A method, system for SiPM testing and calibration for particle detection applications

Abstract

The present application relates to the technical field of particle detection, in particular to a SiPM testing and calibration method and system for particle detection application, comprising the following steps: extracting a slope quadratic difference set to determine a current change inflection point, calculating a local slope change proportion set to obtain a breakdown turning voltage point, analyzing output pulse trigger probability, combining photon number extraction probability growth interval, constructing refined pulse extraction detection efficiency reference ratio, screening high-density trigger data, calculating spatial deviation extreme value, and adjusting system bias compensation voltage. In the present application, the breakdown turning voltage point is located, thereby avoiding measurement interference caused by basic background noise, a trigger probability sequence is established by counting the number of attenuations, and a detection efficiency reference ratio is extracted by combining a preset maximum likelihood estimation model, a channel deviation extreme value is calculated to directly correct parameter fluctuation, and an array global dynamic balance is established to maintain long-term stability of the overall detection working state.

Description

Technical Field

[0001] This invention relates to the field of particle detection technology, and in particular to a SiPM testing and calibration method and system for particle detection applications. Background Technology

[0002] The field of particle and nuclear detection technology involves the observation, measurement, and analysis of high-energy particles, radiative particles, and their interactions. It is widely applied in high-energy physics experiments, nuclear physics experiments, cosmic ray observation, medical imaging, and radiation monitoring. This field typically uses particle detectors as the core device, acquiring and recording electrical or optical signals generated by the interaction of particles with matter through detection materials and readout electronics systems. Common detector types include scintillator detectors, gas detectors, semiconductor detectors, and photomultiplier detectors. In modern particle detection systems, silicon photomultipliers, as solid-state photon detection devices, are widely used in scintillator readout systems and particle detection arrays due to their high gain, small size, and responsiveness to weak light signals. In practical applications, to ensure the consistency and stability of the detector array's output signal, systematic testing and calibration of the detector are required during the setup and use of the experimental setup. This process typically involves setting the bias voltage, measuring the dark count rate, measuring the gain, measuring the single-photon response, and measuring parameter changes under temperature conditions. The detector performance parameters are acquired and processed through electronic readout circuits and data recording equipment.

[0003] The traditional SiPM testing and calibration method used for particle detection applications refers to the performance testing and parameter calibration process of silicon photomultipliers before or during use in particle detection systems. It mainly measures parameters such as device gain, dark count rate, photon detection efficiency, and breakdown voltage. Traditional methods typically involve mounting the SiPM device on a test circuit board, applying an adjustable bias voltage to the SiPM via a stable power supply, and using a pulsed light source or LED to generate weak light pulses that illuminate the SiPM's photosensitive surface. Simultaneously, a preamplifier circuit amplifies the SiPM's output current signal, and an oscilloscope or multi-channel charge integrator records the output pulse waveform amplitude and charge distribution. The position of the single-photon response peak is determined by statistically analyzing the pulse amplitude distribution to calculate the device gain. The number of output pulses per unit time is recorded under no-light conditions to obtain the dark count rate. During breakdown voltage measurement, the bias voltage is typically adjusted gradually, and the output current change curve is recorded to determine the voltage point where the current rises rapidly. In array-type SiPM applications, it is also necessary to connect the test circuit channel by channel and repeat the illumination measurement steps to obtain the gain and response parameters of each channel. The bias voltage is then adjusted channel by channel based on the measured parameters to complete the calibration.

[0004] Traditional calibration methods rely on manually building peripheral circuit boards and adjusting steady-state power supplies. They use external testing equipment to perform static statistics on pulse waveforms and charge distribution. When measuring breakdown voltage, they rely solely on monotonic step voltage adjustment and observing the absolute change amplitude of current. This makes the process susceptible to background noise interference, leading to deviations in feature point judgment. When dealing with array-type probe architectures, they use repeated wiring and illumination testing to obtain parameters. This cumbersome process lengthens the calibration cycle. Furthermore, they lack spatial deviation correlation analysis for abnormal fluctuations in time intervals under multi-channel concurrent conditions, making it difficult to maintain the consistency of bias states of each channel within the device array. Summary of the Invention

[0005] The purpose of this invention is to overcome the shortcomings of the existing technology and to propose a SiPM testing and calibration method and system for particle detection applications.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: a SiPM testing and calibration method for particle detection applications, comprising the following steps: S1: Obtain the steady-state dark current data of the silicon photomultiplier tube under the bias voltage sequence, calculate the current change slope sequence, extract the slope second difference set, determine the position of the value in the slope second difference set that exceeds the differential cutoff threshold, and extract the current change inflection point set. S2: Based on the set of current change inflection points, filter the local scanning voltage range, obtain the secondary dark current data in the local scanning voltage range and calculate the local slope change ratio set, analyze the numerical distribution in the local slope change ratio set, and determine the breakdown inflection voltage point. S3: Based on the breakdown transition voltage point, obtain the output pulse signal of the silicon photomultiplier tube under the optical pulse excitation sequence, calculate and obtain the trigger probability sequence, compare the trigger probability sequence with the number of optical pulse photons in the corresponding optical pulse excitation sequence, and extract the probability growth interval. S4: Based on the probability growth range and fine-tuning photon step size, calculate the local triggering probability sequence, input the ratio parameter of the number of photons in the optical pulse to the local triggering probability sequence into the maximum likelihood estimation model for calculation, and extract the detection efficiency benchmark ratio. S5: Based on the detection efficiency benchmark ratio, obtain the pulse time interval output by the silicon photomultiplier tube in multiple detection channels, determine the value of the time interval less than the preset time threshold as high-density trigger data, calculate the spatial deviation extreme value, and adjust the bias compensation voltage based on the spatial deviation extreme value.

[0007] As a further embodiment of the present invention, the set of inflection points of current change includes critical abrupt change coordinates, response transition index and differential transition node; the breakdown transition voltage point includes conduction attenuation threshold, steady-state level parameter and breakdown edge potential; the probability growth interval includes response ignition lower bound, sensitivity ramp-up band and linear response upper bound; the detection efficiency benchmark ratio includes quantum conversion coefficient, absolute calibration constant and photoelectric gain factor; and the bias compensation voltage includes dark noise reduction bias voltage, array equalization difference and leakage current correction amount.

[0008] As a further aspect of the present invention, the step of obtaining the set of inflection points of current change specifically includes: S101: Collect steady-state dark current data of silicon photomultiplier tubes within the bias voltage sequence, extract voltage difference variables for adjacent nodes within the bias voltage sequence, extract current difference variables at corresponding positions within the steady-state dark current data, perform algebraic division operation on the current difference variables and voltage difference variables to obtain node slope variables, and perform array concatenation on all node slope variables according to the scanning order to obtain the current change slope sequence; S102: Perform offset alignment on adjacent numerical terms within the current change slope sequence, extract the subsequent index value and subtract the preceding index value to obtain a first difference variable, perform adjacent term subtraction operation again on all first difference variables to obtain a second difference variable, arrange all second difference variables according to the original numerical index to generate a slope second difference set. S103: Extract the preset differential cutoff threshold, compare the distributed values ​​in the slope quadratic difference set with the differential cutoff threshold, filter the sequence target index numbers in the slope quadratic difference set whose values ​​exceed the differential cutoff threshold, extract the voltage parameters in the bias voltage sequence that match the sequence target index numbers, and obtain the current change inflection point set.

[0009] As a further aspect of the present invention, the differential cutoff threshold is obtained by extracting multiple quadratic difference variables located at the initial arrangement position from the slope quadratic difference set to construct a baseline sequence; calculating the sequence mean parameter and the baseline standard deviation parameter based on all quadratic difference variables in the baseline sequence; extracting the fluctuation confidence coefficient, which is set by collecting the distribution boundary multiple of the static dark level of the silicon photomultiplier tube in the unshielded state; performing a multiplication operation on the baseline standard deviation parameter and the fluctuation confidence coefficient to obtain the tolerance offset; and performing an addition operation on the sequence mean parameter and the tolerance offset to generate the differential cutoff threshold.

