 byte, 0xCC), and the total frame length is 4N+10 bytes (N is the number of detector units).

[0064] The main processor accumulates the energy spectrum data by channel address to form a cumulative energy spectrum. The cumulative energy spectrum is stored in shared memory as a 4096-channel floating-point array. The energy spectrum data in the shared memory is updated every 1 second of accumulated acquisition time.

[0065] Example 2: The high-voltage power failure emergency protection module is used to monitor the external power supply status in real time and execute emergency power supply switching and soft voltage reduction shutdown procedures when the external power supply is interrupted. Specifically, it includes the following steps:

[0066] The voltage signal of the external power supply after step-down rectification is continuously sampled at a preset period, and the sampled value is compared with a preset threshold to determine whether the external power supply is stable.

[0067] When the sampled value is lower than the preset threshold for three consecutive times, it is determined that the external power supply is interrupted and a seamless switching operation is triggered. That is, the seamless switching from the external power supply to the lithium battery pack is completed immediately through the set target switch within a preset time period. At the same time, a power failure event interruption signal is sent, including the event type and timestamp.

[0068] Upon receiving a power outage event, the Qt interface thread is immediately invoked to pop up a modal dialog box displaying "External power supply has been disconnected. The system will automatically and safely shut down after a preset number of seconds." The remaining countdown time is displayed in real time in the interface status bar. The user can click the "Continue Data Acquisition" button to cancel this automatic shutdown and further monitor the battery voltage. If the voltage is higher than the preset voltage value, the system will continue to operate. If the voltage is lower than the preset voltage value, the system will continuously determine whether the countdown has reached zero and whether the battery voltage is lower than the preset danger threshold. If either condition is met, a forced safe shutdown process will be initiated.

[0069] Soft voltage reduction shutdown procedure:

[0070] Sa: Stop data acquisition and parsing: The main processor pauses the acquisition thread and the parsing thread, and writes the energy spectrum data and operation log in memory to the eMMC storage;

[0071] Sb: Perform soft bucking: The main processor sends a bucking command to the coprocessor through the isolated serial port; the coprocessor outputs a linearly decreasing control voltage to the high voltage module through the DAC pin with a preset fixed slope; the output of the high voltage module drops from the target voltage value to 0V at a constant speed; during the bucking process, the coprocessor reports the current voltage value in real time.

[0072] Sc: System Shutdown: After the high voltage drops to 0V, the main processor completes all data write-to-disk, safely unloads the file system, sends a shutdown command to the coprocessor, and the coprocessor cuts off the power to the main processor and peripherals, and the system is completely shut down.

[0073] Example 3: The parsing thread module is used to obtain the currently accumulated raw energy spectrum data from shared memory, and to perform background subtraction, peak shape correction, and overlapping peak decomposition on the acquired raw energy spectrum data. Specifically, it includes the following steps:

[0074] S1: Preprocess the raw energy spectrum data to obtain the net energy spectrum data. The preprocessing includes median smoothing and noise reduction and SNIP background subtraction.

[0075] Median smoothing noise reduction: The preset ZZSmooth function is called to perform median smoothing on the original energy spectrum data for 3 iterations. The smoothing window width is 5 channels to suppress statistical fluctuation noise.

[0076] SNIP background subtraction: The smoothed energy spectrum is converted to logarithmic space, and the SNIP iterative algorithm is executed with 10 iterations to accurately subtract the Compton scattering background and noise floor, thus obtaining the net energy spectrum data;

[0077] S2: Perform the second derivative peak-finding algorithm on the net energy spectrum data to identify candidate characteristic peaks, and calculate parameters such as peak energy, peak area, full width at half maximum (FWHM), and net count for each candidate characteristic peak.

[0078] S3: Compare the peak energy with the nuclide database to screen the list of candidate nuclides;

[0079] S4: For the selected candidate feature peaks, a pre-set KTW asymmetric Gaussian model is introduced for fine fitting;

[0080] Where E is energy, E0 is peak position, A is peak height, σ is standard deviation, and KTW is tailing coefficient; the optimal values ​​of peak position E0, peak height A, standard deviation σ, and tailing coefficient KTW are obtained by iteratively solving the Levenberg-Marquardt nonlinear least squares method.

[0081] S5: Compare the corrected candidate peak half-width at half-maximum (FWHM) with the calibrated single-energy peak half-width at half-maximum (FWHM):

[0082] If the half-width exceeds the preset calibration threshold, it is determined to be an overlapping peak, and the process proceeds to the intelligent decomposition and analysis step of overlapping peaks.

