A High-Sensitivity Detection Method and System for Energy-Dispersive X-ray Fluorescence Spectrometer

Abstract

This application provides a high-sensitivity detection method and system for energy-dispersive X-ray fluorescence spectrometers, relating to the field of X-ray fluorescence spectroscopy detection technology. This application obtains initial spectral information on the photon number distribution by acquiring spectral data of the sample to be tested, matches and analyzes it with an elemental characteristic spectral library, and resolves chemical shift effects to generate enhanced characteristic spectra with elemental chemical state labels. Based on this, the excitation parameters of the beam filtering system are adjusted, and the crystal reflection conditions and incident angle are optimized to obtain matching single-wavelength excitation condition parameters. These two parameters are then cross-correlated and fused through a dual-channel architecture to form multi-dimensional analytical features. A hierarchical mapping relationship is then established through hierarchical feature space transformation and identification model adaptation to achieve high-sensitivity identification of elemental chemical states. This method enables high-sensitivity identification of elemental chemical states in samples through spectral data analysis, excitation parameter optimization, dual-channel collaborative processing, and hierarchical mapping modeling.

Description

Technical Field

[0001] This application relates to the field of X-ray fluorescence spectroscopy detection technology, and in particular to a high-sensitivity detection method and system for energy-dispersive X-ray fluorescence spectrometers. Background Technology

[0002] In fields such as environmental monitoring, new material development, and geological exploration, the demand for elemental chemical state analysis of unknown and complex samples is increasingly urgent, often facing scenarios where no standard samples are available for reference. Such analyses require accurate identification of the chemical states of elements with low abundance, while also addressing issues such as multi-element interference and complex matrix effects within the sample. This places stringent demands on the sensitivity, anti-interference capabilities, and standard-free compatibility of the detection methods.

[0003] The current mainstream solution for this requirement is a standard-free analysis scheme based on matching a preset standard spectral library with fixed excitation parameters. This scheme compares the acquired sample spectrum with the preset standard spectral library, corrects matrix effects with algorithms, and uses fixed beam excitation parameters and optical component settings to achieve preliminary identification of the chemical state of elements in the absence of a standard.

[0004] The existing scheme has significant drawbacks: First, the fixed excitation parameters cannot adapt to the differences in spectral characteristics corresponding to different chemical states of elements, and it is difficult to cope with the peak position shift and intensity change caused by chemical shift effects, resulting in low excitation efficiency; Second, the preset spectral library matching ignores the spectral distortion caused by chemical shifts in complex samples, which is prone to matching bias and reduces the accuracy of identification; Third, it lacks a dynamic synergistic optimization mechanism for spectral characteristics and excitation conditions, and its sensitivity is insufficient when dealing with low-content elements, which cannot meet the actual needs of high-precision analysis of unknown and complex samples. Summary of the Invention

[0005] The purpose of this application is to provide a high-sensitivity detection method and system for energy-dispersive X-ray fluorescence spectrometers, in order to solve the problems in existing standard-free analysis schemes, such as poor adaptability of fixed excitation parameters, susceptibility of spectral library matching to deviations caused by chemical shifts, and lack of dynamic collaborative optimization, which lead to insufficient sensitivity and accuracy in identifying the chemical states of elements in unknown and complex samples.

[0006] To address the aforementioned technical problems, in a first aspect, this application provides a high-sensitivity detection method for an energy-dispersive X-ray fluorescence spectrometer, comprising:

[0007] Energy dispersive X-ray fluorescence spectral data of the sample to be tested are collected to obtain initial spectral information including the photon number distribution;

[0008] Based on the initial spectral information and the elemental feature map library, a matching analysis is performed. By analyzing the chemical shift effect characterized by the peak position shift and intensity change features of the spectral lines, an enhanced feature spectrum containing elemental chemical state labels is generated.

[0009] Based on the element chemical state markings in the enhanced characteristic spectrum, the excitation parameters of the beam selection system are adjusted and optimized through a feedback control mechanism to generate excitation condition parameters of a single wavelength that match the current element chemical state. The beam selection system is a tunable optical component used to achieve monochromaticity of incident X-rays.

[0010] The enhanced feature spectrum and the excitation condition parameters are processed collaboratively. By establishing a dual-channel data interaction architecture, the spectral features and excitation conditions are cross-correlated and weighted to form a multi-dimensional analysis feature that adapts to the excitation conditions.

[0011] Based on the multidimensional analysis features, a hierarchical mapping relationship between spectral features and elemental chemical states is established through hierarchical feature space transformation and recognition model adaptation, thereby achieving high-sensitivity identification of elemental chemical states in the sample to be tested.

[0012] Optionally, the step of co-processing the enhanced feature spectrum and the excitation condition parameters, and establishing a dual-channel data interaction architecture to cross-correlate and weight the spectral features and excitation conditions to form an adaptive multidimensional analysis feature for the excitation conditions, includes:

[0013] A dual-channel data interaction architecture is established, wherein the first channel is used to transmit spectral feature data in the enhanced feature spectrum, and the second channel is used to transmit excitation setting data in the excitation condition parameters.

[0014] Cross-correlation is performed on the spectral feature data and the excitation setting data in the dual-channel data interaction architecture to establish the correspondence between spectral features and excitation conditions;

[0015] Based on the aforementioned correspondence, the cross-correlated data is dynamically weighted according to a preset weight allocation strategy;

[0016] The weighted data are then fused to form the adaptive multidimensional analysis features of the aforementioned activation conditions.

[0017] Optionally, the step of dynamically weighting the cross-correlated data based on the correspondence and according to a preset weight allocation strategy includes:

[0018] Based on the correlation characteristics between the spectral feature data and the excitation setting data in the correspondence, the correlation strength score is calculated;

[0019] Based on the degree of matching between the spectral features and the excitation conditions in the aforementioned correspondence, a matching weight factor is determined;

[0020] Based on the stability performance of spectral characteristics under different excitation conditions in the aforementioned correspondence, a stability weighting factor is calculated;

[0021] Based on a preset weight allocation strategy, the correlation strength score, matching weight factor, and stable weight factor are combined and calculated to generate a dynamic weight coefficient for each data combination.

[0022] A feedback correlation mechanism is established between the weight coefficients and the corresponding relationship, and the dynamic weight coefficients are adaptively adjusted based on the real-time analysis results.

[0023] Optionally, the step of establishing a hierarchical mapping relationship between spectral features and elemental chemical states based on the multidimensional analysis features, through hierarchical feature space transformation and recognition model adaptation, to achieve high-sensitivity identification of elemental chemical states in the sample to be tested, includes:

[0024] The multidimensional analysis features are input into the hierarchical processing module. In the first layer feature space transformation, the multidimensional analysis features are converted into element category features. In the second layer feature space transformation, the element category features are further converted into chemical state features.

[0025] By identifying model adaptation, a hierarchical mapping relationship is established from the chemical state characteristics to the elemental chemical state;

[0026] Based on the hierarchical mapping relationship, the converted chemical state features are matched with the element chemical state database;

[0027] Based on the feature matching results, the final identification result of the chemical state of the elements in the sample to be tested is output.