[0010] As a further aspect of the present invention, the step of obtaining the breakdown transition voltage point specifically includes: S201: Call the set of current change inflection points and the current change slope sequence, extract voltage parameters for the nodes distributed in the set of current change inflection points, compare the value of each node at the corresponding position in the current change slope sequence, extract the node with the maximum value of the corresponding sequence value as a candidate target, calculate the voltage difference between adjacent nodes of the candidate target, select the node with the minimum voltage difference as the center coordinate, extend the preset voltage step size along both sides of the center coordinate, and establish the neighborhood offset scanning interval; S202: Collect the secondary dark current data corresponding to each voltage node in the neighborhood bias scanning interval, perform difference calculation on adjacent voltage nodes in the neighborhood bias scanning interval to obtain the local voltage increment, perform difference operation on adjacent values ​​in the secondary dark current data to obtain the local current increment, perform division operation on the local current increment and the local voltage increment to obtain the secondary slope parameter, extract the adjacent secondary slope parameters and perform ratio calculation to generate a local slope ratio sequence; S203: Arrange the distributed values ​​within the local slope ratio sequence to form a contrast gradient, extract the minimum value term within the contrast gradient, map the minimum value term to the corresponding voltage parameter node within the neighborhood bias scan interval, extract the digital record pointed to by the corresponding voltage parameter node, and obtain the breakdown inflection voltage point.

[0011] As a further aspect of the present invention, the step of obtaining the probability growth interval specifically includes: S301: Based on the breakdown transition voltage point, detect the initial pulse signal fed back during the period when the optical pulse excitation sequence acts on the silicon photomultiplier tube, extract the pulse amplitude feature data carried in the initial pulse signal, compare the pulse amplitude feature data with the preset attenuation judgment threshold, filter out the background waveforms below the preset attenuation judgment threshold, retain the nodes exceeding the preset attenuation judgment threshold as attenuation signals, count the cumulative value of attenuation signals within the set trigger period, and generate an effective attenuation pulse set sequence. S302: Call the effective attenuation pulse quantity set sequence and the unit trigger count parameter, perform traversal reading for each value in the effective attenuation pulse quantity set sequence, perform algebraic division operation on each value and the unit trigger count parameter to obtain the single-point trigger ratio, and perform array concatenation on all single-point trigger ratios in the original scan time order to establish a trigger probability sequence; S303: Obtain the number of photons in the light pulse corresponding to the light pulse excitation sequence; perform differential operation on adjacent nodes in the trigger probability sequence to obtain the probability change increment; perform a comparison operation between the probability change increment and a preset judgment benchmark value; select a set of coherent nodes whose probability change increment exceeds the preset judgment benchmark value; map the two ends of the coherent node set to the light pulse photon number distribution dimension; extract the interval boundary digital segment restricted by the start and end coordinates; and obtain the probability growth interval.

[0012] As a further aspect of the present invention, the process of setting the attenuation judgment threshold is specifically as follows: collecting a set of dark noise signals in a dark environment, extracting the maximum amplitude distribution value in the set of dark noise signals, obtaining a single photon amplitude reference value under calibration, adding the maximum amplitude distribution value and the single photon amplitude reference value to obtain the amplitude reference sum, performing a division operation of the amplitude reference sum by two to obtain the center level parameter, and establishing the center level parameter as the preset attenuation judgment threshold; The specific method for setting the preset judgment benchmark value is as follows: extract the peak value of background probability fluctuation under the no-light triggering environment, obtain the standard triggering slope constant of the silicon photomultiplier tube under the fixed working state, perform an algebraic summation operation on the standard triggering slope constant and the peak value of background probability fluctuation, and set the numerical result generated by the algebraic summation operation as the preset judgment benchmark value.

[0013] As a further aspect of the present invention, the step of obtaining the detection efficiency benchmark ratio specifically includes: S401: Call the probability growth interval and the locally extracted fine-tuned photon step size, extract the start and end boundary values ​​of the probability growth interval, divide the start and end boundary values ​​into equal distances according to the fine-tuned photon step size to obtain the number of photons associated with each discrete node, and perform a sequence combination operation on all photons according to the node position to construct a refined optical pulse signal; S402: Based on the refined optical pulse signal, obtain the associated detection response waveform, count the total accumulated attenuation of the detection response waveform within a unit number of triggers, perform algebraic division between the total accumulated attenuation and the unit number of triggers to obtain the single-point trigger variable, and perform corresponding mapping splicing on all single-point trigger variables according to the index of the refined optical pulse signal to generate a local trigger probability sequence. S403: Based on the local trigger probability sequence and the number of photons in the light pulse, perform sequence alignment of each node of the local trigger probability sequence with the number of photons in the light pulse, extract the aligned values ​​and divide to obtain the probability photon ratio set, input the probability photon ratio set into a preset maximum likelihood estimation model, call the maximum likelihood estimation model to perform iterative fitting on the probability photon ratio set to obtain the convergent extreme value, and extract the detection efficiency benchmark ratio.

[0014] As a further aspect of the present invention, the step of obtaining the bias compensation voltage specifically includes: S501: Based on the detection efficiency benchmark ratio, collect the timestamps corresponding to the output pulses of multiple detection channels, subtract adjacent pulse timestamps to obtain the single-point time difference, combine the single-point time difference values ​​according to channel category to establish a time interval change sequence, call the preset time truncation threshold, compare the distributed values ​​inside the time interval change sequence with the preset time truncation threshold, filter out the associated value items below the preset time truncation threshold, and obtain high-density trigger data; S502: Calculate the accumulated amount of clustered pulses in the high-density trigger data, collect the total amount of pulses sampled by the detection channel, divide the accumulated amount of pulses by the total amount of sampled pulses to obtain the channel ratio variable, and perform sequence merging and assembly of multiple channel ratio variables to generate the high-density trigger ratio. S503: Calculate the average reference term of each value within the high-density trigger ratio, perform subtraction operation on each value with the average reference term to extract the deviation from the absolute parameter, retrieve the maximum term within all deviations from the absolute parameter to extract the spatial deviation extreme value, perform addition operation on the device reference bias voltage and the spatial deviation extreme value to obtain the array balance correction amount, and establish the bias compensation voltage.

[0015] A SiPM testing and calibration system for particle detection applications, wherein the SiPM testing and calibration system for particle detection applications is used to implement the aforementioned SiPM testing and calibration method for particle detection applications, the system comprising: The dark current feature analysis module acquires the steady-state dark current data of the silicon photomultiplier tube under the bias voltage sequence, calculates the current change slope sequence, extracts the second difference set of the slope, determines the position of the value in the second difference set that exceeds the differential cutoff threshold, and extracts the set of inflection points of current change. The breakdown voltage analysis module, based on the set of current change inflection points, filters the local scanning voltage range, obtains the secondary dark current data within the local scanning voltage range, calculates the local slope change ratio set, analyzes the numerical distribution within the local slope change ratio set, and determines the breakdown inflection voltage point. The trigger probability analysis module, based on the breakdown transition voltage point, obtains the output pulse signal corresponding to the silicon photomultiplier tube under the optical pulse excitation sequence, calculates and obtains the trigger probability sequence, compares the trigger probability sequence with the number of optical pulse photons in the corresponding optical pulse excitation sequence, and extracts the probability growth interval. The detection efficiency benchmark extraction module calculates the local triggering probability sequence based on the probability growth interval and the fine-tuned photon step size. It inputs the ratio parameter of the number of photons in the light pulse to the local triggering probability sequence into the maximum likelihood estimation model for calculation and extracts the detection efficiency benchmark ratio. The spatial deviation compensation module, based on the detection efficiency benchmark ratio, obtains the pulse time interval output by the silicon photomultiplier tube in multiple detection channels, determines the value of the time interval less than the preset time threshold as high-density trigger data, calculates the spatial deviation extreme value, and adjusts the bias compensation voltage based on the spatial deviation extreme value.

