A method and system for determining the composition of titanium alloys based on a spectrometer

Abstract

This invention relates to the field of spectroscopic detection technology, specifically disclosing a method and system for determining the composition of titanium alloys based on a spectrometer. The method includes: acquiring raw spectral data of a titanium alloy sample using a spectrometer, filtering high-frequency noise from the raw spectral data to obtain clean spectral data; detecting local minima in the clean spectral data to construct a baseline profile of the titanium alloy sample, and performing drift calibration on the baseline profile to obtain optimized spectral data; comparing the characteristic wavelength heights of adjacent elements and integrating the obtained peak data into characteristic spectral peaks; performing similarity matching between the characteristic spectral peaks and wavelengths in a preset element characteristic wavelength library to determine the intensity of the characteristic peaks; extracting element concentration data and correlating the element concentration data with the intensity of the characteristic peaks to construct a calibration curve; mapping the optimized spectral data to the calibration curve to generate composition ratio data. This invention can improve the efficiency of spectrometer-based titanium alloy composition determination.

Description

Technical Field

[0001] This invention relates to the field of spectroscopic detection technology, and in particular to a method and system for determining the composition of titanium alloys based on a spectrometer. Background Technology

[0002] In the field of spectrometer-based titanium alloy composition determination, existing technologies lack targeted solutions for noise processing of raw spectral data. They fail to implement differentiated filtering strategies based on noise intensity levels, resulting in residual high-frequency interference in the denoised data and poor alignment of data segment boundaries, directly impacting the quality of clean spectral data. Furthermore, the selection of local minima during baseline profile construction is not precise enough, outlier removal is incomplete, and baseline drift calibration lacks segmented compensation and smooth transition processing, leading to distortion in the optimized spectral data and creating potential problems for subsequent composition analysis.

[0003] Existing technologies have significant shortcomings in the extraction and matching of characteristic spectral peaks. They lack sufficient analysis of the correlation between spectral peaks, easily leading to duplicate peaks or omissions of characteristic peaks. Furthermore, the matching with a pre-defined elemental characteristic wavelength library focuses only on wavelength parameters, neglecting peak shape consistency and signal-to-noise ratio evaluation, resulting in insufficient accuracy in determining characteristic peak intensities. In addition, the selection and utilization of historical detection data during calibration curve construction are not reasonable enough, the correlation logic between concentration and characteristic peak intensity is not rigorous enough, and the outlier handling mechanism during mapping is imperfect. Ultimately, this results in low efficiency in generating titanium alloy composition ratio data, failing to meet the demands of rapid and accurate determination in practical testing scenarios. Therefore, improving the efficiency of spectrometer-based titanium alloy composition determination has become an urgent problem to be solved. Summary of the Invention

[0004] This invention provides a method and system for determining the composition of titanium alloys based on a spectrometer, in order to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides a method for determining the composition of titanium alloys based on a spectrometer, comprising:

[0006] S1. Collect the raw spectral data of the titanium alloy sample using a spectrometer, and filter the high-frequency noise of the raw spectral data to obtain the clean spectral data of the titanium alloy sample.

[0007] S2. Detect local minima in the clean spectral data to construct the baseline profile of the titanium alloy sample, and perform drift calibration on the baseline profile to obtain the optimized spectral data of the titanium alloy sample.

[0008] S3. Based on the characteristic wavelengths of the elements in the optimized spectral data, compare the characteristic wavelength heights of adjacent elements to obtain the spectral peak data of the optimized spectral data, and integrate the spectral peak data into the characteristic spectral peaks of the optimized spectral data.

[0009] S4. Match the characteristic spectral peaks with wavelengths in a preset element characteristic wavelength library to determine the intensity of the characteristic peaks of the elements in the optimized spectral data;

[0010] S5. Extract the elemental concentration data of the titanium alloy sample from the historical detection data, and correlate the elemental concentration data with the characteristic peak intensity to construct the calibration curve of the titanium alloy sample.

[0011] S6. Map the optimized spectral data to the calibration curve to generate the composition ratio data of the titanium alloy sample.

[0012] In a preferred embodiment, the step of acquiring the raw spectral data of the titanium alloy sample using a spectrometer and filtering the high-frequency noise of the raw spectral data to obtain clean spectral data of the titanium alloy sample includes:

[0013] Raw spectral data of titanium alloy samples were acquired using a spectrometer;

[0014] The original spectral data is divided into data segments according to wavelength order to obtain a data segment sequence of the original spectral data;

[0015] Extract the noise fluctuation characteristics of the data segment sequence, and determine the noise intensity level corresponding to the data segment sequence based on the noise fluctuation characteristics;

[0016] Select the noise suppression strategy corresponding to the noise intensity level from the preset filtering strategies;

[0017] Based on the noise suppression strategy, data segments with different noise intensity levels in the data segment sequence are differentially targeted and filtered out to obtain the denoised data segments of the data segment sequence.