[0028] Optionally, adjusting the excitation parameters of the beam filtering system based on the element chemical state markers in the enhanced characteristic spectrum, and optimizing them through a feedback control mechanism to generate excitation condition parameters of a single wavelength matching the current element chemical state, includes:

[0029] Based on the elemental chemical state markers in the enhanced feature spectrum, the target excitation parameters are determined;

[0030] The beam selection system is initially adjusted according to the target excitation parameters;

[0031] The actual output of the beam selection system is monitored, and the actual output is compared with the target excitation parameters to generate a feedback adjustment signal;

[0032] The beam selection system is adjusted in a closed loop based on the feedback adjustment signal until the actual output and the target excitation parameters reach a stable matching state.

[0033] Record the crystal reflection conditions and incident angle when a stable matching state is achieved, as the excitation condition parameters.

[0034] Optionally, the step of performing closed-loop adjustment of the beam selection system based on the feedback adjustment signal until the actual output and the target excitation parameters reach a stable matching state includes:

[0035] Based on the feedback adjustment signal, the coordinated adjustment amount of the crystal incident angle and reflection conditions is calculated, and the coordinated adjustment amount is converted into a control command for the crystal driving mechanism to synchronously fine-tune the crystal incident angle and reflection conditions.

[0036] After each fine-tuning, the actual output X-ray wavelength of the beam filtering system is re-monitored;

[0037] The re-monitored X-ray wavelength is compared with the target excitation wavelength to update the feedback adjustment signal;

[0038] The process of adjusting the crystal incident angle and reflection conditions is iteratively executed until the X-ray wavelength and the target excitation parameters reach a stable matching state.

[0039] Optionally, the step of performing matching analysis based on the initial spectral information and the elemental characteristic spectral library, and generating an enhanced characteristic spectrum containing elemental chemical state labels by analyzing the chemical shift effect characterized by spectral peak shifts and intensity changes, includes:

[0040] Analyze the characteristic changes of the initial spectral information relative to the reference spectrum in the elemental feature map library, including spectral line peak shifts and intensity changes.

[0041] Based on the aforementioned feature changes, a preset chemical state feature pattern is matched to determine the elemental chemical state label of the target element in the sample to be tested;

[0042] The element chemical state label is associated with the initial spectral information to generate the enhanced feature spectrum.

[0043] Secondly, this application provides a high-sensitivity detection system for an energy-dispersive X-ray fluorescence spectrometer, comprising:

[0044] The acquisition module is used to acquire energy dispersive X-ray fluorescence spectral data of the sample to be tested, and obtain initial spectral information including the photon number distribution.

[0045] The generation module is used to perform matching analysis between the initial spectral information and the elemental feature map library, and generate an enhanced feature spectrum containing elemental chemical state labels by analyzing the chemical shift effect characterized by spectral line peak shift and intensity change characteristics.

[0046] The filtering module is used to adjust the excitation parameters of the beam filtering system according to the element chemical state markings in the enhanced feature spectrum, and to optimize them through a feedback control mechanism to generate excitation condition parameters of a single wavelength that match the current element chemical state. The beam filtering system is a tunable optical component for achieving monochromaticity of incident X-rays.

[0047] The forming module is used to collaboratively process the enhanced feature spectrum and the excitation condition parameters. By establishing a dual-channel data interaction architecture, the spectral features and excitation conditions are cross-correlated and weighted to form multi-dimensional analysis features that are adaptive to the excitation conditions.

[0048] The identification module is used to establish a hierarchical mapping relationship between spectral features and elemental chemical states based on the multidimensional analysis features, through hierarchical feature space transformation and identification model adaptation, and to complete the high-sensitivity identification of elemental chemical states in the sample to be tested.

[0049] Thirdly, this application provides an electronic device, comprising:

[0050] Memory, used to store computer programs;

[0051] A processor, configured to execute the computer program to implement the steps of the high-sensitivity detection method for an energy-dispersive X-ray fluorescence spectrometer as described in the first aspect above.

[0052] Fourthly, this application provides a computer-readable storage medium storing a computer program that, when executed by a processor, can implement the steps of the high-sensitivity detection method for energy-dispersive X-ray fluorescence spectrometer as described in the first aspect above.

[0053] The high-sensitivity detection method for energy-dispersive X-ray fluorescence spectrometers provided in this application acquires the initial spectral information of the sample to be tested, combines it with an elemental characteristic spectral library to analyze chemical shift effects and generate an enhanced characteristic spectrum with elemental chemical state markings, accurately capturing key changes in spectral lines to clarify the basic characteristics of elemental chemical states. Based on this marking, the excitation parameters of the beam selection system are optimized through feedback control to generate single-wavelength excitation conditions adapted to the current chemical state, solving the problem of poor adaptability of fixed excitation parameters. Subsequently, a dual-channel architecture is used to achieve the synergistic fusion of the enhanced characteristic spectrum and excitation condition parameters to form multi-dimensional analytical features. At the same time, a hierarchical mapping relationship is established through hierarchical feature space transformation and recognition model adaptation, effectively avoiding spectral library matching bias and making up for the lack of dynamic synergistic optimization in existing schemes. Ultimately, this significantly improves the sensitivity and accuracy of elemental chemical state identification in scenarios with unknown and complex samples and no standards, fully meeting the practical needs of low-content element analysis and anti-interference.

[0054] Furthermore, a dual-channel data interaction architecture is established. The first channel transmits spectral feature data for enhanced characteristic spectra, while the second channel transmits excitation setting data for excitation condition parameters. First, the data from the two channels are cross-correlated to establish the correspondence between spectral features and excitation conditions. Then, the correlated data is dynamically weighted according to a preset weighting strategy. Finally, the weighted data is fused to form a multidimensional analysis feature that adapts to the excitation conditions. This scheme achieves precise separation and transmission of spectral feature data and excitation setting data through a dual-channel architecture. Cross-correlation clarifies the correspondence between the two, and dynamic weighting adapts to the importance of different data. The multidimensional analysis feature formed by data fusion is both targeted and adaptable, effectively improving the accuracy and efficiency of data processing. This provides high-quality data support for subsequently establishing a hierarchical mapping relationship between spectral features and elemental chemical states, and achieving high-sensitivity identification. Attached Figure Description

[0055] To more clearly illustrate the technical solutions of the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0056] Figure 1 A schematic flowchart illustrating a high-sensitivity detection method for an energy-dispersive X-ray fluorescence spectrometer provided in this application embodiment;

[0057] Figure 2 A flowchart illustrating a specific embodiment of a high-sensitivity detection method for an energy-dispersive X-ray fluorescence spectrometer provided in this application;

[0058] Figure 3 A schematic diagram of a specific embodiment of a high-sensitivity detection method for an energy-dispersive X-ray fluorescence spectrometer provided in this application;

[0059] Figure 4 This is a schematic diagram of the structure of a high-sensitivity detection system for an energy-dispersive X-ray fluorescence spectrometer provided in an embodiment of this application. Detailed Implementation

[0060] In standard-free analysis scenarios involving unknown and complex sample elements with varying chemical states, existing analytical schemes based on matching preset standard spectral libraries with fixed excitation parameters have significant shortcomings: fixed excitation parameters cannot adapt to the differences in spectral characteristics of different element chemical states, making it difficult to cope with spectral line changes caused by chemical shifts, resulting in low excitation efficiency; preset spectral library matching is easily affected by spectral line distortion caused by chemical shifts, leading to matching deviations; at the same time, there is a lack of a dynamic synergistic optimization mechanism between spectral characteristics and excitation conditions, resulting in insufficient sensitivity when dealing with low-content elements, failing to meet the actual needs of high-precision analysis.

