Semiconductor detector particle pulse waveform data generation and analysis platform based on digital twinning technology

Abstract

The invention provides a semiconductor detector particle pulse waveform data generation and analysis platform based on a digital twinning technology, and the platform comprises an automatic full-link multi-working-condition semiconductor detector particle pulse waveform simulation system which is used for controlling GEANT4, Weightfield2 and PyLTSPICE software, and outputting simulated energy data and differential current data; the dual-channel pulse waveform automatic acquisition system is used for acquiring and storing energy data and current data actually measured by the accelerator by using a physical entity and performing multi-channel display; and the digital twinborn and physical twinborn data comparison and analysis system is used for performing waveform matching degree analysis by utilizing the actually measured waveform of the accelerator and the simulation waveform under the same working condition and the same circuit parameter. The method has the advantage that mass pulse waveform data under different working conditions can be efficiently and automatically generated.

Description

Technical Field

[0001] The present application belongs to the field of particle detection technology, and specifically relates to a semiconductor detector particle pulse waveform data generation and analysis platform based on digital twin technology. Background Art

[0002] Charged particles in space include galactic cosmic rays, solar cosmic rays, solar wind and Van Allen radiation belts. Galactic cosmic rays are composed of high-energy charged particle flows from all directions of the Milky Way, and the main components are protons, alpha particles and other nuclear components; solar cosmic rays are high-energy charged particle flows emitted during solar flares, mainly composed of protons; solar wind is the plasma flow that the sun continuously ejects outward, mainly composed of hydrogen plasma, including a small amount of other components; Van Allen radiation belts, also known as the Earth radiation belts, are located near the earth and are radiation belts captured by the geomagnetic field. They are divided into inner radiation belts and outer radiation belts, and are mainly composed of protons and electrons. Faced with a variety of charged particles in space, it is necessary to accurately understand their energy spectra, types, and pitch angles in order to further study the origin of particles, acceleration and propagation processes, lifetimes and dynamics.

[0003] Accelerator experiments are currently the most direct and commonly used method for obtaining pulse waveform data of charged particles incident on semiconductor detectors. They are widely used in particle physics, high-energy physics experiments, and detector performance evaluation. By providing high-energy particle beams through accelerators, researchers can accurately measure the signal response generated by particles in the detector, providing reliable reference data for the design, optimization, and experimental verification of detectors. Although accelerator experiments are of great significance, they still have significant limitations in practical applications:

[0004] 1. Limitations on particle types and energy range:

[0005] The particle types and energy ranges generated in accelerator experiments are usually limited by the performance of the equipment and cannot cover all possible physical conditions. Complex experimental requirements may require a lot of time to adjust.

[0006] 2. High costs and resource sharing issues:

[0007] The operating costs of accelerators are extremely high, including the cost of equipment maintenance, energy consumption, and operating personnel;

[0008] The experimental time of equipment is usually a shared resource and needs to wait for scheduling and allocation, which leads to delays in research progress;

[0009] Multiple experimental adjustments and tests often take a lot of time to cover a wide range of parameters, further increasing costs and cycles.

[0010] 3. Detector loss and stability of experimental results:

[0011] Long-term high-energy particle beam incidence may cause the accumulation of detector lattice defects, thus affecting the accuracy of test results and the long-term performance of the equipment.

[0012] In view of the high cost and limited experimental conditions of accelerator experiments, many scholars have begun to use simulation methods to study and analyze the physical process and electronic performance of charged particles incident on semiconductor detectors. Among them, GEANT4, as an efficient tool based on the Monte Carlo method, can accurately simulate the motion trajectory of charged particles in the detector and its energy deposition process by defining the particle source, detector material and geometric structure. At the same time, combined with the physical process model, GEANT4 can describe the electromagnetic interaction and nuclear reaction process between particles and detector materials in detail, providing important support for the design optimization and performance prediction of detectors.

[0013] However, the simulation results of GEANT4 are limited to the physical process data of the energy deposition of particles in the detector, and cannot directly reflect the output signal of the detector. Therefore, many researchers have tried to combine GEANT4 with SPICE software and use SPICE tools (such as PSPICE or LTSPICE) to evaluate the circuit performance of the preamplifier and the main amplifier. SPICE tools can study the gain characteristics, time response and noise effects of the circuit on the input signal through precise transient simulation analysis, thereby optimizing the signal readout path. In addition, this combination can achieve a comprehensive simulation of the system-level performance of particle detectors to a certain extent, and improve the reliability and efficiency of the design. Nevertheless, the simulation method combining GEANT4 with SPICE software can usually only obtain the peak information of the pulse signal, but lacks accurate simulation of the full time domain waveform, and cannot fully reflect the dynamic changes during the particle incidence process, which limits the in-depth understanding of particle behavior.