[0016] Compared with the prior art, the advantages and positive effects of the present invention are as follows: In this invention, a second-order difference set of slopes is extracted based on steady-state dark current and inflection points are selected. A local scanning voltage range is constructed to analyze the second-order dark current data, accurately locating the breakdown inflection voltage point to avoid basic noise interference. At the same time, a trigger probability sequence is established based on the effective attenuation of the output pulse signal. The probability growth range is extracted by combining the photon number and a refined optical pulse signal is generated. The detection efficiency benchmark ratio is iteratively extracted, the multi-channel pulse time interval is extracted and high-density trigger data is selected. The trigger ratio and spatial deviation extreme value are calculated based on the pulse ratio. The deviation fluctuation is directly corrected based on the bias compensation voltage to establish a dynamic balance of global array parameters to maintain stable operation. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a schematic diagram of the workflow of the present invention; Figure 2 This is a detailed flowchart of S1 of the present invention; Figure 3 This is a detailed flowchart of the S2 process of the present invention; Figure 4 This is a detailed flowchart of the S3 process of the present invention; Figure 5 This is a detailed flowchart of the S4 process of the present invention; Figure 6 This is a detailed flowchart of S5 of the present invention; Figure 7 This is a system flowchart of the present invention. Detailed Implementation

[0019] The technical solution of the present invention will now be described with reference to the accompanying drawings.

[0020] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0021] Please see Figure 1 This invention provides a technical solution: a SiPM testing and calibration method for particle detection applications, comprising the following steps: S1: Obtain the steady-state dark current data of the silicon photomultiplier tube under the bias voltage sequence. Based on the bias voltage sequence and the steady-state dark current data, perform first-order differential calculation to obtain the current change slope sequence. Input the current change slope sequence into Newton's forward difference formula to calculate and extract the slope second difference set. Determine the position of the value in the slope second difference set that exceeds the difference cutoff threshold to extract the current change inflection point set. S2: Based on the comparison of voltage parameters within the set of current change inflection points, select parameter nodes in the set of current change inflection points that correspond to the peak value of the current change slope sequence and the minimum value of the corresponding voltage parameter change amplitude to construct a local scanning voltage interval. Obtain secondary dark current data within the local scanning voltage interval and calculate the local slope change ratio set. Analyze the numerical distribution within the local slope change ratio set to determine the breakdown inflection voltage point. S3: Based on the breakdown transition voltage point, obtain the output pulse signal of the silicon photomultiplier tube under the optical pulse excitation sequence, calculate the number of effective attenuated pulses of the output pulse signal within a unit number of triggers, calculate the trigger probability sequence by comparing the number of effective attenuated pulses with the number of triggers per unit number of triggers, and compare the trigger probability sequence with the probability growth range of the number of optical pulse photons in the corresponding optical pulse excitation sequence. S4: Based on the probability growth range and fine-tuning photon step size, a refined optical pulse signal is constructed. The refined optical pulse signal and the number of unit triggers are analyzed to calculate the local trigger probability sequence. The ratio parameter of the number of optical pulse photons to the local trigger probability sequence is input into the maximum likelihood estimation model to calculate and extract the detection efficiency benchmark ratio. S5: Based on the detection efficiency benchmark ratio, obtain the pulse time interval of the silicon photomultiplier tube output in multiple detection channels and calculate the time interval change sequence. Values ​​in the time interval change sequence that are less than a preset time threshold are identified as high-density trigger data. Calculate the high-density trigger ratio based on the number of pulses contained in the high-density trigger data and the total number of sampled pulses. Compare the high-density trigger ratios corresponding to multiple detection channels to calculate the spatial deviation extreme value. Adjust the bias compensation voltage based on the spatial deviation extreme value.

[0022] The set of inflection points for current change includes critical abrupt change coordinates, response transition exponent, and differential transition nodes; the breakdown transition voltage point includes conduction attenuation threshold, steady-state level parameter, and breakdown edge potential; the probability growth range includes response ignition lower bound, sensitivity ramp-up band, and linear response upper bound; the detection efficiency benchmark ratio includes quantum conversion coefficient, absolute calibration constant, and photoelectric gain factor; and the bias compensation voltage includes dark noise reduction bias voltage, array equalization difference, and leakage current correction.

[0023] Please see Figure 2 The specific steps for obtaining the set of inflection points of current change are as follows: S101: Collect steady-state dark current data of silicon photomultiplier tubes within the bias voltage sequence, extract voltage difference variables for adjacent nodes within the bias voltage sequence, extract current difference variables at corresponding positions within the steady-state dark current data, perform algebraic division operation on the current difference variables and voltage difference variables to obtain node slope variables, and perform array concatenation on all node slope variables according to the scanning order to obtain the current change slope sequence; The test signal of the silicon photomultiplier tube, continuously varying within a bias voltage sequence, is extracted. A digital source meter continuously applies voltage excitation in fixed voltage steps of 0.1 volts, with a sampling frequency set to 10 times per second, for 100 consecutive sampling cycles. The acquired analog signal is converted into a digital sequence using a 16-bit analog-to-digital converter to obtain steady-state dark current data. For adjacent nodes within the bias voltage sequence, the voltage difference variable is extracted. The bias voltage value of the second node is subtracted from the bias voltage value of the first node. Setting the bias voltage of the first node to 50.1 volts and the bias voltage of the second node to 50.2 volts, the voltage difference variable = 50.2 - 50.1, yielding a voltage difference variable of 0.1 volts. Extract the current difference variable at the corresponding location within the steady-state dark current data. Subtract the dark current value of the first node from the dark current value of the second node. Set the dark current of the first node to 1.2 μA and the dark current of the second node to 1.5 μA. The current difference variable = 1.5 - 1.2, resulting in a current difference variable of 0.3 μA. Perform algebraic division on the current difference variable and the voltage difference variable. The node slope variable = current difference variable / voltage difference variable. The node slope variable = 0.3 / 0.1, resulting in a node slope variable of 3.0 μA per volt. Perform array concatenation on all node slope variables according to the scanning order. Store the slope variables of the first node to the 999th node in a one-dimensional continuous memory space according to the ascending voltage acquisition time.

[0024] Table 1 Initial Voltage and Current Data of Test Nodes

[0025] Table 1 shows the collected data of bias voltage and dark current and the corresponding slope calculation results.

[0026] S102: Perform offset alignment on adjacent numerical terms within the current change slope sequence, extract the subsequent index value and subtract the preceding index value to obtain a first difference variable, perform adjacent term subtraction operation again on all first difference variables to obtain a second difference variable, arrange all second difference variables according to the original numerical index to generate a slope second difference set. For adjacent numerical terms within the generated current change slope sequence, offset alignment is performed, placing the numerical term with index 2 and the numerical term with index 1 into the same data processing window of length 2. The first difference variable is obtained by subtracting the preceding index value from the subsequent index value. The preceding index value corresponds to a slope variable of 3.0 μA / V for the first node, and the subsequent index value corresponds to a slope variable of 4.0 μA / V for the second node. The first difference variable = 4.0 - 3.0, resulting in a first first difference variable of 1.0 μA / V squared. Similarly, for the slope variable of 4.0 μA / V for the second node and 5.0 μA / V for the third node, the first difference variable = 5.0 - 4.0, resulting in a second first difference variable of 1.0 μA / V squared. For all first-order difference variables, perform adjacent term subtraction again, reading the second first-order difference variable and subtracting the first. Set the first first-order difference variable to 1.0 microamps per volt squared and the second to 1.5 microamps per volt squared. The second-order difference variable is then calculated as 1.5 - 1.0, resulting in a second-order difference variable of 0.5 microamps per volt cubed. Arrange all second-order difference variables according to their original numerical indices, loading them into the set variable according to the original voltage node numbers from 1 to 997.