[0061] To address the aforementioned issues, this application proposes a high-sensitivity detection method for energy-dispersive X-ray fluorescence spectrometers. This method first acquires sample spectral data and analyzes the chemical shift effect to generate enhanced feature spectra with elemental chemical state labels. Then, based on these labels, the excitation parameters of the beam filtering system are optimized through feedback control. Subsequently, a dual-channel architecture is used to achieve the synergistic fusion of spectral features and excitation condition parameters to form multi-dimensional analytical features. Finally, high-sensitivity identification is achieved through hierarchical mapping modeling. This scheme adapts dynamically adjustable excitation parameters to the spectral features of different chemical states, improves spectral line matching accuracy by analyzing the chemical shift effect, and enhances data processing performance through a dual-channel synergistic optimization mechanism. It fundamentally solves the shortcomings of existing schemes, such as poor adaptability, large matching deviations, and insufficient sensitivity, significantly improving the accuracy and reliability of elemental chemical state identification in scenarios with unknown and complex samples and no standards.

[0062] To enable those skilled in the art to better understand the present application, the present application will be further described in detail below with reference to the accompanying drawings and specific embodiments. Obviously, the described embodiments are merely some embodiments of the present application, and not all embodiments. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0063] The core of this application is to provide a high-sensitivity detection method for energy-dispersive X-ray fluorescence spectrometers, and a flowchart of one specific implementation is shown below. Figure 1 As shown, the method includes:

[0064] S101. Collect energy dispersive X-ray fluorescence spectral data of the sample to be tested to obtain initial spectral information including photon number distribution;

[0065] In the above scheme, energy-dispersive X-ray fluorescence spectroscopy data refers to the relevant data obtained by detecting the fluorescent X-rays generated after the sample is excited by X-rays, including information such as the number of photons corresponding to different energies. Initial spectral information is a set of information that reflects the distribution of photon quantity with energy, formed after preliminary processing of the above data. Photon quantity distribution refers to the specific arrangement of photons within different energy ranges.

[0066] In this embodiment, the X-ray emitting device is first started and stable tube voltage and tube current parameters are set to emit a uniform X-ray beam to the unknown sample. For example, an X-ray beam with a tube voltage of 50kV and a tube current of 1mA is emitted to a solid sample containing unknown elements. The X-ray photons collide with the atoms inside the sample, causing the inner shell electrons of the atoms to be knocked out to form vacancies, thereby exciting various elements in the sample.

[0067] Secondly, an energy-resolved detector (such as a silicon drift detector) is used to receive the fluorescence X-ray signal excited from the sample. The sensitive layer of the detector captures fluorescence X-rays of different energies emitted by the sample. When a photon strikes the sensitive layer, it generates a charge pulse. The detector converts these charge pulses into a processable electrical signal that is proportional to the photon energy and records the arrival time of the signal simultaneously. For example, the detector can distinguish between 1 keV and 1.1 keV fluorescence X-ray signals and convert them into electrical signals of corresponding amplitudes.

[0068] Next, based on the energy resolution of the detector, the energy region division rules are preset, and the fluorescence X-ray signal is divided into multiple continuous energy intervals (i.e., energy regions) according to the energy value. When dividing, it is necessary to balance energy resolution and statistical efficiency. For example, if the energy resolution of the detector is 0.1 keV, the energy range of 0-10 keV can be divided into 100 continuous energy regions, each with an energy range of 0.1 keV. This can accurately distinguish photons of different energies, while ensuring that each energy region has enough signal for statistics.

[0069] Then, the received fluorescence X-ray electrical signals in each energy region are processed by the data processing module. First, the electrical signals are amplified and filtered to eliminate noise interference. Then, the energy region to which the signal belongs is determined based on the signal amplitude. Then, the number of effective photons in each energy region is counted. For example, for the energy region with an interval of 0.1 keV, the electrical signals with amplitudes corresponding to 0-0.1 keV are counted. After eliminating noise signals with abnormal amplitudes, 45 effective photons are counted in this energy region. Similarly, 32 effective photons are counted in the 0.1-0.2 keV energy region.

[0070] Finally, all energy regions are arranged in order of increasing energy. The central energy value of each energy region is used as the horizontal axis and the effective photon count of the corresponding energy region is used as the vertical axis. A continuous spectral curve is generated through the data visualization module to form initial spectral information containing the distribution of photon count with energy. For example, the horizontal axis is labeled with 0.05keV, 0.15keV...9.95keV (central values ​​of each energy region), and the vertical axis is labeled with the corresponding photon count. The final result is a fluctuating spectral graph, which serves as the initial spectral information.

[0071] In a practical application, Research Institute A needed to analyze an unknown mixed non-metallic material sample. Staff operated the laboratory's X-ray emission device to emit an X-ray beam towards the sample. Then, an energy-resolution detector was activated to receive the fluorescence X-ray signal generated after the sample was excited. Based on the detector's energy resolution parameters, the received fluorescence X-ray signal was divided into multiple continuous energy regions with 0.08 keV intervals. A data processing module then counted the photons in each energy region, determining the number of photons corresponding to each region. Finally, all energy regions were arranged in ascending order of energy, successfully obtaining the initial spectral information of the mixed non-metallic material sample.

[0072] S102. Based on the initial spectral information and the elemental feature spectral library, a matching analysis is performed. By analyzing the chemical shift effect characterized by the peak position shift and intensity change of spectral lines, an enhanced feature spectrum containing elemental chemical state labels is generated.

[0073] Optionally, step S102 may specifically include the following steps:

[0074] S1021. Analyze the characteristic changes of the initial spectral information relative to the reference spectrum in the elemental characteristic spectral library, the characteristic changes including spectral line peak position shifts and intensity changes;

[0075] S1022. Based on the aforementioned feature changes, a preset chemical state feature pattern is matched to determine the elemental chemical state label of the target element in the sample to be tested;

[0076] S1023. Associate the element chemical state label with the initial spectral information to generate the enhanced feature spectrum.

[0077] In the above scheme, the elemental characteristic spectral library is a database storing standard spectral information for various known elements, including reference data such as the position and intensity of characteristic spectral lines corresponding to different elements and their chemical states. Peak position shift refers to the shift in the position of characteristic spectral lines in the initial spectrum relative to the reference spectrum. Intensity variation refers to the increase or decrease in the height or area of ​​characteristic spectral lines in the initial spectrum compared to the reference spectrum. Chemical shift effect is the phenomenon where changes in the peak position and intensity of characteristic spectral lines occur due to changes in the element's chemical environment, and can be characterized by peak position shift and intensity variation. Element chemical state labeling is tagging information used to identify the specific chemical form of the target element in the sample. Enhanced characteristic spectrum is spectral information that makes the element's chemical state characteristics more prominent in the spectrum by associating the element chemical state labeling with the initial spectral information.

[0078] In this embodiment, firstly, in step S1021, reference spectra that may be related to the initial spectrum are screened from the elemental characteristic spectral library. The screening criterion is the approximate energy range of the main characteristic peaks in the initial spectrum. A spectral alignment algorithm (such as the correlation coefficient method) is used to align the initial spectrum with each reference spectrum to ensure that their energy axes are consistent. Then, the characteristic spectral lines in each reference spectrum are located, and the corresponding spectral line positions are found in the initial spectrum. The peak position shift is calculated. For example, if the energy of a characteristic peak in the reference spectrum is 5 keV, and the corresponding peak energy in the initial spectrum is 5.02 keV, the peak position shift is 0.02 keV. Simultaneously, by measuring the peak height or peak area of ​​the corresponding characteristic peaks in the initial and reference spectra, the intensity increase / decrease ratio, i.e., the intensity change characteristic, is obtained. For example, if the peak height of the reference spectrum is 100, and the peak height of the initial spectrum is 80, the intensity change is a decrease of 20%.