[0014] In order to solve this problem, some scholars began to combine the Weightfield2 tool to simulate the movement of carriers in semiconductor detectors under the action of electric fields. Weightfield2 is an efficient tool that can accurately predict the characteristics of the detector output signal by quickly calculating the electric field distribution and the drift and diffusion process of carriers, providing important support for detector design and optimization. Compared with the simple combination of GEANT4 and SPICE, this method can more directly simulate the transient response characteristics of the sensor.

[0015] However, the current simulation methods still have the following important shortcomings:

[0016] 1. Lack of automated data generation capabilities for multiple working conditions:

[0017] The current simulation process usually requires researchers to manually adjust parameters such as particle type, energy, and incident angle, which cannot efficiently cover a variety of experimental conditions, especially under extreme conditions. The lack of an automated simulation system leads to inefficient data generation and cannot meet the needs of data-driven technologies such as deep learning for large-scale, high-quality data.

[0018] 2. Difficulty in achieving seamless integration across tools:

[0019] The functions of tools such as GEANT4, Weightfield2 and SPICE have different focuses, but their combination often requires complex interface conversion and unification of intermediate data formats, which leads to fragmentation of the entire process and reduces the ease of use and efficiency of simulation.

[0020] 3. Experimental noise and lack of actual conditions:

[0021] Current simulation data is usually idealized, without considering the inevitable noise interference and equipment defects in physical experiments, such as electromagnetic noise, material inhomogeneity, etc. The lack of these factors will cause the simulation results to deviate from the actual application scenarios and reduce the actual reference value of the model. Summary of the invention

[0022] The purpose of this application is to overcome the defects of the prior art that it lacks the ability to automatically generate data for multiple working conditions and is difficult to achieve seamless integration across tools.

[0023] In order to achieve the above objectives, the present application proposes a semiconductor detector particle pulse waveform data generation and analysis platform based on digital twin technology, the platform comprising:

[0024] An automated full-link multi-operation semiconductor detector particle pulse waveform simulation system, used to control GEANT4, Weightfield2 and PyLTSPICE software, and output simulated energy data and differential current data;

[0025] A dual-channel pulse waveform automated acquisition system for collecting and storing accelerator measured energy and current data using physical entities and for multi-channel display; and

[0026] The digital twin and physical twin data comparison and analysis system is used to perform waveform matching analysis on the accelerator measured waveform obtained by the dual-channel pulse waveform automatic acquisition system and the simulated waveform under the same working condition and circuit parameters obtained by the automated full-link multi-working condition semiconductor detector particle pulse waveform simulation system. The analysis process includes comparing the rising edge slope of energy information, the differential current pulse width, the mean square error and the wavelet change.

[0027] As an improvement of the above platform, the processing process of the automated full-link multi-condition semiconductor detector particle pulse waveform simulation system includes:

[0028] Run GEANT4 software to simulate and output multiple-incident energy loss data and draw curves;

[0029] Run the Weightfield2 software to simulate and output the detector time domain waveform data based on the incident energy loss data;

[0030] Run PyLTSpice software to simulate and output the final dual-channel energy and differential current data based on the time domain waveform data.

[0031] As an improvement of the above platform, the dual-channel pulse waveform automatic acquisition system includes:

[0032] Si-PN junction ion implantation type semiconductor sensor, used to input the particles of the calibration source and generate the corresponding pulse current signal;

[0033] A preamplifier, used to amplify the pulse circuit signal and output an energy signal and a current signal;

[0034] A main amplifier, used to re-amplify the energy signal and current signal output by the preamplifier;

[0035] A high-speed acquisition board is used to collect the energy signal and circuit signal output by the main amplifier.

[0036] As an improvement of the above platform, the data processing process of the dual-channel pulse waveform automatic acquisition system includes:

[0037] The waveform of each energy signal and current signal collected by the high-speed acquisition board is preprocessed to obtain the peak-to-peak value and draw a peak-to-peak value distribution bar graph;

[0038] According to the distribution of the peak-to-peak value histogram, the interval of the maximum peak-to-peak value count is obtained and a counting interval is expanded on both sides as valid data to eliminate low-quality waveforms obtained due to noise or abnormal acquisition;

[0039] Find the maximum value of the waveform and extract several points to the left and right as output information. If there are not enough points, fill them according to the end value.