[0027] S103: Extract the preset differential cutoff threshold, compare the distributed values ​​in the slope quadratic difference set with the differential cutoff threshold, filter the target index numbers of the sequence corresponding to the values ​​in the slope quadratic difference set that exceed the differential cutoff threshold, extract the voltage parameters in the bias voltage sequence that match the target index numbers of the sequence, and obtain the set of inflection points of current change. The differential cutoff threshold is obtained as follows: Multiple quadratic difference variables located at the initial position in the slope quadratic difference set are extracted to construct a baseline sequence; the sequence mean parameter and the baseline standard deviation parameter are calculated based on all quadratic difference variables within the baseline sequence; the fluctuation confidence coefficient is extracted, and the fluctuation confidence coefficient is set by collecting the distribution boundary multiple of the static dark level of the silicon photomultiplier tube in an unshielded state; the baseline standard deviation parameter and the fluctuation confidence coefficient are multiplied to obtain the tolerance offset; the sequence mean parameter and the tolerance offset are added to generate the differential cutoff threshold.

[0028] Ten quadratic difference variables, ranked from 1st to 10th in the slope quadratic difference set, were extracted to construct a baseline sequence. These ten variables were summed and divided by 10; the sequence mean parameter was calculated as the sum of all quadratic difference variables divided by 10. Setting the sum of the ten variables to 2.0 μA / V cubic, the sequence mean parameter was calculated as 0.2 μA / V cubic. The square root of the sum of the squares of the differences between each quadratic difference variable and the sequence mean parameter, divided by 9, was used as the baseline standard deviation parameter, set to 0.05 μA / V cubic. Static dark level waveforms of a silicon photomultiplier tube under unshielded conditions were continuously observed for 100 seconds. The ratio of the highest dark level amplitude to the average dark level amplitude was obtained and set to 5.0, which was then used as the fluctuation confidence coefficient. Multiply the baseline standard deviation parameter and the fluctuation confidence coefficient, and set the tolerance offset to 0.05 × 5.0, resulting in a tolerance offset of 0.25 μA / V cubic. Add the sequence mean parameter and the tolerance offset, and set the difference cutoff threshold to 0.2 + 0.25, resulting in a difference cutoff threshold of 0.45 μA / V cubic. Compare each value in the slope quadratic difference set with the difference cutoff threshold, and retrieve quadratic difference variables with values ​​greater than 0.45 μA / V cubic, extracting the index of the corresponding node in the set. Set the quadratic difference variable corresponding to index 45 to be greater than 0.45 μA / V cubic, and extract 45 as the target index of the sequence. Extract the voltage parameter corresponding to the target index of sequence 45 in the bias voltage sequence, obtaining a voltage value of 54.5 volts, and store 54.5 volts in an array.

[0029] Please see Figure 3 The specific steps for obtaining the breakdown inflection voltage point are as follows: S201: Call the set of inflection points of current change and the sequence of slope of current change, extract voltage parameters for the nodes distributed in the set of inflection points of current change, compare the value of each node at the corresponding position in the sequence of slope of current change, extract the node with the maximum value of the corresponding sequence value as a candidate target, calculate the voltage difference between adjacent nodes of the candidate target, select the node with the minimum value of voltage difference as the center coordinate, extend the preset voltage step size along both sides of the center coordinate, and establish the neighborhood bias scanning interval. The process of establishing a neighborhood offset scanning interval by extending a preset voltage step size along both sides of the center coordinate is as follows: Extract the voltage difference between the preceding and following adjacent nodes corresponding to the center coordinate in the offset scan; compare the values ​​within the voltage difference between the preceding and following adjacent nodes to obtain a maximum value parameter; multiply the maximum value parameter with a preset expansion factor to obtain a preset voltage step size; the preset expansion factor is set by extracting the micro-pixel attenuation recovery time constant of the silicon photomultiplier tube and combining it with the detection channel gain tolerance parameter through proportional weighting calculation; subtract the preset voltage step size from the center voltage value corresponding to the center coordinate to obtain the lower boundary value of the scan; add the preset voltage step size to the center voltage value corresponding to the center coordinate to obtain the upper boundary value of the scan; integrate the lower boundary value and the upper boundary value of the scan to form a boundary span interval, which is set as the neighborhood offset scanning interval; For nodes distributed within the current change inflection point set, a voltage parameter of 54.5 volts is extracted. This voltage is matched to the value at position 45 in the current change slope sequence. The slope values ​​at positions 40 to 50 are examined, and the node with the highest value is selected as a candidate target. The slope of node 48 is set to 15.0 microamps per volt, and it is also selected as a candidate target. The voltage difference between nodes 47 and 49, adjacent to candidate target 48, is calculated. The node with the smallest voltage difference is assigned to the center coordinate, and its corresponding center voltage value is set to 54.8 volts. The voltage difference between the preceding and following adjacent nodes (0.1 volts) corresponding to the center coordinate in the offset scan is extracted, along with the voltage difference between the following and preceding adjacent nodes. The 0.15 volt value is then extracted as the maximum parameter. The micropixel attenuation recovery time constant, calibrated from the silicon photomultiplier tube's factory parameters, is extracted to be 50.0 nanoseconds. The probe channel gain tolerance parameter is extracted to be 0.02. The preset expansion factor is calculated as: micropixel attenuation recovery time constant × probe channel gain tolerance parameter = 50.0 × 0.02, resulting in a preset expansion factor of 1.0. The maximum value parameter of 0.15 volts is multiplied by the preset expansion factor of 1.0, resulting in a preset voltage step size of 0.15 × 1.0, also resulting in a preset voltage step size of 0.15 volts. The center voltage value of 54.8 volts is subtracted from the preset voltage step size of 0.15 volts, resulting in a lower scan boundary value of 54.65 volts (54.8 - 0.15). The center voltage value is then added to the preset voltage step size, resulting in an upper scan boundary value of 54.95 volts (54.8 + 0.15). The 54.65 volt and 54.95 volt ranges are combined to form the boundary span.

[0030] S202: Collect the secondary dark current data corresponding to each voltage node in the neighborhood bias scan interval, perform difference calculation on adjacent voltage nodes in the neighborhood bias scan interval to obtain the local voltage increment, perform difference operation on adjacent values ​​in the secondary dark current data to obtain the local current increment, perform division operation on the local current increment and local voltage increment to obtain the secondary slope parameter, extract the adjacent secondary slope parameters and perform ratio calculation to generate the local slope ratio sequence; Secondary dark current data for each voltage node within the 54.65 volt to 54.95 volt range was collected, with the data acquisition frequency increased to 100 times per second. Difference calculations were performed on adjacent voltage nodes within the neighborhood bias scan interval. With the second-level node voltage set at 54.68 volts and the first-level node voltage at 54.65 volts, the local voltage increment was calculated as 54.68 - 54.65, yielding a local voltage increment of 0.03 volts. Differential operations were performed on adjacent values ​​within the secondary dark current data. With the second-level dark current set at 8.5 microamps and the first-level dark current at 8.1 microamps, the local current increment was calculated as 8.5 - 8.1, yielding a local current increment of 0.4 microamps. A division operation was then performed between the local current increment and the local voltage increment, with the secondary slope parameter calculated as 0.4 / 0.03, resulting in a secondary slope parameter of 13.33 microamps per volt. The adjacent second-order slope parameters are extracted and their ratios are calculated. The subsequent second-order slope parameter is set to 15.0 μA / V, and the preceding second-order slope parameter is set to 13.33 μA / V. The local slope ratio is calculated as 15.0 / 13.33, resulting in a local slope ratio of 1.125. The local slope ratios of all node pairs are then sorted by voltage and loaded into an array.