[0079] Secondly, in step S1022, based on the element to which the characteristic change belongs as determined in step S1021, the characteristic mode parameters corresponding to all chemical states of the element are retrieved from the preset chemical state characteristic mode library. Next, the peak position shift and intensity change ratio obtained in step S1021 are compared one by one with the retrieved chemical state characteristic mode parameters to calculate the similarity between the two. The similarity calculation can be performed using methods such as Euclidean distance, and the smaller the distance, the higher the similarity. The chemical state corresponding to the chemical state characteristic mode with the highest similarity is selected as the chemical state of the target element, and the corresponding element chemical state label is generated.

[0080] Finally, step S1023 determines the specific position of the characteristic spectral line corresponding to the element's chemical state label in the initial spectrum. This position is the same as the spectral line position corresponding to the characteristic change analyzed in step S1021. Using data annotation technology, chemical state label text is added to this spectral line position in the initial spectrum. The annotated spectral data is then integrated to ensure a one-to-one correspondence between the label information and the corresponding energy and photon quantity data, preventing misalignment. Finally, an enhanced characteristic spectrum file is generated. This file contains complete data from the initial spectrum and clear chemical state labels at key spectral lines, and can be directly used for subsequent processing steps.

[0081] In practical applications, when analyzing an unknown metal oxide sample, Detection Center B first obtained the initial spectral information of the sample, and then matched it with the laboratory's elemental characteristic spectral library. It was found that a certain characteristic spectral line in the initial spectrum shifted to a higher energy end by 0.03 keV relative to the reference spectral peak of a certain metal element in the library, and the intensity decreased by 15%. Next, these characteristic changes were matched with the preset chemical state characteristic patterns of the metal element, and it was found that the characteristic pattern corresponding to the +3 valence chemical state of the metal was consistent, thus determining that the chemical state label of the target element was "a certain metal +3 valence". Finally, this chemical state label was associated with the corresponding characteristic spectral line of the initial spectrum, successfully generating an enhanced characteristic spectrum containing the element's chemical state label.

[0082] The overall scheme of S102 described above, through matching analysis of the initial spectrum and the elemental characteristic spectral library, accurately captured the chemical shift effect characterized by the shift of spectral peak positions and intensity changes, successfully determined the chemical state marker of the target element in the sample to be tested and generated an enhanced characteristic spectrum. This not only highlighted and clarified the chemical state characteristics of the element in the spectral information, providing a precise chemical state basis for subsequent adjustment of the excitation parameters of the beam screening system, but also made up for the deficiency of the initial spectrum, which only contains photon number distribution and lacks chemical state information. This has promoted the entire detection process from simple spectral data acquisition to elemental chemical state identification, taking a key step forward.

[0083] S103. Based on the element chemical state markings in the enhanced characteristic spectrum, adjust the excitation parameters of the beam selection system and optimize them through a feedback control mechanism to generate excitation condition parameters of a single wavelength that match the current element chemical state. The beam selection system is a tunable optical component used to achieve monochromaticity of incident X-rays.

[0084] Optionally, step S103 may specifically include the following steps:

[0085] S1031. Determine the target excitation parameters based on the element chemical state markers in the enhanced feature spectrum;

[0086] S1032. Perform initial adjustment of the beam selection system according to the target excitation parameters;

[0087] S1033. Monitor the actual output of the beam filtering system and compare the actual output with the target excitation parameters to generate a feedback adjustment signal;

[0088] S1034. Based on the feedback adjustment signal, the beam selection system is adjusted in a closed loop until the actual output and the target excitation parameters reach a stable matching state.

[0089] S1035. Record the crystal reflection conditions and incident angle when a stable matching state is achieved, as the excitation condition parameters.

[0090] Specifically, step S1034 includes the following process: calculating the coordinated adjustment amount of the crystal incident angle and reflection conditions based on the feedback adjustment signal; converting the coordinated adjustment amount into a control command for the crystal driving mechanism to synchronously fine-tune the crystal incident angle and reflection conditions; re-monitoring the actual output X-ray wavelength of the beam filtering system after each fine-tuning; comparing the re-monitored X-ray wavelength with the target excitation wavelength to update the feedback adjustment signal; iteratively executing the adjustment process of the crystal incident angle and reflection conditions until the X-ray wavelength and the target excitation parameters reach a stable matching state.

[0091] In the above scheme, the beam sorting system is a tunable optical component used to achieve monochromaticity of incident X-rays, capable of sorting out X-rays of specific wavelengths. Excitation condition parameters are stable beam sorting system settings that match the current chemical state of the element, including crystal reflection conditions and incident angle. Target excitation parameters are the desired excitation parameter values ​​achieved by the beam sorting system, determined based on the element's chemical state label. Actual output is the actual X-ray characteristic data generated after the beam sorting system is adjusted. Feedback adjustment signal is a signal generated based on the comparison between the actual output and the target excitation parameters, used to adjust the system. Crystal reflection conditions are the relevant state settings of the crystal in the beam sorting system for reflecting X-rays. Incident angle is the angle at which X-rays are incident on the crystal of the beam sorting system. Coordinated adjustment amount is the adjustment amount required to simultaneously adjust the crystal's incident angle and reflection conditions. Crystal driving mechanism is a mechanical structure used to drive the crystal to change the incident angle and reflection conditions.

[0092] In this embodiment, firstly, a clear element chemical state marker is extracted from the enhanced feature spectrum in step S1031; a preset database of element chemical states and excitation parameters is called to retrieve the corresponding excitation parameter information; and the target excitation parameter is determined based on the retrieval results. For example, if the element chemical state marker in the enhanced feature spectrum is "a certain element with a +2 valence", the target excitation wavelength corresponding to the +2 valence of the element is determined to be 8 keV based on the preset correspondence between element chemical states and excitation parameters. This is the target excitation parameter.

[0093] Secondly, the determined target excitation parameters are obtained through step S1032. The optical characteristics manual or database of the beam selection system is consulted to understand the approximate setting range of the crystal incident angle and reflection conditions at the target excitation wavelength. Based on the optical characteristics information obtained, the beam selection system is initially adjusted by setting the crystal incident angle to an initial angle close to the theoretical value and the reflection conditions to the corresponding initial state, thus completing the initial adjustment of the system.

[0094] Next, in step S1033, the wavelength monitoring device is activated and aligned with the output of the beam filtering system to collect the wavelength data of the system's output X-rays in real time. The deviation between the collected actual output wavelength data and the previously determined target excitation wavelength is calculated. For example, if the actual measured wavelength is 7.03 keV and the target wavelength is 7 keV, the calculated deviation is +0.03 keV. Finally, based on the magnitude and direction of the deviation, a feedback adjustment signal is generated according to a preset signal generation rule. This signal contains information about the direction and approximate magnitude of the adjustment.