[0040] As an improvement of the above platform, the analysis process of the digital twin and physical twin data comparison and analysis system includes:

[0041] Perform wavelet transform on the simulated and measured energy and current data respectively, and generate denoised waveform data after determining the number of decomposition layers;

[0042] The energy waveform data is analyzed for matching. The different time discrete slopes of the rising edge of the energy output waveforms of the two modes are calculated, and the difference of the discrete slopes is squared to judge the difference. The peak-to-peak difference is calculated to evaluate the deviation of the absolute energy information.

[0043] The current waveform data is analyzed for matching. The two modes of waveform data are normalized by softmax and the cross entropy is calculated to determine the difference in shape. The peak-to-peak value difference of the current is calculated to measure the overall consistency of the current intensity.

[0044] Compared with the prior art, the advantages of this application are:

[0045] 1. Efficiently and automatically generate massive pulse waveform data under different working conditions (different temperatures, different incident angles, different energies, different particles, front and rear incidents, and different bias voltages);

[0046] 2. The simulation model can be evaluated and optimized using measured data;

[0047] 3. The data generation and analysis platform can be used to obtain data results that are difficult to simulate with ground accelerators;

[0048] 4. The data obtained from the data generation and analysis platform can be applied to newly developed equipment as a data supplement for the calibration system. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 Shown is the structural diagram of the semiconductor detector particle pulse waveform data generation and analysis platform based on digital twin technology. DETAILED DESCRIPTION

[0050] The technical solution of the present application is described in detail below with reference to the accompanying drawings.

[0051] With the rapid development of deep learning technology, more and more scholars have begun to use multi-layer perceptrons, convolutional neural networks or recurrent neural networks for particle identification or position analysis. These methods rely heavily on massive pulse waveform data, but the existing simulation technology is difficult to efficiently generate data with high coverage, multiple scenarios and matching the actual test state, which has become one of the bottlenecks restricting the application of deep learning technology, and digital twin technology is expected to solve this problem. Digital twin technology is an emerging technology that combines physical systems with virtual models to achieve full life cycle optimization through data interaction and simulation analysis. Its core components include physical entities, digital models and data interaction interfaces. The core feature is the high-precision mapping and real-time interaction between virtual entities and physical entities, so that the state of complex systems can be intuitively reflected and dynamically optimized in digital models. Although this technology has been widely used in industries such as industry, aerospace and energy, it has not yet been applied in the field of particle detection, especially in the field of pulse waveform particle identification.

[0052] In order to solve the problems of high cost, high time consumption, low parameter coverage of accelerator experiments in particle detection, non-automation of multi-conditions in particle simulation, no cross-tool data interface, and deviation of simulation results from actual results, this application provides a semiconductor detector particle pulse waveform data generation and analysis platform based on digital twin technology, including the digital twin GEANT4+Weightfield2+PyLTSpice automated full-link multi-condition semiconductor detector particle pulse waveform generation system, the physical twin calibration source+semiconductor sensor+preamplifier and main amplifier+fast acquisition board+host computer software automated acquisition system, and digital twin and physical twin data comparison and analysis system. Based on the original full-link semiconductor detector particle pulse waveform generation system, the digital twin embeds the multi-condition automation function into each simulation step, unifies the data interface of each tool, and uses the physical twin to automatically collect the measured data of the same condition and evaluate and analyze it, laying a solid foundation for a comprehensive and in-depth understanding of the characteristics of different charged particle pulse waveforms and the application of deep learning in the field of particle detection.

[0053] The semiconductor detector particle pulse waveform data generation and analysis platform based on digital twin technology mainly includes three systems, namely, the automated full-link multi-working condition semiconductor detector particle pulse waveform simulation system, the dual-channel pulse waveform automatic acquisition system, and the digital twin and physical twin data comparison and analysis system.