[0031] S203: Arrange the distributed values ​​within the local slope ratio sequence to form a contrast gradient, extract the minimum value term within the contrast gradient, map the minimum value term to the corresponding voltage parameter node within the neighborhood bias scan interval, extract the digital record pointed to by the corresponding voltage parameter node, and obtain the breakdown inflection voltage point. A one-dimensional contrast gradient array is constructed by arranging all the ratio values ​​distributed within the local slope ratio sequence. A traversal algorithm is used to compare the magnitudes of all elements in the contrast gradient array, extracting the data item with the smallest value as the minimum value. After traversal comparison, the ratio value of 1.01 corresponding to the 8th position is identified as the minimum value. This minimum value of 1.01 is mapped to the 8th voltage parameter node within the neighborhood bias scan interval. The numerical record pointed to by this node is extracted, and the voltage test value corresponding to this node is obtained as 54.74 volts, which is then output to a register.

[0032] Please see Figure 4 The specific steps for obtaining the probability growth interval are as follows: S301: Based on the breakdown transition voltage point, detect the initial pulse signal fed back during the period when the optical pulse excitation sequence is applied to the silicon photomultiplier tube, extract the pulse amplitude feature data carried in the initial pulse signal, compare the pulse amplitude feature data with the preset attenuation judgment threshold, filter out the background waveforms below the preset attenuation judgment threshold, retain the nodes exceeding the preset attenuation judgment threshold as attenuation signals, count the cumulative value of attenuation signals within the set trigger period, and generate a set sequence of effective attenuation pulse numbers; The specific process of setting the attenuation judgment threshold is as follows: collect a set of dark noise signals in a dark environment, extract the maximum amplitude distribution value in the set of dark noise signals, obtain the single photon amplitude reference value under the calibration state, add the maximum amplitude distribution value and the single photon amplitude reference value to obtain the amplitude reference sum, perform a division operation of the amplitude reference sum by two to obtain the center level parameter, and establish the center level parameter as the preset attenuation judgment threshold. The initial pulse signal fed back during the excitation sequence of a 405 nm optical pulse applied to a silicon photomultiplier tube was detected. The voltage amplitude corresponding to the peak value of the waveform within the initial pulse signal was extracted as pulse amplitude feature data. A set of dark noise signals was collected over 1000 consecutive trigger cycles in a dark environment, and the maximum amplitude distribution value in the dark noise signal set was extracted to be 5.0 mV. The single-photon amplitude reference value under calibration conditions was obtained as 25.0 mV. The maximum amplitude distribution value was added to the single-photon amplitude reference value, resulting in a total amplitude reference sum of 5.0 + 25.0, which is 30.0 mV. This sum was then divided by 2, resulting in a centering level parameter of 30.0 / 2, which is 15.0 mV. This 15.0 mV was established as the preset attenuation threshold. The pulse amplitude feature data was compared with the preset attenuation threshold of 15.0 mV, and background waveforms below 15.0 mV were filtered out. Nodes exceeding 15.0 mV are retained. A node with a pulse amplitude set to 18.0 mV, exceeding 15.0 mV, is retained. The accumulated attenuation signal value within a set trigger period of 100 microseconds is counted, yielding 45 effective attenuation pulses, which are then stored in a sequence.

[0033] Table 2. Test Table of Dark Noise and Photon Amplitude Characteristics

[0034] Table 2 shows the data items that the generation operation depends on for determining the threshold.

[0035] S302: Call the effective attenuation pulse quantity set sequence and the unit trigger count parameter, perform traversal reading for each value in the effective attenuation pulse quantity set sequence, perform algebraic division operation on each value and the unit trigger count parameter to obtain the single-point trigger ratio, and perform array concatenation on all single-point trigger ratios in the original scan time order to establish a trigger probability sequence; For each value in the effective attenuation pulse count sequence, a traversal reading is performed, setting the first effective attenuation pulse count to 45. The unit trigger count parameter is extracted and set to a preset fixed value of 100 times by the acquisition device. An algebraic division operation is performed between the effective attenuation pulse count and the unit trigger count parameter, resulting in a single-point trigger ratio of 45 / 100, which is 0.45. The same division operation is performed on the second value in the sequence (52 values), resulting in a single-point trigger ratio of 52 / 100, which is 0.52. 0.45 and 0.52 are then concatenated into an array according to the original scan timestamps of 0.1 seconds and 0.2 seconds.

[0036] S303: Obtain the number of photons in the light pulse corresponding to the light pulse excitation sequence, perform differential operation on adjacent nodes in the trigger probability sequence to calculate the probability change increment, perform a comparison operation between the probability change increment and the preset judgment benchmark value, select a set of coherent nodes whose probability change increment exceeds the preset judgment benchmark value, map the two ends of the coherent node set to the light pulse photon number distribution dimension, extract the interval boundary digital segment restricted by the start and end coordinates, and obtain the probability growth interval; The specific method for setting the preset judgment benchmark value is as follows: extract the peak value of background probability fluctuation under the no-light triggering environment, obtain the standard triggering slope constant of the silicon photomultiplier tube under the fixed working state, perform an algebraic summation operation on the standard triggering slope constant and the peak value of background probability fluctuation, and set the numerical result generated by the algebraic summation operation as the preset judgment benchmark value. Obtain the photon number distribution of the light pulse corresponding to the light pulse excitation sequence, setting the distribution range to 1 to 100 photons. Perform a difference operation on adjacent nodes within the trigger probability sequence, setting the trigger probability of the second node to 0.52 and the trigger probability of the first node to 0.45, with the probability change increment = 0.52 - 0.45, resulting in a probability change increment of 0.07. Extract the background probability fluctuation peak value of 0.01 for 50 consecutive observation cycles under no-light triggering conditions. Obtain the standard trigger slope constant of 0.03 for the silicon photomultiplier tube in a fixed operating state of 54.74 volts. Perform an algebraic summation operation on the standard trigger slope constant and the background probability fluctuation peak value, with a preset judgment benchmark value = 0.03 + 0.01, resulting in a preset judgment benchmark value of 0.04. Compare the probability change increment of 0.07 with the preset judgment benchmark value of 0.04; 0.07 is greater than 0.04. Select all nodes with probability change increments greater than 0.04 for three consecutive times to construct a coherent node set. Map the positions of the first and last ends of the coherent node set to the positions of the 15th to 60th photons in the photon number distribution dimension of the light pulse, and extract the corresponding interval boundary digital segments from 15 to 60.

[0037] Please see Figure 5 The specific steps for obtaining the detection efficiency benchmark ratio are as follows: S401: Call the probability growth interval and the local extraction of fine-tuned photon step size, extract the start and end boundary values ​​of the probability growth interval, divide the start and end boundary values ​​into equal distances according to the fine-tuned photon step size to obtain the number of photons associated with each discrete node, and perform sequence combination operation on all photons according to the node position to construct a refined optical pulse signal; Extract the start and end boundary values ​​of the probability growth interval, setting the start boundary value to 15 and the end boundary value to 60. Extract the local fine-tuning photon step size, setting it to 5. Divide the start and end boundary values ​​into equidistant intervals according to the fine-tuning photon step size. The number of discrete nodes = end boundary value - start boundary value / fine-tuning photon step size. The number of discrete nodes = 60 - 15 / 5, resulting in 9 discrete intervals. Obtain the number of photons associated with each discrete node, which are 15, 20, 25, etc., up to 60. Arrange all photon counts (15, 20, 25, etc.) according to the node's index position in the array and perform a sequence combination operation to generate a one-dimensional tensor of length 10.