[0095] Then, in step S1034, a feedback adjustment signal is received. Based on this signal, a parameter adjustment algorithm is used to calculate the coordinated adjustment amount of the crystal incident angle and reflection conditions. The calculated coordinated adjustment amount is converted into a control command that the crystal driving mechanism can recognize. Then, the crystal driving mechanism is driven to synchronously fine-tune the crystal incident angle and reflection conditions according to the control command. After the fine-tuning is completed, the wavelength monitoring device is restarted to monitor the actual output X-ray wavelength of the monitoring system. The re-monitored wavelength is compared with the target excitation wavelength, and the deviation is calculated. If the deviation still exceeds the allowable range, the feedback adjustment signal is updated. The above process of calculating the coordinated adjustment amount, converting the control command, fine-tuning, monitoring, comparing, and updating the signal is iteratively executed until the monitored X-ray wavelength is stable within the allowable error range of the target excitation wavelength, achieving a stable matching state.

[0096] Specifically, during the closed-loop adjustment process, Laboratory A received a feedback adjustment signal "wavelength needs to be reduced by 0.04 keV". The algorithm calculated that the crystal incident angle needed to be reduced by 0.6°, and the reflection condition needed to be adjusted by ΔR1. After converting these adjustment amounts into control commands for the drive mechanism, the drive mechanism performed synchronous fine-tuning of the crystal. After fine-tuning, the measured wavelength was 6.82 keV, the target wavelength was 6.8 keV, and the deviation was 0.02 keV, indicating further adjustment was needed. Therefore, the feedback adjustment signal was updated to "wavelength needs to be reduced by 0.02 keV". The coordinated adjustment amount was recalculated, the incident angle was reduced by 0.3°, and the reflection condition was adjusted by ΔR2. After fine-tuning, the measured wavelength was 6.80 keV, reaching a stable matching state, and the iterative adjustment stopped. Here, ΔR1 and ΔR2 both refer to the fine-tuning amounts of the crystal reflection condition.

[0097] Finally, S1035 confirms that the actual output of the beam selection system has reached a stable state matching the target excitation parameters, meaning the wavelength measured by the wavelength monitoring device is within the allowable error range and remains stable. The parameter recording module is then activated. This module, connected to the control unit of the beam selection system, can read the specific values ​​of the crystal reflection conditions and incident angle at this time. Finally, the parameter recording module stores the read crystal reflection conditions and incident angle, marking them as excitation condition parameters matching the current element's chemical state, thus completing the recording operation.

[0098] In practical applications, when an analytical institution was conducting elemental chemical state detection on an unknown ore sample, it first extracted the chemical state marker of "element D+2 valence" from the enhanced characteristic spectrum and used the corresponding relational database to determine the target excitation wavelength as 7.1 keV. Then, based on the target wavelength and the system's optical characteristics, the crystal incident angle was initially adjusted to 26°, and the reflection condition was set to mode E. Next, the actual output wavelength was measured to be 7.15 keV using a wavelength monitoring device, deviating from the target wavelength by 0.05 keV, generating a feedback adjustment signal. Then, through iterative closed-loop adjustment, the crystal incident angle and reflection condition were fine-tuned twice, ultimately stabilizing the actual output wavelength at 7.1 keV. Finally, the crystal incident angle was recorded as 26.3°, and the reflection condition as mode E, as the excitation condition parameters.

[0099] The overall scheme of S103 described above, based on the element chemical state markers in the enhanced characteristic spectrum, successfully generated excitation condition parameters matching the current element chemical state through a series of adjustments and optimizations, solving the problem of poor adaptability of fixed excitation parameters in existing schemes. Through closed-loop adjustment of the feedback control mechanism, the accuracy and stability of the excitation condition parameters are ensured, enabling the beam selection system to output single-wavelength X-rays adapted to the element chemical state. This provides high-quality parameter support for the subsequent collaborative processing of the enhanced characteristic spectrum and excitation condition parameters, further improving the sensitivity and accuracy of the entire detection method for element chemical state identification.

[0100] S104. The enhanced feature spectrum and the excitation condition parameters are processed collaboratively. By establishing a dual-channel data interaction architecture, the spectral features and excitation conditions are cross-correlated and weighted to form a multi-dimensional analysis feature that adapts to the excitation conditions.

[0101] Optionally, step S104 may specifically include the following steps:

[0102] S1041. Establish a dual-channel data interaction architecture, wherein the first channel is used to transmit spectral feature data in the enhanced feature spectrum, and the second channel is used to transmit excitation setting data in the excitation condition parameters.

[0103] S1042. Cross-correlate the spectral feature data and the excitation setting data in the dual-channel data interaction architecture to establish the correspondence between spectral features and excitation conditions.

[0104] S1043. Based on the correspondence, the cross-correlated data is dynamically weighted according to a preset weight allocation strategy.

[0105] S1044. The weighted data are fused to form the adaptive multidimensional analysis features of the excitation conditions.

[0106] Specifically, step S1043 includes the following processes: calculating the correlation strength score based on the correlation characteristics between the spectral feature data and the excitation setting data in the correspondence; determining the matching weight factor based on the degree of matching between the spectral features and the excitation conditions in the correspondence; calculating the stability weight factor based on the stability performance of the spectral features under different excitation conditions in the correspondence; performing a composite calculation of the correlation strength score, the matching weight factor, and the stability weight factor according to a preset weight allocation strategy to generate a dynamic weight coefficient for each data combination; establishing a feedback correlation mechanism between the weight coefficient and the correspondence, and adaptively adjusting the dynamic weight coefficient based on the real-time analysis results.

[0107] In the above scheme, the dual-channel data interaction architecture is an interactive structure composed of two independent data transmission channels, which transmit different types of data respectively. Spectral feature data includes key data such as peak positions and intensities of spectral lines reflecting the chemical states of elements in the enhanced feature spectrum. Excitation setting data includes system setting information such as crystal reflection conditions and incident angles in the excitation condition parameters. The preset weight allocation strategy is a pre-defined rule used to determine the weights of various data types. The correlation strength score is a quantitative value that measures the degree of correlation between spectral feature data and excitation setting data. The matching weight factor is a weight coefficient determined based on the degree of matching between spectral features and excitation conditions. The stability weight factor is a weight coefficient determined based on the stability performance of spectral features under different excitation conditions. The dynamic weight coefficient is a data weight value that can be adjusted in real time, calculated by integrating multiple factors.

[0108] like Figure 2 As shown in the embodiment of this application, a dual-channel data interaction architecture is first established through step S1041. In the dual-channel data interaction architecture, the first data channel is used to transmit the spectral feature data in the enhanced feature spectrum, and the second data channel is used to transmit the excitation setting data in the excitation condition parameters.

[0109] Secondly, in step S1042, the spectral feature data in the dual-channel data interaction architecture is cross-correlated with the spectral feature data in the first data channel and the excitation setting data in the second data channel to establish the correspondence between spectral features and excitation conditions. For example, a data association algorithm is used to match the spectral peak position of the first data channel (5.02keV, intensity 80, etc.) with the incident angle of the second data channel (25°, reflection condition mode C, etc.) to clarify the correspondence between "spectral peak position 5.02keV - incident angle 25°" and "intensity 80 - reflection condition mode C".