[0083] If the half-width at half-maximum does not exceed the preset calibration threshold, it is determined to be a non-overlapping peak, and the nuclide is directly identified based on the peak position and branching ratio.

[0084] The steps for intelligent disassembly and analysis of overlapping peaks are as follows:

[0085] T1: Peak Fitting and Subtraction: In overlapping peak groups, the characteristic peak with the highest energy is identified as the main peak. If the energies are similar, the peak with the largest net count (meaning the count generated purely by the characteristic γ-rays of the nuclide after subtracting the background) is selected as the main peak.

[0086] The main peak was accurately fitted using the KTW asymmetric Gaussian model to determine its peak position, peak area, and tailing coefficient.

[0087] The fitted main peak curve is subtracted from the original net energy spectrum channel by channel to generate the residual energy spectrum;

[0088] T2: Secondary peak finding of residual spectrum: For the residual energy spectrum after removing the main peak, the second derivative peak finding algorithm is re-executed to extract the secondary peak information that is covered by the main peak, including the secondary peak position, peak area, half width at half maximum and other parameters. If there are still abnormal peak widths in the residual spectrum, T1-T2 is repeated for multiple rounds of iterative stripping.

[0089] T3: Branch ratio verification and nuclide confirmation:

[0090] All the extracted peak energies are matched with the nuclidesData database to obtain a list of candidate nuclides for each peak.

[0091] For multiple characteristic peaks of the same candidate nuclide, the ratio of their peak areas is calculated as the measured branch ratio.

[0092] The measured branch ratio is compared with the standard branch ratio in the database to calculate the matching degree; when the matching degree is greater than the preset matching degree threshold (ThresholdBR=0.75), the presence of the nuclide is confirmed.

[0093] For the confirmed nuclide, the identification confidence level is calculated based on the branch ratio matching degree and peak area statistical error, and the nuclide identification result is output.

[0094] T4: Quantitative Analysis:

[0095] For a confirmed nuclide, the nuclide activity concentration is calculated based on parameters such as peak area, detector efficiency, measurement time, and decay correction factor. The formula for calculating the nuclide activity concentration is: A = NM / (ε × t × B × η), where NM is the net peak area, ε is the detection efficiency, t is the measurement time, B is the decay correction factor, and η is the branching ratio.

[0096] The energy spectrum UI display module uses a preset LTTB algorithm to downsample the energy spectrum data and achieves smooth graphical display and interaction, specifically including:

[0097] The raw energy spectrum data in the preset energy spectrum format is divided into M data buckets (M>0, such as M set to 512 by default), and each bucket contains k = preset energy spectrum format / M raw data points;

[0098] The first bucket selects the first data point, and the last bucket selects the last data point.

[0099] For each intermediate bucket, take the representative point of the previous bucket and the first point of the next bucket as the two vertices of a triangle, calculate the area of ​​the triangle formed by each point in the current bucket and these two vertices, and select the point with the largest area as the representative point of the bucket.

[0100] The representative points of all the buckets are combined in sequence to form the downsampled energy spectrum data;

[0101] Use the downsampled energy spectrum data as the data source for the QCPGraph object, and set the curve color, line width, and fill style;

[0102] Simultaneously, it listens to the user's two-finger zoom gesture on the touchscreen to obtain the zoom ratio and center point, recalculates the energy range of the energy spectrum display, and listens to the user's single-finger drag gesture to obtain the offset, translates the current display window, and refreshes the energy spectrum curve in real time.

[0103] And when the user clicks or long-presses the energy spectrum curve, the energy address corresponding to the clicked position is calculated, the nuclide information near that energy in the current analysis result is queried, and a label box pops up on the interface to display information such as nuclide name, activity, and confidence level;