[0054] The simulation process of the automated full-link multi-condition semiconductor detector particle pulse waveform simulation system includes three key steps: first, use the script control method to set the conditional space that GEANT4 needs to simulate, such as energy, direction, particle type, step size, etc., and read the corresponding conditions to start the multi-incident event simulation. The output energy loss information and the corresponding conditional space are recorded in txt format and sent to the next step; the script determines in real time whether the GEANT4 simulation is completed. When it is completed, open the Weightfield2 software and read the energy loss txt file and the corresponding input conditions (energy, direction, particle type, step size) at the same time, and set the influencing parameter space of the carrier transport process (temperature, front and rear end incidence, bias, magnetic field, etc.), use the Batch mode to simulate the multi-incident events, and finally output the result txt file, including the hole induced time domain current waveform, the electron induced time domain current waveform, and the total induced time domain current waveform; finally, use the script to monitor the Weightfield2 simulation log status file. When it is completed, start PyLTSPICE to read the time domain current waveform of multiple incidences and combine the circuit parameters such as the preamplifier, main amplifier and shaping module to finally output the energy data and differential current data.

[0055] The dual-channel pulse waveform automatic acquisition system uses low-noise energy output and differential current dual-channel preamplifiers, low-noise main amplifiers, high-speed acquisition boards, and multi-channel fast-view and fully automatic data storage host computers to achieve accelerator high-speed pulse waveform acquisition and form a particle waveform database through preprocessing methods.

[0056] The digital twin and physical twin data comparison and analysis system uses the dual-channel pulse waveform automatic acquisition module to obtain the accelerator's measured waveform and the automated full-link multi-condition semiconductor detector particle pulse waveform simulation system to perform waveform matching analysis on the simulated waveform under the same operating conditions and circuit parameters. The analysis methods include comparing the rising edge slope of the energy information, the differential current pulse width, the mean square error, and the wavelet change.

[0057] The systems of the semiconductor detector particle pulse waveform data generation and analysis platform based on digital twin technology are as follows:

[0058] 1. Automated full-link multi-operating-condition semiconductor detector particle pulse waveform simulation system

[0059] The system uses a top-level script to control the overall process. The controlled software includes GEANT4, Weightfield2 and PyLTSPICE, and finally outputs energy and differential current waveforms.

[0060] 1)GEANT4 generates conditional space .mac file

[0061] GEANT4 generates .mac files with different energies, directions, particle types, and step sizes. The energy is 1MeV, 3MeV, and 5MeV. The directions are 0°, 15°, 30°, -15°, and -30°. The particle types include H, He, B, C, O, Al, Si, P, Fe, and Cu. The step size is 1um for H, and 0.1um for other heavy ions because of their shorter range. The specific parameter space needs to be adjusted according to the actual test status.

[0062] 2) Run GEANT4 simulation to output multiple incident energy loss data and draw curves

[0063] Construct the incident particles, sensor model, event trigger function, event end function, step start function, step end function, etc., and run the .mac file under the current conditions to obtain the energy loss step data txt file, range count curve, energy loss step curve, simulation condition txt file and simulation log txt file after multiple incident events.

[0064] 3) Run Weightfield2 software and configure the same operating parameters as GEANT4

[0065] Use the top-level script file to monitor the GEANT4 simulation log file in real time. When the simulation is completed, start the Weightfield2 program and use the simulation conditions txt entered in the previous step as simulation parameters (including energy, direction, step size, type). At the same time, it is necessary to further generate the conditional space of Weightfield2 itself, including temperature, bias and magnetic field). The temperature conditional space selects three levels, namely 273K, 300K, and 327K; the bias is generally set according to the characteristics of the sensor. For example, the recommended working voltage of the Aotek ion implantation PN junction detector is 100V, and 60V and 80V are set as the conditional space; the magnetic field is generally not considered.

[0066] 4) Run Weightfield2 simulation to output detector time domain waveform data

[0067] The top-level script is used to start the detector construction of Weightfield2, the calculation of the built-in potential field and weight field, and the current calculation in sequence, and the batch processing mode is started to simulate the input energy loss step data in sequence, and all current time domain waveforms under the current working condition are output as txt files (the sampling rate uses the default 40GS / s). After the program detects that the simulation of a certain working condition is completed, the current simulation of the next working condition (such as temperature change) is started until it ends. The final current time domain waveform includes the hole induction time domain current waveform, the electron induction time domain current waveform, and the total induction time domain current waveform.