[0110] Next, in step S1043, based on the correspondence, the cross-correlated data is dynamically weighted according to a preset weight allocation strategy. Specifically, dynamic weight coefficients are first assigned to the cross-correlated data according to the degree of correlation between the data in the correspondence. These weight coefficients are adjusted based on the importance of the spectral feature data and the matching degree of the excitation setting data. Specifically, the importance of the spectral feature data in the correspondence (e.g., the criticality of spectral peak position for chemical state identification) and the matching degree of the excitation setting data (e.g., the fit between the incident angle and the peak position) are considered, and the changes in both are combined. The correlation strength score is calculated based on correlation characteristics such as trend consistency and data fluctuation synchronization. For example, "spectral peak position 5.02keV - incident angle 25°" is highly important and closely correlated due to its good matching degree. Then, the matching weight factor is determined according to the matching degree. For example, if the matching degree is high, a weight factor of 0.8 is used. Then, the stability weight factor is calculated based on the stability of spectral characteristics. For example, if the spectral peak position is stable under different excitation conditions, a weight factor of 0.7 is used. Finally, the three factors are combined according to the preset strategy to calculate the dynamic weight coefficient, such as (8 / 10) × 0.8 × 0.7 = 0.448, and a feedback mechanism is established to adjust the coefficient according to the real-time effect.

[0111] Finally, the weighted data is fused in S1044 to form an excitation condition adaptive multidimensional analysis feature spectral feature data and excitation setting data. These are then fused to form a unified multidimensional data set. For example, the cross-correlation data and their dynamic weight coefficients are integrated to form a unified data set containing multidimensional information such as spectral peak position, incident angle, and corresponding weight. The multidimensional data set is then adaptively adjusted to reflect the feature changes under different excitation conditions, ultimately forming the excitation condition adaptive multidimensional analysis feature set.

[0112] In practical applications, an analytical institution establishes a dual-channel data interaction architecture when analyzing unknown alloy samples. The first data channel transmits spectral characteristic data such as "peak position 5.8 keV, intensity 95, and marked metal E+2 valence," while the second data channel transmits excitation setting data matching this chemical state, such as "incident angle 24° and reflection condition mode F." Next, using an S1042 data association algorithm, the two channels are cross-correlated to establish a correspondence between "peak position 5.8 keV - incident angle 24°" and "intensity 95 - reflection condition mode F." Then, dynamic weighting coefficients are assigned based on this correspondence. According to the importance of spectral characteristics and the matching degree of excitation settings, a correlation strength score of 8, a matching weight factor of 0.8, and a stability weight factor of 0.7 are calculated. A coefficient of 0.448 is obtained by combining these coefficients according to a preset strategy, and then adjusted to 0.48 through a feedback mechanism. Finally, the weighted cross-correlated data is integrated to form a unified multidimensional dataset, which is adaptively adjusted to reflect characteristic changes under different excitation conditions, ultimately forming an adaptive multidimensional analytical feature for excitation conditions.

[0113] The overall scheme of S104 described above achieves independent transmission and cross-correlation of two types of data through a dual-channel data interaction architecture. Combined with dynamic weighting processing and data fusion, it successfully forms multidimensional analysis features that are adaptive to excitation conditions. This process not only integrates key information from spectral features and excitation conditions, but also ensures the prominent role of important data through dynamic weighting. It solves the problem of fragmented processing of spectral data and excitation parameters in existing schemes, making the analysis features more comprehensive and accurate in reflecting the correlation between elemental chemical states and excitation conditions, and providing strong multidimensional data support for subsequent high-sensitivity identification of elemental chemical states.

[0114] S105. Based on the multidimensional analysis features, a hierarchical mapping relationship between spectral features and elemental chemical states is established through hierarchical feature space transformation and identification model adaptation, thereby achieving high-sensitivity identification of elemental chemical states in the sample to be tested.

[0115] Optionally, step S105 may specifically include the following steps:

[0116] S1051. Input the multidimensional analysis features into the hierarchical processing module, convert the multidimensional analysis features into element category features in the first layer feature space transformation, and further convert the element category features into chemical state features in the second layer feature space transformation.

[0117] S1052. By identifying model adaptation, establish a hierarchical mapping relationship from the chemical state characteristics to the element chemical state;

[0118] S1053. Based on the hierarchical mapping relationship, the converted chemical state features are matched with the element chemical state database;

[0119] S1054. Based on the feature matching results, output the final identification result of the chemical state of the elements in the sample to be tested.

[0120] In the above scheme, hierarchical feature space transformation is a process of gradually extracting more targeted features from multidimensional analysis features through multi-level transformations. The hierarchical mapping relationship is a correspondence established in stages from feature data to element chemical states. The hierarchical processing module is a functional module used to perform hierarchical feature space transformation on multidimensional analysis features. Element category features are feature data that can distinguish different element categories. Chemical state features are feature data that can reflect the specific chemical existence form of an element. The element chemical state database is a database that stores various chemical states of known elements and their corresponding feature data.

[0121] In this embodiment, the multidimensional analysis features are first input into the hierarchical processing module in step S1051. In the first layer feature space transformation, a feature extraction algorithm (such as principal component analysis algorithm) is used to extract key information that can distinguish element categories from the multidimensional analysis features and convert it into element category features. For example, the element category features corresponding to "metal A" are extracted from the multidimensional features containing information such as peak position and incident angle. In the second layer feature space transformation, a feature refinement algorithm (such as linear discriminant analysis algorithm) is used to further convert the element category features into chemical state features that can reflect the chemical state. For example, the element category features of "metal A" are converted into the chemical state features of "metal A+2".

[0122] Secondly, through step S1052, a hierarchical mapping relationship from chemical state features to element chemical states is established by identifying the model. First, a suitable model for chemical state identification (such as a support vector machine model) is selected. Then, the converted chemical state features are adapted and adjusted to match the input format of the model. Finally, a corresponding mapping between chemical state features and specific element chemical states is established in the model. For example, a mapping relationship of "metal A+2 valence chemical state features - metal A+2 valence" is established.

[0123] Next, based on the hierarchical mapping relationship in step S1053, the converted chemical state features are matched with the element chemical state database. The standard feature data of various element chemical states in the element chemical state database are called, and a similarity matching algorithm (such as Euclidean distance matching algorithm) is used to compare the converted chemical state features with the standard feature data and calculate the similarity value. For example, the "metal A+2 valence chemical state feature" is compared with the "metal A+2 valence standard feature" in the database, and the similarity is over 90%.

[0124] Finally, based on the feature matching results, step S1054 outputs the final identification result of the chemical state of the element in the sample to be tested. A similarity threshold is set. If the matching similarity is higher than the threshold, it is determined that the corresponding chemical state of the element exists in the sample to be tested, and the identification result is output. For example, because the similarity is more than 90%, the final identification result "the sample to be tested contains metal A+2 valence" is output.

[0125] In practical applications, a testing agency, during the detection of an unknown alloy sample, inputs multidimensional analytical features including a peak position of 6.3 keV, an intensity of 102, and an incident angle of 27° into a hierarchical processing module. The first layer extracts the principal components that distinguish metal I from other metals using principal component analysis, obtaining the element category feature of "metal I". The second layer is refined and converted into the chemical state feature of "metal I + 2 valence" using linear discriminant analysis. Subsequently, a support vector machine model is selected, and the chemical state feature is adjusted to the model input format and paired with the "metal I + 2 valence" label for training, establishing a hierarchical mapping relationship between the two. Then, based on the hierarchical mapping, the standard features of metal I in the element chemical state database are called, and the similarity with the +2 valence standard feature is calculated to be 93% using the Euclidean distance matching algorithm. This similarity is compared with a preset threshold of 85%. If it is higher than the threshold, the final identification result "the sample contains metal I + 2 valence" is output.