[0038] S402: Based on the refined optical pulse signal, obtain the associated detection response waveform, count the total accumulated attenuation of the detection response waveform within a unit number of triggers, perform algebraic division between the total accumulated attenuation and the unit number of triggers to obtain the single-point trigger variable, perform corresponding mapping and splicing of all single-point trigger variables according to the index of the refined optical pulse signal to generate a local trigger probability sequence. Refined optical pulse signals with corresponding photon counts of 15, 20, and 25 are sent to the detection input terminal to obtain the associated digital sampling sequence of the detection response waveform. The total number of cumulative attenuations exceeding the judgment threshold within 1000 unit trigger counts of this digital sampling sequence is counted. The total cumulative attenuation is set to 850 when the photon count is 20. The total cumulative attenuation is divided algebraically by the unit trigger count, resulting in a single-point trigger variable of 850 / 1000, which is 0.85. All single-point trigger variables are arranged according to the index order of the input photon counts of 15, 20, and 25, and 0.85 is placed at index 2 for corresponding mapping and splicing.

[0039] S403: Based on the local trigger probability sequence and the number of photons in the light pulse, perform sequence alignment of each node in the local trigger probability sequence with the number of photons in the light pulse, extract the aligned values ​​and divide them to obtain the probability photon ratio set, input the probability photon ratio set into the preset maximum likelihood estimation model, call the maximum likelihood estimation model to perform iterative fitting on the probability photon ratio set to obtain the convergent extreme value, and extract the detection efficiency benchmark ratio. The process of using the maximum likelihood estimation model to iteratively fit the probabilistic photon proportion set to obtain the convergent extreme value is as follows: Analyze the statistical distribution characteristics of the historical calibration test set; configure the maximum likelihood estimation model based on the matching expression architecture according to the statistical distribution characteristics; set the proportion values ​​within the probabilistic photon proportion set as iterative samples and map them to the input nodes of the maximum likelihood estimation model; set the detection efficiency parameter as the parameter to be estimated; establish a joint probability correlation by combining the parameter to be estimated and the iterative samples; perform a logarithmic transformation on the joint probability correlation to obtain the log-likelihood mapping; solve the partial derivative of the log-likelihood mapping with respect to the parameter to be estimated to obtain the optimization gradient direction; perform numerical updates of the parameter to be estimated along the optimization gradient direction; calculate the parameter variation difference between adjacent update rounds; generate a preset convergence tolerance threshold by multiplying the device noise floor parameter and the calculated bit width limit; compare the parameter variation difference with the preset convergence tolerance threshold; when the parameter variation difference is lower than the preset convergence tolerance threshold, stop the numerical update and set the current parameter to be estimated as the convergence extreme value.

[0040] The set of probability photon proportions is obtained by dividing aligned values. The local trigger probability sequence node value is set to 0.85, the number of light pulse photons is set to 20, and the probability photon proportion = 0.85 / 20, resulting in a proportion of 0.0425. The statistical distribution characteristics of dark count rate and crosstalk rate in the historical calibration test set are analyzed. A maximum likelihood estimation model is configured, internally constructed as a multilayer perceptron network structure containing one input layer, three hidden layers, and one output layer. The input layer has 1 neuron, receiving the probability photon proportion value; the first hidden layer contains 64 neurons, connected by fully connected logic, and uses the ReLU activation function to filter negative signals; the second hidden layer contains 32 neurons, using the ReLU activation function; the third hidden layer contains 16 neurons, using the Sigmoid activation function to compress the signal to between 0 and 1; the output layer contains 1 neuron, outputting the parameter to be estimated, i.e., the detection efficiency parameter. The value of 0.0425 in the probability photon proportion set is set as the iterative sample and mapped to the input node. A joint probability correlation following a Poisson distribution is established between the combined detection efficiency parameter and the iterative samples. A natural logarithmic transformation is performed on the joint probability correlation to obtain a log-likelihood mapping. The partial derivative of the log-likelihood mapping with respect to the detection efficiency parameter is solved to obtain the direction of the fastest numerical descent optimization gradient. Along the optimization gradient direction, the Adam optimization algorithm is used to update the detection efficiency parameter numerically with a learning rate of 0.001. The difference in parameter variation between the first and second update rounds is calculated, set to 0.405 for the second round and 0.400 for the first round. The difference is calculated as 0.405 - 0.400, resulting in a value of 0.005. The device noise floor parameter is extracted as 0.0001, the bit width limit is calculated to be 10.0, and the preset convergence tolerance threshold is 0.0001 × 10.0, resulting in a preset convergence tolerance threshold of 0.001. The parameter variation difference of 0.005 is compared with the preset convergence tolerance threshold of 0.001. If 0.005 is greater than 0.001, the update continues. By the 50th iteration, the parameter variation difference is 0.0008, which is lower than 0.001. The numerical update is stopped and the current detection efficiency parameter of 0.428 is set as the convergence extreme value.

[0041] Table 3. Parameters for Iterative Operation of Maximum Likelihood Estimation Model

[0042] Table 3 details the specific convergence process of parameter changes during iterative optimization.

[0043] Please see Figure 6 The specific steps for obtaining the bias compensation voltage are as follows: S501: Based on the detection efficiency benchmark ratio, collect the timestamps corresponding to the output pulses of multiple detection channels, subtract adjacent pulse timestamps to obtain the single-point time difference, combine the single-point time difference values ​​according to channel category to establish a time interval change sequence, call the preset time truncation threshold, compare the internal distribution values ​​of the time interval change sequence with the preset time truncation threshold, filter out the associated value items below the preset time truncation threshold, and obtain high-density trigger data; The specific method for setting the time cutoff threshold is as follows: obtain the set of dark count pulses of the detection channel under no-light conditions, count the average pulse interval parameter of the dark count pulse set within a fixed observation window, extract the preset tolerance scaling factor, perform a multiplication operation on the average pulse interval parameter and the tolerance scaling factor to obtain the product value, and set the product value as the preset time cutoff threshold. The timestamps corresponding to the output pulses from 16 detection channels were collected under the operating condition of the detection efficiency benchmark ratio. Subtraction was performed between the second timestamp (1050 nanoseconds) and the first timestamp (1000 nanoseconds) of adjacent pulses within channel 1, resulting in a single-point time difference of 50 nanoseconds (1050 - 1000). The calculated single-point time differences for all 16 channels were sequentially stored in an array to establish a time interval variation sequence. A set of 10,000 consecutive dark-count pulses was obtained from the detection channels under conditions of no light source and ambient temperature of 25 degrees Celsius. The timestamp distance of this set within a fixed 1-second observation window was calculated, yielding an average pulse interval parameter of 2000 nanoseconds. A preset tolerance scaling factor of 0.05 was extracted. The average pulse interval parameter and the tolerance scaling factor were multiplied, resulting in a product of 2000 × 0.05, which was 100 nanoseconds. This 100 nanosecond value was set as the preset time truncation threshold. The 50 nanosecond intervals distributed within the time interval variation sequence are compared with the preset time cutoff threshold of 100 nanoseconds. Values ​​with 50 nanosecond intervals lower than 100 nanosecond intervals are filtered out and stored in an array.

[0044] S502: Statistically calculate the accumulated pulse amount within the high-density trigger data, collect the total sampled pulse amount of the detection channel, perform division between the accumulated pulse amount and the total sampled pulse amount to obtain the channel proportion variable, and perform sequence merging and assembly of multiple channel proportion variables to generate the high-density trigger ratio. The accumulated number of clustered pulses with time intervals less than 100 nanoseconds within the high-density trigger data is statistically analyzed, with 150 clustered pulses counted for channel 1. The total number of sampled pulses output by channel 1 within the same time period is collected, with a total sampled pulse count of 2000. A division is performed between the accumulated pulse count and the total sampled pulse count, resulting in a channel proportion variable of 150 / 2000, yielding a channel proportion variable of 0.075 for channel 1. The same statistical calculation is performed on channel 2, yielding a channel proportion variable of 0.082. The channel proportion variables of 0.075 (channel 1) and 0.082 (channel 2) are then arranged in channel number order and merged into a sequence.