[0104] In summary, a PID control algorithm is used to achieve uniform voltage rise and fall and precise voltage stabilization, avoiding damage to the CZT crystal from voltage surges. Real-time feedback of the operating status ensures the detector operates under optimal conditions, extending hardware lifespan. Furthermore, the array-type CZT detector paired with a dedicated ASIC readout circuit achieves high-precision, high-resolution conversion of gamma-ray energy. Multiple data verifications eliminate erroneous frames, and real-time accumulation and shared memory storage provide complete and accurate raw data for analysis. The system also features real-time power status monitoring and seamless switching, supporting user-intervention shutdown design. The soft-step-down shutdown process first writes data to disk before gradually reducing voltage and shutting down. This approach protects the detector hardware and prevents the loss of critical energy spectrum data and operational logs. After multiple preprocessing steps to remove noise, combined with a refined peak shape fitting and a multi-round iterative overlapping peak decomposition algorithm, it effectively identifies masked characteristic peaks, thus solving the problem that traditional methods cannot accurately analyze overlapping peaks. By combining the branch ratio verification of the nuclide database, the existence of nuclides is determined by the matching degree between measured values ​​and standard values. At the same time, the identification confidence and activity concentration are calculated, achieving the dual goals of accurate qualitative identification and accurate quantitative analysis of nuclides. Furthermore, it downsamples massive amounts of energy spectrum data, significantly reducing the amount of data while ensuring the complete preservation of energy spectrum features, and enabling smooth graphical display of energy spectrum curves.

[0105] The threshold is set for comparative analysis of results to determine whether they are good or bad. The value of the threshold is determined by a combination of large-scale model analysis of sample data and human experience. It can also be adjusted appropriately based on seasonal or common-sense influencing factors.

[0106] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A portable nuclide identification system based on an array-type CZT detector, characterized in that, It includes a detector bias high voltage establishment module, a raw energy spectrum acquisition module, a high voltage power failure emergency protection module, an analysis thread module, and an energy spectrum UI display module; The detector bias high voltage establishment module is used to receive the high voltage application command sent by the main processor through the coprocessor, and to perform the detector bias high voltage establishment operation in response to the high voltage application command, and output the high voltage ready status frame. The raw energy spectrum acquisition module uses an array-type CZT detector and a built-in dedicated ASIC readout circuit to convert gamma-ray energy into electrical pulse signals and digitize them into energy spectrum data. The high-voltage power failure emergency protection module is used to monitor the status of the external power supply in real time and execute emergency power supply switching and soft voltage reduction shutdown procedures when the external power supply is interrupted. The parsing thread module is used to obtain the currently accumulated raw energy spectrum data from shared memory, and to perform background subtraction, peak shape correction and overlapping peak decomposition on the collected raw energy spectrum data; The energy spectrum UI display module uses a preset LTTB algorithm to downsample the energy spectrum data and realizes graphical display and interaction; The background subtraction, peak shape correction, and overlapping peak decomposition process of the original energy spectrum data is as follows: The raw energy spectrum data is preprocessed to obtain the net energy spectrum data. The second derivative peak finding algorithm is executed on the net energy spectrum data to identify candidate characteristic peaks. For each candidate characteristic peak, the peak energy, peak area, half width at half maximum (FWHM), and net count are calculated. The peak energy is preliminarily compared with the nuclide database to screen the candidate nuclide list. For the selected candidate characteristic peaks, a preset KTW asymmetric Gaussian model is introduced for fine fitting. The corrected candidate peak half-width is compared with the calibrated single-energy peak half-width, and the output is transferred to the intelligent decomposition analysis step of overlapping peaks or to perform nuclide qualitative identification. The intelligent disassembly and analysis steps for overlapping peaks are also included as follows: T1: Peak Fitting and Subtraction: In the overlapping peak group, the characteristic peak with the highest energy is identified as the main peak. If the energies are similar, the peak with the largest net count is selected as the main peak. The KTW asymmetric Gaussian model is used to accurately fit the main peak, and its peak position, peak area and tailing coefficient are determined. The fitted main peak curve is subtracted from the original net energy spectrum one channel at a time to generate the residual energy spectrum. T2: Secondary peak finding of residual spectrum: For the residual energy spectrum after removing the main peak, the second derivative peak finding algorithm is re-executed to extract the secondary peak information that is covered by the main peak. If there are still abnormal peak widths in the residual spectrum, T1-T2 is repeated for multiple rounds of iterative stripping. T3: Branch ratio verification and nuclide confirmation: Match the energy of all extracted peak positions with the nuclide database to obtain a list of candidate nuclides for each peak position. For multiple characteristic peaks of the same candidate nuclide, calculate the peak area ratio as the measured branch ratio. Compare the measured branch ratio with the standard branch ratio in the database to calculate the matching degree. When the matching degree is greater than the preset matching degree threshold, the existence of the nuclide is confirmed. For the confirmed nuclide, the identification confidence level is calculated based on the branch ratio matching degree and peak area statistical error, and the nuclide identification result is output. T4: Quantitative analysis: For confirmed nuclides, calculate the nuclide activity concentration A based on its peak area, detector detection efficiency, measurement time, and decay correction factor.