[0018] The denoised data segments are then boundary aligned to obtain clean spectral data of the titanium alloy sample.

[0019] In a preferred embodiment, the step of detecting local minima in the clean spectral data to construct a baseline profile of the titanium alloy sample, and performing drift calibration on the baseline profile to obtain optimized spectral data of the titanium alloy sample, includes:

[0020] The clean spectral data is scanned point by point according to a sliding window of fixed width, and the point with the lowest signal intensity within the sliding window is taken as a candidate local minimum point to form a candidate point set of the clean spectral data.

[0021] By removing outliers from the candidate point set that deviate from the main distribution area, the reference points of the clean spectral data are obtained;

[0022] Curve fitting is performed on the reference points according to the wavelength coordinate order to generate the initial baseline profile of the clean spectral data;

[0023] Assess the degree of deviation of the characteristic wavelength positions of the initial baseline profile relative to the overall trend;

[0024] Based on the degree of offset, the initial baseline profile is adjusted in segments to obtain optimized spectral data of the titanium alloy sample.

[0025] In a preferred embodiment, the step of segmentally compensating and adjusting the initial baseline profile based on the degree of offset to obtain optimized spectral data of the titanium alloy sample includes:

[0026] Based on the degree of deviation of the characteristic wavelength position, the initial baseline profile is divided into continuous adjustment segments;

[0027] The compensation direction and magnitude of the adjustment segment are determined based on the average deviation direction and magnitude of the characteristic wavelength points in the adjustment segment.

[0028] Based on the compensation direction and the amplitude, directional amplitude compensation is performed on the baseline contour points in the adjustment section to obtain the compensation section of the initial baseline contour;

[0029] At the boundary of adjacent adjustment sections, the contour points of the compensation section are smoothed to obtain optimized spectral data of the titanium alloy sample.

[0030] In a preferred embodiment, the step of comparing the characteristic wavelength heights of adjacent elements based on the characteristic wavelengths of elements in the optimized spectral data to obtain the spectral peak data of the optimized spectral data, and integrating the spectral peak data into the characteristic spectral peaks of the optimized spectral data, includes:

[0031] Identify the spectral peak points of elements in the optimized spectral data to form an initial set of spectral peak points for the optimized spectral data;

[0032] Using the spectral peaks in the initial set of spectral peaks as centers, construct a neighborhood relation matrix for the optimized spectral data;

[0033] Based on the neighborhood relation matrix, the intensity gradient changes between adjacent spectral peak points are analyzed to establish the correlation between the spectral peak points;

[0034] Based on the aforementioned correlation, spectral peaks with continuously changing characteristics are merged into a set of characteristic spectral peaks of the optimized spectral data.

[0035] Eliminate the repetitive peak structures in the set of characteristic peaks to form the characteristic peaks of the optimized spectral data.

[0036] In a preferred embodiment, the step of performing similarity matching between the characteristic spectral peak and wavelengths in a preset elemental characteristic wavelength library to determine the characteristic peak intensity of elements in the optimized spectral data includes:

[0037] Extract the reference wavelength ranges from the preset element feature wavelength library that correspond to the elements in the optimized spectral data;

[0038] The wavelength parameters of the characteristic spectral peaks are mapped to the preset element characteristic wavelength library to obtain the initial matching pairs of the optimized spectral data;

[0039] Based on the peak shape profile of the characteristic spectral peak, the consistency between the characteristic spectral peak in the initial matching pair and the reference wavelength range is evaluated, and candidate matching pairs with matching peak shapes are selected from the initial matching pairs.

[0040] The signal-to-noise ratio of the candidate matching pairs is evaluated, and the signal amplitude of the characteristic spectral peak in the high-confidence candidate matching pairs is extracted as the characteristic peak intensity of the element.

[0041] In a preferred embodiment, the step of extracting elemental concentration data of the titanium alloy sample from historical detection data and correlating the elemental concentration data with the characteristic peak intensity to construct a calibration curve for the titanium alloy sample includes:

[0042] Based on the material grade of the titanium alloy sample, historical test records with the same material grade are retrieved from the historical database of the titanium alloy sample to obtain the initial historical data of the titanium alloy sample.

[0043] By removing historical data whose concentration values ​​exceed the conventional composition range of titanium alloys from the initial historical data, the effective historical data of the titanium alloy sample is obtained.

[0044] The effective historical data is quantified to determine the concentration values ​​of elements in the effective historical data, and the concentration values ​​are integrated into the elemental concentration data of the titanium alloy sample.