[0045] S503: Calculate the average reference term of each value within the high-density trigger ratio, perform subtraction operation on each value with the average reference term to extract the deviation from the absolute parameter, retrieve the maximum term within all deviations from the absolute parameter to extract the spatial deviation extreme value, perform addition operation on the device reference bias voltage and the spatial deviation extreme value to obtain the array balance correction amount, and establish the bias compensation voltage. The average value of the 16 channel proportion variables within the high-density trigger ratio is calculated, with the sum set to 1.6. The average reference term is calculated as 1.6 / 16, resulting in an average reference term of 0.1. The proportion variable of channel 1 (0.075) is subtracted from the average reference term (0.1), yielding the absolute deviation parameter as the absolute value of 0.075 - 0.1, which is 0.025. The highest value among the 16 absolute deviation parameters is retrieved, and the highest deviation parameter for channel 5 is set to 0.04. This 0.04 is extracted as the spatial deviation extreme value. The pre-set device reference bias voltage is set to 54.0 volts. The device reference bias voltage is then added to the spatial deviation extreme value, resulting in an array balance correction of 54.0 + 0.04, which is 54.04 volts. This 54.04 volts is written to the controller register.

[0046] Please see Figure 7 A SiPM testing and calibration system for particle detection applications, the SiPM testing and calibration system for particle detection applications is used to perform the aforementioned SiPM testing and calibration method for particle detection applications, the system comprising: The dark current feature analysis module acquires the steady-state dark current data of the silicon photomultiplier tube under the bias voltage sequence, calculates the current change slope sequence, extracts the second difference set of the slope, determines the position of the value in the second difference set that exceeds the differential cutoff threshold, and extracts the set of inflection points of current change. The breakdown voltage analysis module, based on the set of inflection points of current change, filters the local scanning voltage range, obtains the secondary dark current data in the local scanning voltage range, calculates the set of local slope change ratios, analyzes the numerical distribution in the set of local slope change ratios, and determines the breakdown inflection voltage point. The trigger probability analysis module, based on the breakdown transition voltage point, obtains the output pulse signal of the silicon photomultiplier tube under the optical pulse excitation sequence, calculates the trigger probability sequence, compares the trigger probability sequence with the number of optical pulse photons in the corresponding optical pulse excitation sequence, and extracts the probability growth interval. The detection efficiency benchmark extraction module calculates the local triggering probability sequence based on the probability growth interval and fine-tuning photon step size. It inputs the ratio parameter of the number of photons in the light pulse to the local triggering probability sequence into the maximum likelihood estimation model for calculation and extracts the detection efficiency benchmark ratio. The spatial deviation compensation module obtains the pulse time interval of the silicon photomultiplier tube in multiple detection channels based on the detection efficiency benchmark ratio, determines the value of the time interval less than the preset time threshold as high-density trigger data, calculates the spatial deviation extreme value, and adjusts the bias compensation voltage based on the spatial deviation extreme value.

[0047] It should be understood that the term "and / or" in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A alone, A and B simultaneously, or B alone. A and B can be singular or plural. Additionally, the character " / " in this article generally indicates an "or" relationship between the preceding and following related objects, but it can also represent an "and / or" relationship. Please refer to the context for a more accurate understanding.

[0048] In this invention, "at least one" means one or more, and "more than one" means two or more. "At least one of the following" or similar expressions refer to any combination of these items, including any combination of a single item or a plurality of items. For example, at least one of a, b, or c can represent: a, b, c, ab, ac, bc, or abc, where a, b, and c can be a single item or multiple items.

[0049] It should be understood that, in various embodiments of the present invention, the order of the above-mentioned process numbers does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0050] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0051] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the devices, apparatuses, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0052] In the several embodiments provided by this invention, it should be understood that the disclosed devices, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another device, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0053] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0054] In addition, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0055] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0056] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A SiPM testing and calibration method for particle detection applications, characterized by, Includes the following steps: S1: Obtain the steady-state dark current data of the silicon photomultiplier tube under the bias voltage sequence, calculate the current change slope sequence, extract the slope second difference set, determine the position of the value in the slope second difference set that exceeds the differential cutoff threshold, and extract the current change inflection point set. S2: Based on the set of current change inflection points, filter the local scanning voltage range, obtain the secondary dark current data in the local scanning voltage range and calculate the local slope change ratio set, analyze the numerical distribution in the local slope change ratio set, and determine the breakdown inflection voltage point. S3: Based on the breakdown transition voltage point, obtain the output pulse signal of the silicon photomultiplier tube under the optical pulse excitation sequence, calculate and obtain the trigger probability sequence, compare the trigger probability sequence with the number of optical pulse photons in the corresponding optical pulse excitation sequence, and extract the probability growth interval. S4: Based on the probability growth range and fine-tuning photon step size, calculate the local triggering probability sequence, input the ratio parameter of the number of photons in the optical pulse to the local triggering probability sequence into the maximum likelihood estimation model for calculation, and extract the detection efficiency benchmark ratio. S5: Based on the detection efficiency benchmark ratio, obtain the pulse time interval output by the silicon photomultiplier tube in multiple detection channels, determine the value of the time interval less than the preset time threshold as high-density trigger data, calculate the spatial deviation extreme value, and adjust the bias compensation voltage based on the spatial deviation extreme value.

2. The SiPM testing and calibration method for particle detection applications according to claim 1, characterized in that, The set of inflection points for current change includes critical abrupt change coordinates, response transition exponent, and differential transition nodes. The breakdown transition voltage point includes conduction attenuation threshold, steady-state level parameter, and breakdown edge potential. The probability growth interval includes response ignition lower bound, sensitivity ramp-up band, and linear response upper bound. The detection efficiency benchmark ratio includes quantum conversion coefficient, absolute calibration constant, and photoelectric gain factor. The bias compensation voltage includes dark noise reduction bias voltage, array balance correction amount, and leakage current correction amount.

3. The SiPM testing and calibration method for particle detection applications according to claim 1, characterized in that, The specific steps for obtaining the set of inflection points of current change are as follows: S101: Collect steady-state dark current data of silicon photomultiplier tubes within the bias voltage sequence, extract voltage difference variables for adjacent nodes within the bias voltage sequence, extract current difference variables at corresponding positions within the steady-state dark current data, perform algebraic division operation on the current difference variables and voltage difference variables to obtain node slope variables, and perform array concatenation on all node slope variables according to the scanning order to obtain the current change slope sequence; S102: Perform offset alignment on adjacent numerical terms within the current change slope sequence, extract the subsequent index value and subtract the preceding index value to obtain a first difference variable, perform adjacent term subtraction operation again on all first difference variables to obtain a second difference variable, arrange all second difference variables according to the original numerical index to generate a slope second difference set. S103: Extract the preset differential cutoff threshold, compare the distributed values ​​in the slope quadratic difference set with the differential cutoff threshold, filter the sequence target index numbers in the slope quadratic difference set whose values ​​exceed the differential cutoff threshold, extract the voltage parameters in the bias voltage sequence that match the sequence target index numbers, and obtain the current change inflection point set.

4. The SiPM testing and calibration method for particle detection applications according to claim 3, characterized in that, The specific method for obtaining the differential cutoff threshold is as follows: extract multiple quadratic difference variables located at the initial arrangement position from the slope quadratic difference set to construct a baseline sequence; calculate the sequence mean parameter and the baseline standard deviation parameter based on all quadratic difference variables in the baseline sequence; Extract the fluctuation confidence coefficient, which is set by collecting the distribution boundary multiple of the static dark level of the silicon photomultiplier tube in the unshielded state; multiply the background standard deviation parameter and the fluctuation confidence coefficient to obtain the tolerance offset; add the sequence mean parameter and the tolerance offset to generate the differential truncation threshold.