[0126] The overall scheme of S105 described above accurately extracts chemical state features through hierarchical feature space transformation, establishes a hierarchical mapping relationship by combining recognition model adaptation, and then achieves high-sensitivity identification of the chemical state of elements in the sample under test through database feature matching and threshold judgment. The entire process is progressive, ensuring the accuracy of feature extraction and improving the accuracy of recognition through mapping and matching, thus solving the problem of insufficient sensitivity of traditional recognition methods and providing a complete and reliable technical process for the accurate determination of the chemical state of elements in the sample under test.

[0127] The following is a complete example for steps 101-105, such as Figure 3 As shown, an analytical institution conducts elemental chemical state identification on an unknown mineral sample. First, step S101 is performed: the mineral sample to be tested is excited using an X-ray excitation device, with initial excitation parameters set to wavelength 7.0 keV and intensity 120 μA. After excitation, the original feature spectrum generated by the sample is acquired through a spectral detection component. Then, the original feature spectrum is processed by background noise removal (using wavelet denoising algorithm) and spectral line enhancement (using adaptive filtering algorithm), finally generating an enhanced feature spectrum containing information such as element J and element K. The spectral line peak position of element J is 6.5 keV and intensity 98, and the spectral line peak position of element K is 5.9 keV and intensity 85.

[0128] Next, step S102 is executed: feature extraction and labeling are performed on the generated enhanced feature spectrum. First, the 6.5 keV peak position of element J in the enhanced feature spectrum is identified by the spectral line analysis algorithm, which corresponds to its +3 valence chemical state, and the 5.9 keV peak position of element K corresponds to its +2 valence chemical state. Then, exclusive labels are added to the chemical states of these two elements in the enhanced feature spectrum, forming enhanced feature spectra with labels "element J +3 valence" and "element K +2 valence", providing a basis for subsequent adjustment of excitation parameters.

[0129] Then, step S103 is executed: the excitation parameters of the beam filtering system are adjusted according to the "element J+3 valence" mark in the enhanced feature spectrum. First, the target excitation parameter is determined to be a wavelength of 6.5 keV. The beam filtering system is initially adjusted (the crystal incident angle is set to 26.5° and the reflection condition is set to mode G). Then, the actual output wavelength is monitored to be 6.55 keV. The feedback adjustment signal is generated by comparing it with the target parameter. The coordinated adjustment amount (the incident angle is reduced by 0.3° and the reflection condition is fine-tuned by ΔR3) is calculated based on the signal and fine-tuned synchronously. After three iterations, the actual output wavelength is stabilized at 6.5 keV. At this time, the crystal incident angle is 26.2° and the reflection condition mode G is recorded as the excitation condition parameter.

[0130] Next, execute step S104: Establish a dual-channel data interaction architecture. The first channel transmits the spectral feature data of "element J+3 valence (peak position 6.5keV, intensity 98)" in the enhanced feature spectrum, and the second channel transmits the excitation setting data of "incident angle 26.2°, reflection condition mode G" in the excitation condition parameters. Cross-correlate the data from the two channels to establish the correspondence between "peak position 6.5keV - incident angle 26.2°" and "intensity 98 - reflection condition mode G". Based on the correspondence, assign dynamic weight coefficients, calculate the correlation intensity score of 8.2 points, the matching weight factor of 0.85, and the stability weight factor of 0.75, and calculate the coefficient of 0.52 according to the preset strategy. After feedback, adjust it to 0.55. Merge the weighted data to form a unified multidimensional data set, and after adaptive adjustment, obtain the multidimensional analysis features of the excitation conditions.

[0131] Finally, step S105 is executed: the multidimensional analysis features are input into the hierarchical processing module. The first layer extracts the element category features of "element J" through principal component analysis algorithm. The second layer is refined into the chemical state features of "element J + 3 valence" through linear discriminant analysis algorithm. Support vector machine model adaptation is selected to establish a hierarchical mapping relationship between the chemical state features and "element J + 3 valence". According to the mapping, the standard features of element J in the element chemical state database are called. The similarity with the + 3 valence standard features is calculated to be 94% through Euclidean distance matching algorithm. The similarity is compared with the preset 85% threshold. If it is higher than the threshold, the final identification result "the sample to be tested contains element J + 3 valence" is output, completing the entire element chemical state identification process.

[0132] Figure 4 This is a schematic diagram illustrating a specific embodiment of a high-sensitivity detection system for an energy-dispersive X-ray fluorescence spectrometer provided in this application. (Refer to...) Figure 4 The system may include:

[0133] The acquisition module 41 is used to acquire energy dispersive X-ray fluorescence spectral data of the sample to be tested, and obtain initial spectral information including the photon number distribution.

[0134] The generation module 42 is used to perform matching analysis with the elemental feature map library based on the initial spectral information, and generate an enhanced feature spectrum containing elemental chemical state labels by analyzing the chemical shift effect characterized by the peak position shift and intensity change characteristics of the spectral lines.

[0135] The filtering module 43 is used to adjust the excitation parameters of the beam filtering system according to the element chemical state markings in the enhanced feature spectrum, and to optimize them through a feedback control mechanism to generate excitation condition parameters of a single wavelength that match the current element chemical state. The beam filtering system is a tunable optical component for achieving monochromaticity of incident X-rays.

[0136] The forming module 44 is used to collaboratively process the enhanced feature spectrum and the excitation condition parameters, and to cross-correlate and weight the spectral features and excitation conditions by establishing a dual-channel data interaction architecture to form an adaptive multidimensional analysis feature of the excitation conditions.

[0137] The identification module 45 is used to establish a hierarchical mapping relationship between spectral features and elemental chemical states based on the multidimensional analysis features, through hierarchical feature space transformation and identification model adaptation, and to complete the high-sensitivity identification of elemental chemical states in the sample to be tested.

[0138] The high-sensitivity detection system of the energy-dispersive X-ray fluorescence spectrometer in this application embodiment is used to implement the aforementioned high-sensitivity detection method of the energy-dispersive X-ray fluorescence spectrometer. Therefore, the specific implementation of the high-sensitivity detection system of the energy-dispersive X-ray fluorescence spectrometer can be found in the embodiment section of the high-sensitivity detection method of the energy-dispersive X-ray fluorescence spectrometer above. The specific implementation can be referred to the description of the corresponding embodiment, and will not be repeated here.

[0139] This application also provides an electronic device, comprising: a memory for storing a computer program; and a processor for executing the computer program to implement the steps of the high-sensitivity detection method of any of the above-described energy-dispersive X-ray fluorescence spectrometers.

[0140] This application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the high-sensitivity detection method for energy-dispersive X-ray fluorescence spectrometer described above.

[0141] In one exemplary embodiment, the aforementioned computer-readable storage medium may include, but is not limited to, various media capable of storing computer programs, such as USB flash drives, read-only memory, random access memory, portable hard drives, magnetic disks, or optical disks.

[0142] Embodiments of the present invention also provide a computer program product, which includes a computer program that, when executed by a processor, implements the steps in any of the embodiments of the high-sensitivity detection method for energy-dispersive X-ray fluorescence spectrometers described above.

[0143] Those skilled in the art will further 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, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. 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.

[0144] The above provides a detailed description of a high-sensitivity detection method and system for energy-dispersive X-ray fluorescence spectrometer. Specific examples have been used to illustrate the principles and implementation methods of this application. The descriptions of the embodiments above are merely for the purpose of helping to understand the method and its core ideas. It should be noted that those skilled in the art can make various improvements and modifications to this application without departing from its principles, and these improvements and modifications also fall within the protection scope of this application.