[0045] By binding the concentration value with the corresponding characteristic peak intensity, a concentration-intensity correspondence relationship for the titanium alloy sample is generated.

[0046] Based on the concentration-intensity correspondence, the variation trend between the concentration value and the corresponding characteristic peak intensity is analyzed to construct the calibration curve of the titanium alloy sample.

[0047] In a preferred embodiment, the calibration curve is calculated using the following formula:

[0048] ;

[0049] In the formula, Indicating the first in the optimized spectral data The final concentration of each element, Indicating the first in the optimized spectral data The preset linear fitting slope coefficients for each element. Indicating the first in the optimized spectral data Characteristic peak intensities of the elements Indicates the first of the valid historical data The average characteristic peak intensity of the elements Indicating the first in the optimized spectral data The preset linear fitting intercept coefficients for each element. Indicating the first in the optimized spectral data The preset baseline offset correction coefficient for each element. Indicating the first in the optimized spectral data The degree of shift in the characteristic wavelength of a species. Indicating the first in the optimized spectral data The standard characteristic wavelength of the element Indicating the first in the optimized spectral data The signal-to-noise ratio (SNR) of the characteristic spectral peaks of each element is compared to the reference SNR. Indicating the first in the optimized spectral data The reference signal-to-noise ratio of each element. Indicating the first in the optimized spectral data The preset signal-to-noise ratio weighting coefficients for each element Indicates the first of the valid historical data The average concentration of each element.

[0050] In a preferred embodiment, mapping the optimized spectral data to the calibration curve to generate the compositional ratio data of the titanium alloy sample includes:

[0051] The optimized spectral data is input into the effective mapping range of the calibration curve as the mapping data to be processed for the calibration curve;

[0052] Abnormal mapping points are removed from the mapping data to be processed to obtain a valid mapping subset of the mapping data to be processed;

[0053] The elemental composition data of the effective mapping subset are fitted using the calibration curve to form the initial composition ratio dataset of the titanium alloy sample.

[0054] The initial composition ratio dataset is normalized to generate the composition ratio data of the titanium alloy sample.

[0055] To address the above problems, the present invention also provides a titanium alloy composition determination system based on a spectrometer, the system comprising:

[0056] The spectral data denoising module is used to acquire the original spectral data of the titanium alloy sample using a spectrometer, and filter the high-frequency noise of the original spectral data to obtain clean spectral data of the titanium alloy sample.

[0057] The baseline calibration module is used to detect local minima in the clean spectral data to construct the baseline profile of the titanium alloy sample, and to perform drift calibration on the baseline profile to obtain the optimized spectral data of the titanium alloy sample.

[0058] The feature peak extraction module is used to compare the feature wavelength heights of adjacent elements based on the feature wavelengths of the elements in the optimized spectral data, obtain the spectral peak data of the optimized spectral data, and integrate the spectral peak data into the feature spectral peaks of the optimized spectral data.

[0059] The peak intensity determination module is used to perform similarity matching between the characteristic spectral peaks and wavelengths in a preset element characteristic wavelength library to determine the characteristic peak intensity of elements in the optimized spectral data.

[0060] The calibration curve construction module is used to extract the elemental concentration data of the titanium alloy sample from historical detection data and correlate the elemental concentration data with the characteristic peak intensity to construct the calibration curve of the titanium alloy sample.

[0061] The composition ratio generation module is used to map the optimized spectral data to the calibration curve to generate the composition ratio data of the titanium alloy sample.

[0062] Compared with the prior art, the present invention has the following beneficial effects:

[0063] 1. This invention achieves efficient acquisition of clean spectral data by implementing differentiated targeted filtering based on noise intensity levels and aligning data segment boundaries; it optimizes spectral data quality by combining sliding window screening of reference points, segmented compensation calibration of baselines, and smooth transition processing; and it eliminates duplicate spectral peaks and accurately extracts characteristic spectral peaks by analyzing peak correlation through neighborhood relation matrix analysis, providing reliable data support for component determination.

[0064] 2. This invention accurately determines the intensity of characteristic peaks by taking into account wavelength, peak shape profile, and signal-to-noise ratio during the matching process; it constructs a calibration curve with multi-dimensional correction coefficients based on the screening of valid historical data with the same material grade, and combines outlier removal and normalization processing to achieve efficient mapping of optimized spectral data to composition ratio data, significantly improving the efficiency and accuracy of titanium alloy composition determination. Attached Figure Description

[0065] Figure 1 This is a schematic flowchart of a method for determining the composition of titanium alloys based on a spectrometer, provided in an embodiment of the present invention.