2. The portable nuclide identification system based on an array-type CZT detector according to claim 1, characterized in that, The analysis process of the high-voltage ready state frame is as follows: Based on the high voltage application command sent by the main processor, the coprocessor outputs a linearly increasing control voltage to the high voltage module through the DAC pin. The high voltage module generates a corresponding high voltage output according to the control voltage and applies it to the array-type CZT detector. The voltage boosting process increases at a set slope at a constant speed. The output voltage feedback value of the high-voltage module is sampled in real time, and a PID control algorithm is used to compare the feedback value with the target value and dynamically adjust the DAC output. Once the target voltage value is reached and stabilized, the coprocessor sends a high-voltage ready status frame to the main processor.

3. The portable nuclide identification system based on an array-type CZT detector according to claim 1, characterized in that, The process of acquiring and analyzing the energy spectrum data is as follows: The array-type CZT detector operates under the bias of the target voltage value. After the gamma rays are incident, they interact with the crystal to generate electron-hole pairs. The electrons drift to the anode under the action of the electric field, forming a charge pulse signal. The preamplifier of the ASIC readout circuit converts the charge pulse into a voltage pulse, and the shaping circuit shapes the voltage pulse into a quasi-Gaussian waveform, which is compared with a set energy threshold to eliminate noise pulses. The output voltage of the peak hold circuit is converted into a digital value by the ADC, and the digital value is converted into the corresponding channel address value. The energy spectrum accumulator inside the ASIC circuit increments the counter corresponding to the channel address value by 1 to form energy spectrum data.

4. The portable nuclide identification system based on an array-type CZT detector according to claim 3, characterized in that, The ASIC circuit encapsulates the data frame according to the preset data frame format and performs integrity verification on the encapsulated data frame: check whether the frame header and frame tail match, calculate the CRC check value and compare it with the CRC field in the frame, check whether the detection unit number is within the valid range, parse the data frame that passes the verification, and extract the energy spectrum data block. Data frames that fail verification are discarded, the number of dropped frames is recorded, and an error log is reported. The main processor accumulates the energy spectrum data by channel address to form a cumulative energy spectrum, which is stored in shared memory as a 4096-channel floating-point array.

5. The portable nuclide identification system based on an array-type CZT detector according to claim 1, characterized in that, The analysis process of the high-voltage power failure emergency protection module is as follows: The voltage signal of the external power supply after step-down rectification is continuously sampled at a preset period. The sampled value is compared with a preset threshold to determine whether the external power supply is stable. When the sampled value is lower than the preset threshold for three consecutive times, the external power supply is determined to be interrupted, triggering a seamless switching operation. After the seamless switching operation, the user can click the "Continue Data Collection" button to cancel the automatic shutdown and further monitor the battery voltage. If the voltage is higher than the preset voltage value, the operation will continue. If the voltage is lower than the preset voltage value, the countdown will be continuously checked to see if it has reached zero and if the battery voltage is lower than the preset danger threshold. If either condition is met, a forced safety shutdown process will be initiated.

6. The portable nuclide identification system based on an array-type CZT detector according to claim 1, characterized in that, The soft voltage reduction shutdown procedure is as follows: Sa: Stop data acquisition, parsing, and writing; Sb: Execute soft voltage reduction; Sc: System shutdown.

7. The portable nuclide identification system based on an array-type CZT detector according to claim 1, characterized in that, The analysis process of the energy spectrum UI display module is as follows: The raw energy spectrum data in the preset energy spectrum format is divided into M data buckets (M>0). Each bucket contains k = preset energy spectrum format / M raw data points. The first data point is selected for the first bucket, and the last data point is selected for the last bucket. For each intermediate bucket, the representative point of the previous bucket and the first point of the next bucket are used as the two vertices of a triangle. The area of ​​the triangle formed by each point in the current bucket and these two vertices is calculated, and the point with the largest area is selected as the representative point of the bucket. Combine the representative points of all buckets in order to form downsampled energy spectrum data. Use the downsampled energy spectrum data as the data source of the QCPGraph object and set the curve color, line width, and fill style. The system obtains the user's zoom level and center point on the touchscreen, recalculates the energy range of the energy spectrum display, listens for the user's single-finger drag gesture, obtains the offset, translates the current display window, and refreshes the energy spectrum curve in real time. When a user clicks or long-presses the energy spectrum curve, the system calculates the energy address corresponding to the clicked location, queries the nuclide information near that energy in the current analysis results, and displays the nuclide name, activity, and confidence level in a pop-up label box on the interface.

Images