Claims

1. A high-sensitivity detection method for an energy-dispersive X-ray fluorescence spectrometer, characterized in that, include: Energy dispersive X-ray fluorescence spectral data of the sample to be tested are collected to obtain initial spectral information including the photon number distribution; Based on the initial spectral information and the elemental feature map library, a matching analysis is performed. By analyzing the chemical shift effect characterized by the peak position shift and intensity change features of the spectral lines, an enhanced feature spectrum containing elemental chemical state labels is generated. Based on the element chemical state markings in the enhanced characteristic spectrum, the excitation parameters of the beam selection system are adjusted and optimized through a feedback control mechanism to generate excitation condition parameters of a single wavelength that match the current element chemical state. The beam selection system is a tunable optical component used to achieve monochromaticity of incident X-rays. The enhanced feature spectrum and the excitation condition parameters are processed collaboratively. By establishing a dual-channel data interaction architecture, the spectral features and excitation conditions are cross-correlated and weighted to form a multi-dimensional analysis feature that adapts to the excitation conditions. Based on the multidimensional analysis features, a hierarchical mapping relationship between spectral features and elemental chemical states is established through hierarchical feature space transformation and recognition model adaptation, thereby achieving high-sensitivity identification of elemental chemical states in the sample to be tested.

2. The method according to claim 1, characterized in that, The process of co-processing the enhanced feature spectrum with the excitation condition parameters, and establishing a dual-channel data interaction architecture to cross-correlate and weight the spectral features and excitation conditions, forms an adaptive multidimensional analysis feature for the excitation conditions, including: A dual-channel data interaction architecture is established, wherein the first channel is used to transmit spectral feature data in the enhanced feature spectrum, and the second channel is used to transmit excitation setting data in the excitation condition parameters. Cross-correlation is performed on the spectral feature data and the excitation setting data in the dual-channel data interaction architecture to establish the correspondence between spectral features and excitation conditions; Based on the aforementioned correspondence, the cross-correlated data is dynamically weighted according to a preset weight allocation strategy; The weighted data are then fused to form the adaptive multidimensional analysis features of the aforementioned activation conditions.

3. The method according to claim 2, characterized in that, The step of dynamically weighting the cross-correlated data based on the correspondence and according to a preset weight allocation strategy includes: Based on the correlation characteristics between the spectral feature data and the excitation setting data in the correspondence, the correlation strength score is calculated; Based on the degree of matching between the spectral features and the excitation conditions in the aforementioned correspondence, a matching weight factor is determined; Based on the stability performance of spectral characteristics under different excitation conditions in the aforementioned correspondence, a stability weighting factor is calculated; Based on a preset weight allocation strategy, the correlation strength score, matching weight factor, and stable weight factor are combined and calculated to generate a dynamic weight coefficient for each data combination. A feedback correlation mechanism is established between the weight coefficients and the corresponding relationship, and the dynamic weight coefficients are adaptively adjusted based on the real-time analysis results.

4. The method according to claim 1, characterized in that, Based on the multidimensional analysis features, a hierarchical mapping relationship between spectral features and elemental chemical states is established through hierarchical feature space transformation and recognition model adaptation, thereby achieving high-sensitivity identification of elemental chemical states in the sample to be tested, including: The multidimensional analysis features are input into the hierarchical processing module. In the first layer feature space transformation, the multidimensional analysis features are converted into element category features. In the second layer feature space transformation, the element category features are further converted into chemical state features. By identifying model adaptation, a hierarchical mapping relationship is established from the chemical state characteristics to the elemental chemical state; Based on the hierarchical mapping relationship, the converted chemical state features are matched with the element chemical state database; Based on the feature matching results, the final identification result of the chemical state of the elements in the sample to be tested is output.

5. The method according to claim 1, characterized in that, The step of adjusting the excitation parameters of the beam selection system based on the element chemical state markers in the enhanced characteristic spectrum, and optimizing them through a feedback control mechanism to generate single-wavelength excitation condition parameters that match the current element chemical state, includes: Based on the elemental chemical state markers in the enhanced feature spectrum, the target excitation parameters are determined; The beam selection system is initially adjusted according to the target excitation parameters; The actual output of the beam selection system is monitored, and the actual output is compared with the target excitation parameters to generate a feedback adjustment signal; The beam selection system is adjusted in a closed loop based on the feedback adjustment signal until the actual output and the target excitation parameters reach a stable matching state. Record the crystal reflection conditions and incident angle when a stable matching state is achieved, as the excitation condition parameters.

6. The method according to claim 5, characterized in that, The closed-loop adjustment of the beam selection system based on the feedback adjustment signal until the actual output and the target excitation parameters reach a stable matching state includes: Based on the feedback adjustment signal, the coordinated adjustment amount of the crystal incident angle and reflection conditions is calculated, and the coordinated adjustment amount is converted into a control command for the crystal driving mechanism to synchronously fine-tune the crystal incident angle and reflection conditions. After each fine-tuning, the actual output X-ray wavelength of the beam filtering system is re-monitored; The re-monitored X-ray wavelength is compared with the target excitation wavelength to update the feedback adjustment signal; The process of adjusting the crystal incident angle and reflection conditions is iteratively executed until the X-ray wavelength and the target excitation parameters reach a stable matching state.

7. The method according to claim 1, characterized in that, The matching analysis based on the initial spectral information and the elemental characteristic spectral library, by analyzing the chemical shift effect characterized by spectral peak shifts and intensity changes, generates an enhanced characteristic spectrum containing elemental chemical state labels, including: Analyze the characteristic changes of the initial spectral information relative to the reference spectrum in the elemental feature map library, including spectral line peak shifts and intensity changes. Based on the aforementioned feature changes, a preset chemical state feature pattern is matched to determine the elemental chemical state label of the target element in the sample to be tested; The element chemical state label is associated with the initial spectral information to generate the enhanced feature spectrum.

8. A high-sensitivity detection system for an energy-dispersive X-ray fluorescence spectrometer, characterized in that, include: The acquisition module is used to acquire energy dispersive X-ray fluorescence spectral data of the sample to be tested, and obtain initial spectral information including the photon number distribution. The generation module is used to perform matching analysis between the initial spectral information and the elemental feature map library, and generate an enhanced feature spectrum containing elemental chemical state labels by analyzing the chemical shift effect characterized by spectral line peak shift and intensity change characteristics. The filtering module is used to adjust the excitation parameters of the beam filtering system according to the element chemical state markings in the enhanced feature spectrum, and to optimize them through a feedback control mechanism to generate excitation condition parameters of a single wavelength that match the current element chemical state. The beam filtering system is a tunable optical component for achieving monochromaticity of incident X-rays. The forming module is used to collaboratively process the enhanced feature spectrum and the excitation condition parameters. By establishing a dual-channel data interaction architecture, the spectral features and excitation conditions are cross-correlated and weighted to form multi-dimensional analysis features that are adaptive to the excitation conditions. The identification module is used to establish a hierarchical mapping relationship between spectral features and elemental chemical states based on the multidimensional analysis features, through hierarchical feature space transformation and identification model adaptation, and to complete the high-sensitivity identification of elemental chemical states in the sample to be tested.

9. An electronic device, characterized in that, include: Memory, used to store computer programs; A processor, configured to execute the computer program to implement the steps of the high-sensitivity detection method for an energy-dispersive X-ray fluorescence spectrometer as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, enables the implementation of a high-sensitivity detection method for an energy-dispersive X-ray fluorescence spectrometer as described in any one of claims 1 to 7.

Images