[0068] 5) Use LTSpice to build the preamplifier, main amplifier and shaping circuit

[0069] The premise of using PyLTSpice for automated simulation is to build a corresponding circuit diagram based on the actual circuit architecture. The preamplifier mainly includes two outputs, namely the charge-sensitive preamplifier part and the current-sensitive preamplifier part. The main components include the front-end input pulse waveform current source, operational amplifier, feedback resistor and feedback capacitor, shaping resistor and capacitor. The main amplifier includes an operational amplifier and an amplification ratio resistor. After the circuit is configured, export the .cir file as the input of PyLTSpice.

[0070] 6) Run PyLTSpice to output the final dual-channel energy and differential current data

[0071] First, set up the back-end circuit parameter traversal space. Its main parameters include the feedback resistor, feedback capacitor, and shaping resistor and capacitor of the preamplifier. Load a single current pulse waveform file through the program, import it into the pulse waveform current source, and finally output the corresponding energy and differential current data. After completing all simulation outputs of a certain set of parameters, automatically call the new parameters to re-simulate and repeat the above steps until the simulation ends. All output data files are named with the current full-process simulation input parameters, including particle type, energy, incident direction, step size, temperature, bias, front and rear ends, preamplifier feedback resistor and capacitor, shaping resistor and capacitor, and main amplifier proportional resistor.

[0072] 2. Dual-channel pulse waveform automatic acquisition system

[0073] The system uses physical entities to automatically collect and store accelerator measured data and perform multi-channel quick viewing.

[0074] Physical entity dual-channel pulse waveform acquisition and storage architecture design:

[0075] The dual-channel pulse waveform automatic acquisition system includes Si-PN junction ion implantation semiconductor sensor, low-noise energy output and differential current dual-channel preamplifier, low-noise main amplifier, high-speed acquisition board, multi-channel quick view and fully automatic data storage host computer. In the actual experiment, the particles of the calibration source are first incident into the front-end Si-PN junction ion implantation semiconductor sensor to generate the corresponding pulse current signal. The semiconductor sensor uses CU-013-150-300 of Aotek; the current signal is amplified by the preamplifier to output energy signal and current signal. The preamplifier uses the low-noise dual-channel preamplifier Model 2003BT of Canberra; because the current signal is small, the main amplifier is used to further amplify the current signal. The main amplifier uses Aotek's 9302, and the magnification is 20 times; the final output signal is collected by the high-speed acquisition board and stored in the matching quick view waveform at the same time. The high-speed acquisition board uses the 1GS / s sampling rate board of Zhongke Caixiang.

[0076] The data automatic preprocessing functions of the dual-channel pulse waveform automatic acquisition system include:

[0077] Each waveform obtained by fast sampling is preprocessed to obtain the peak-to-peak value and draw a peak-to-peak distribution bar graph. Due to the volatility of particle energy and the influence of noise, the peak-to-peak value will also be presented in the form of a spectrum. According to the distribution of the peak-to-peak bar graph, the interval of the maximum peak-to-peak count is first obtained and a counting interval is expanded to both sides as valid data to eliminate low-quality waveforms obtained due to noise or abnormal acquisition; in order to further reduce the amount of data and improve the quality of the database, the position of the waveform maximum value is first found and C points and D points are extracted to the left and right respectively as key information. If it is insufficient, it is filled according to the end value and a new excel file is generated, in which each line is a waveform data; in addition, in order to more conveniently send this data to the subsequent deep learning model architecture for particle identification, a normalized excel file will be output at the same time.

[0078] 3. Digital twin and physical twin data comparison and analysis system

[0079] The system uses the measured waveform of the accelerator obtained by the dual-channel pulse waveform automatic acquisition system and the simulated waveform under the same operating conditions and circuit parameters obtained by the automated full-link multi-operating condition semiconductor detector particle pulse waveform simulation system to perform waveform matching analysis. The analysis methods include comparing the rising edge slope of the energy information, the differential current pulse width, the mean square error, and the wavelet change.

[0080] 1) Data import into data analysis system

[0081] The system pre-saves the energy and current waveform data of various working conditions measured by the accelerator. After completing the digital simulation of a certain working condition, the energy waveform, current waveform and simulation parameters output by the digital twin are directly input into the data analysis system, and compared with the measured waveform.

[0082] 2) Comparative analysis of simulation data and measured data

[0083] The main purpose of analyzing the simulation data and measured data is to analyze the matching degree of the two waveforms, and the data will be analyzed from several dimensions.

[0084] In the first step, wavelet transform is performed on the simulated and measured energy and current data respectively, and the denoised waveform data is generated after the number of decomposition layers is determined, so as to reduce the influence of noise on the waveform matching analysis.