5. The SiPM testing and calibration method for particle detection applications according to claim 1, characterized in that, The specific steps for obtaining the breakdown inflection voltage point are as follows: S201: Call the set of current change inflection points and the current change slope sequence, extract voltage parameters for the nodes distributed in the set of current change inflection points, compare the value of each node at the corresponding position in the current change slope sequence, extract the node with the maximum value of the corresponding sequence value as a candidate target, calculate the voltage difference between adjacent nodes of the candidate target, select the node with the minimum voltage difference as the center coordinate, extend the preset voltage step size along both sides of the center coordinate, and establish the neighborhood offset scanning interval; S202: Collect the secondary dark current data corresponding to each voltage node in the neighborhood bias scanning interval, perform difference calculation on adjacent voltage nodes in the neighborhood bias scanning interval to obtain the local voltage increment, perform difference operation on adjacent values ​​in the secondary dark current data to obtain the local current increment, perform division operation on the local current increment and the local voltage increment to obtain the secondary slope parameter, extract the adjacent secondary slope parameters and perform ratio calculation to generate a local slope ratio sequence; S203: Arrange the distributed values ​​within the local slope ratio sequence to form a contrast gradient, extract the minimum value term within the contrast gradient, map the minimum value term to the corresponding voltage parameter node within the neighborhood bias scan interval, extract the digital record pointed to by the corresponding voltage parameter node, and obtain the breakdown inflection voltage point.

6. The SiPM testing and calibration method for particle detection applications according to claim 1, characterized in that, The specific steps for obtaining the probability growth interval are as follows: S301: Based on the breakdown transition voltage point, detect the initial pulse signal fed back during the period when the optical pulse excitation sequence acts on the silicon photomultiplier tube, extract the pulse amplitude feature data carried in the initial pulse signal, compare the pulse amplitude feature data with the preset attenuation judgment threshold, filter out the background waveforms below the preset attenuation judgment threshold, retain the nodes exceeding the preset attenuation judgment threshold as attenuation signals, count the cumulative value of attenuation signals within the set trigger period, and generate an effective attenuation pulse set sequence. S302: Call the effective attenuation pulse quantity set sequence and the unit trigger count parameter, perform traversal reading for each value in the effective attenuation pulse quantity set sequence, perform algebraic division operation on each value and the unit trigger count parameter to obtain the single-point trigger ratio, and perform array concatenation on all single-point trigger ratios in the original scan time order to establish a trigger probability sequence; S303: Obtain the number of photons in the light pulse corresponding to the light pulse excitation sequence; perform differential operation on adjacent nodes in the trigger probability sequence to obtain the probability change increment; perform a comparison operation between the probability change increment and a preset judgment benchmark value; select a set of coherent nodes whose probability change increment exceeds the preset judgment benchmark value; map the two ends of the coherent node set to the light pulse photon number distribution dimension; extract the interval boundary digital segment restricted by the start and end coordinates; and obtain the probability growth interval.

7. The SiPM testing and calibration method for particle detection applications according to claim 6, characterized in that, The process of setting the attenuation judgment threshold is as follows: collect a set of dark noise signals in a dark environment, extract the maximum amplitude distribution value in the set of dark noise signals, obtain the single photon amplitude reference value under the calibration state, add the maximum amplitude distribution value and the single photon amplitude reference value to obtain the amplitude reference sum, perform a division operation of the amplitude reference sum by two to obtain the center level parameter, and establish the center level parameter as the preset attenuation judgment threshold. The specific method for setting the preset judgment benchmark value is as follows: extract the peak value of background probability fluctuation under the no-light triggering environment, obtain the standard triggering slope constant of the silicon photomultiplier tube under the fixed working state, perform an algebraic summation operation on the standard triggering slope constant and the peak value of background probability fluctuation, and set the numerical result generated by the algebraic summation operation as the preset judgment benchmark value.

8. The SiPM testing and calibration method for particle detection applications according to claim 1, characterized in that, The specific steps for obtaining the detection efficiency benchmark ratio are as follows: S401: Call the probability growth interval and the locally extracted fine-tuned photon step size, extract the start and end boundary values ​​of the probability growth interval, divide the start and end boundary values ​​into equal distances according to the fine-tuned photon step size to obtain the number of photons associated with each discrete node, and perform a sequence combination operation on all photons according to the node position to construct a refined optical pulse signal; S402: Based on the refined optical pulse signal, obtain the associated detection response waveform, count the total accumulated attenuation of the detection response waveform within a unit number of triggers, perform algebraic division between the total accumulated attenuation and the unit number of triggers to obtain the single-point trigger variable, and perform corresponding mapping splicing on all single-point trigger variables according to the index of the refined optical pulse signal to generate a local trigger probability sequence. S403: Based on the local trigger probability sequence and the number of photons in the light pulse, perform sequence alignment of each node of the local trigger probability sequence with the number of photons in the light pulse, extract the aligned values ​​and divide to obtain the probability photon ratio set, input the probability photon ratio set into a preset maximum likelihood estimation model, call the maximum likelihood estimation model to perform iterative fitting on the probability photon ratio set to obtain the convergent extreme value, and extract the detection efficiency benchmark ratio.

9. The SiPM testing and calibration method for particle detection applications according to claim 1, characterized in that, The specific steps for obtaining the bias compensation voltage are as follows: S501: Based on the detection efficiency benchmark ratio, collect the timestamps corresponding to the output pulses of multiple detection channels, subtract adjacent pulse timestamps to obtain the single-point time difference, combine the single-point time difference values ​​according to channel category to establish a time interval change sequence, call the preset time truncation threshold, compare the distributed values ​​inside the time interval change sequence with the preset time truncation threshold, filter out the associated value items below the preset time truncation threshold, and obtain high-density trigger data; S502: Calculate the accumulated amount of clustered pulses in the high-density trigger data, collect the total amount of pulses sampled by the detection channel, divide the accumulated amount of pulses by the total amount of sampled pulses to obtain the channel ratio variable, and perform sequence merging and assembly of multiple channel ratio variables to generate the high-density trigger ratio. S503: Calculate the average reference term of each value within the high-density trigger ratio, perform subtraction operation on each value with the average reference term to extract the deviation from the absolute parameter, retrieve the maximum term within all deviations from the absolute parameter to extract the spatial deviation extreme value, perform addition operation on the device reference bias voltage and the spatial deviation extreme value to obtain the array balance correction amount, and establish the bias compensation voltage.

10. A SiPM testing and calibration system for particle detection applications, characterized in that, The system is used to implement the SiPM testing and calibration method for particle detection applications as described in any one of claims 1-9, and the system comprises: The dark current feature analysis module acquires the steady-state dark current data of the silicon photomultiplier tube under the bias voltage sequence, calculates the current change slope sequence, extracts the second difference set of the slope, determines the position of the value in the second difference set that exceeds the differential cutoff threshold, and extracts the set of inflection points of current change. The breakdown voltage analysis module, based on the set of current change inflection points, filters the local scanning voltage range, obtains the secondary dark current data within the local scanning voltage range, calculates the local slope change ratio set, analyzes the numerical distribution within the local slope change ratio set, and determines the breakdown inflection voltage point. The trigger probability analysis module, based on the breakdown transition voltage point, obtains the output pulse signal corresponding to the silicon photomultiplier tube under the optical pulse excitation sequence, calculates and obtains the trigger probability sequence, compares the trigger probability sequence with the number of optical pulse photons in the corresponding optical pulse excitation sequence, and extracts the probability growth interval. The detection efficiency benchmark extraction module calculates the local triggering probability sequence based on the probability growth interval and the fine-tuned photon step size. It inputs the ratio parameter of the number of photons in the light pulse to the local triggering probability sequence into the maximum likelihood estimation model for calculation and extracts the detection efficiency benchmark ratio. The spatial deviation compensation module, based on the detection efficiency benchmark ratio, obtains the pulse time interval output by the silicon photomultiplier tube in multiple detection channels, determines the value of the time interval less than the preset time threshold as high-density trigger data, calculates the spatial deviation extreme value, and adjusts the bias compensation voltage based on the spatial deviation extreme value.

Images