[0085] The second step is to analyze the matching of energy waveform data by calculating the different time discrete slopes of the rising edge of the energy output waveform of the two modes, such as calculating the slopes of the waveform at 10% height, 30% height, 50% height and 90% height, and square the difference of the above four data to judge the difference. In addition, the peak-to-peak difference is calculated to evaluate the deviation of the absolute energy information.

[0086] The third step is to perform a matching analysis on the current waveform data. With the help of the concept of cross entropy in the deep learning classification model, the two mode waveform data are first softmax normalized and the cross entropy is calculated to determine the difference in shape; in addition, the difference in current peak-to-peak value is calculated to measure the overall consistency of current intensity.

[0087] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present application and are not intended to limit it. Although the present application is described in detail with reference to the embodiments, a person skilled in the art should understand that any modification or equivalent replacement of the technical solution of the present application does not depart from the spirit and scope of the technical solution of the present application and should be included in the scope of the claims of the present application.

Claims

1. A semiconductor detector particle pulse waveform data generation and analysis platform based on digital twin technology, characterized in that: The platform includes: An automated full-link multi-operation semiconductor detector particle pulse waveform simulation system, used to control GEANT4, Weightfield2 and PyLTSPICE software, and output simulated energy data and differential current data; A dual-channel pulse waveform automated acquisition system for collecting and storing accelerator measured energy and current data using physical entities and for multi-channel display; and The digital twin and physical twin data comparison and analysis system is used to perform waveform matching analysis on the accelerator measured waveform obtained by the dual-channel pulse waveform automatic acquisition system and the simulated waveform under the same working condition and circuit parameters obtained by the automated full-link multi-working condition semiconductor detector particle pulse waveform simulation system. The analysis process includes comparing the rising edge slope of energy information, the differential current pulse width, the mean square error and the wavelet change.

2. The semiconductor detector particle pulse waveform data generation and analysis platform based on digital twin technology according to claim 1 is characterized in that: The processing process of the automated full-link multi-operating-mode semiconductor detector particle pulse waveform simulation system includes: Run GEANT4 software to simulate and output multiple-incident energy loss data and draw curves; Run the Weightfield2 software to simulate and output the detector time domain waveform data based on the incident energy loss data; Run PyLTSpice software to simulate and output the final dual-channel energy and differential current data based on the time domain waveform data.

3. The semiconductor detector particle pulse waveform data generation and analysis platform based on digital twin technology according to claim 1 is characterized in that: The dual-channel pulse waveform automatic acquisition system comprises: Si-PN junction ion implantation type semiconductor sensor, used to input the particles of the calibration source and generate the corresponding pulse current signal; A preamplifier, used to amplify the pulse circuit signal and output an energy signal and a current signal; A main amplifier, used to re-amplify the energy signal and current signal output by the preamplifier; A high-speed acquisition board is used to collect the energy signal and circuit signal output by the main amplifier.

4. The semiconductor detector particle pulse waveform data generation and analysis platform based on digital twin technology according to claim 3 is characterized in that: The data processing process of the dual-channel pulse waveform automatic acquisition system includes: The waveform of each energy signal and current signal collected by the high-speed acquisition board is preprocessed to obtain the peak-to-peak value and draw a peak-to-peak value distribution bar graph; According to the distribution of the peak-to-peak value histogram, the interval of the maximum peak-to-peak value count is obtained and a counting interval is expanded on both sides as valid data to eliminate low-quality waveforms obtained due to noise or abnormal acquisition; Find the maximum value of the waveform and extract several points to the left and right as output information. If there are not enough points, fill them according to the end value.

5. The semiconductor detector particle pulse waveform data generation and analysis platform based on digital twin technology according to claim 1, characterized in that: The analysis process of the digital twin and physical twin data comparison and analysis system includes: Perform wavelet transform on the simulated and measured energy and current data respectively, and generate denoised waveform data after determining the number of decomposition layers; The energy waveform data is analyzed for matching. The different time discrete slopes of the rising edge of the energy output waveforms of the two modes are calculated, and the difference of the discrete slopes is squared to judge the difference. The peak-to-peak difference is calculated to evaluate the deviation of the absolute energy information. The current waveform data is analyzed for matching. The two modes of waveform data are normalized by softmax and the cross entropy is calculated to determine the difference in shape. The peak-to-peak value difference of the current is calculated to measure the overall consistency of the current intensity.

Images