[0066] Figure 2 A functional block diagram of a titanium alloy composition determination system based on a spectrometer provided in an embodiment of the present invention;

[0067] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0068] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0069] This application provides a method for determining the composition of titanium alloys based on a spectrometer. The execution entity of this spectrometer-based titanium alloy composition determination method includes, but is not limited to, at least one of the following electronic devices that can be configured to execute the method provided in this application: a server, a terminal, etc. In other words, the spectrometer-based titanium alloy composition determination method can be executed by software or hardware installed on a terminal device or a server device. The server includes, but is not limited to, a single server, a server cluster, a cloud server, or a cloud server cluster. The server can be an independent server or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, content delivery networks (CDNs), and big data and artificial intelligence platforms.

[0070] Reference Figure 1 The diagram shown is a schematic flowchart of a spectrometer-based method for determining the composition of titanium alloys according to an embodiment of the present invention. In this embodiment, the spectrometer-based method for determining the composition of titanium alloys includes:

[0071] S1. Collect the raw spectral data of the titanium alloy sample using a spectrometer, and filter the high-frequency noise of the raw spectral data to obtain the clean spectral data of the titanium alloy sample.

[0072] In this embodiment of the invention, the step of acquiring the raw spectral data of the titanium alloy sample using a spectrometer and filtering the high-frequency noise of the raw spectral data to obtain clean spectral data of the titanium alloy sample includes:

[0073] Raw spectral data of titanium alloy samples were acquired using a spectrometer;

[0074] The original spectral data is divided into data segments according to wavelength order to obtain a data segment sequence of the original spectral data;

[0075] Extract the noise fluctuation characteristics of the data segment sequence, and determine the noise intensity level corresponding to the data segment sequence based on the noise fluctuation characteristics;

[0076] Select the noise suppression strategy corresponding to the noise intensity level from the preset filtering strategies;

[0077] Based on the noise suppression strategy, data segments with different noise intensity levels in the data segment sequence are differentially targeted and filtered out to obtain the denoised data segments of the data segment sequence.

[0078] The denoised data segments are then boundary aligned to obtain clean spectral data of the titanium alloy sample.

[0079] Clean the surface of the titanium alloy sample thoroughly, removing oil, oxide layers, and other impurities. Ensure the sample testing surface is flat and unobstructed. Fix the treated titanium alloy sample in the center of the spectrometer's sample testing stage, maintaining a preset fixed distance between the sample testing surface and the spectrometer's probe. This distance must ensure the probe can fully receive the sample's spectral response signal. Start the spectrometer, set the wavelength range for spectral acquisition, and the probe emits continuous spectral rays within this range onto the titanium alloy sample surface. The sample produces specific absorption, reflection, or transmission responses to rays of different wavelengths. The spectrometer's signal receiving module captures these response signals in real time, converts them into electrical signals, and then the signal conversion module converts these electrical signals into raw spectral data containing wavelengths and corresponding response intensities. This data is directly stored in a standard data format, forming the raw spectral data of the titanium alloy sample.

[0080] The original spectral data is divided into continuous data segments in natural order of wavelength from shortest to longest. Each segment is divided into fixed wavelength intervals, with each interval having a consistent length to ensure that each segment contains the same number of original spectral response intensity data corresponding to the same wavelength. During the segmentation process, starting with the response intensity data corresponding to the smallest wavelength in the original spectral data, all data within a fixed-length wavelength interval are sequentially extracted as an independent data segment. Adjacent data segments are joined end-to-end, with the maximum wavelength of the preceding segment completely matching the minimum wavelength of the following segment, ensuring no overlap or omission of wavelength ranges. This results in multiple continuous data segments arranged in wavelength order. These data segments are combined to form a sequence of original spectral data segments, each segment clearly labeled with its corresponding wavelength range and the original spectral response intensity data it contains.

[0081] For each data segment in the data segment sequence, fluctuation analysis is performed on the original spectral response intensity data it contains. Starting from the first response intensity data point in the data segment, the difference in response intensity between each data point and its immediately following data point is calculated sequentially. The absolute values ​​of all differences within the data segment are counted, and the maximum value of the absolute difference and the frequency of occurrence of large differences are recorded. These maximum values ​​of absolute differences and the frequency of occurrence of large differences together constitute the noise fluctuation characteristics of the data segment. The preset noise fluctuation standard is divided into three distinct levels, each with a fixed range of maximum absolute difference values ​​and a range of occurrences of large differences. The noise fluctuation characteristics of each data segment are compared one by one with the preset three-level standard. When the noise fluctuation characteristics of a data segment fall completely within the standard range of a certain level, the noise intensity level corresponding to that data segment can be determined. After all data segments have completed the level determination, a noise intensity level sequence corresponding one-to-one with the data segment sequence is formed.

[0082] The preset filtering strategy includes three different noise suppression strategies, corresponding to low, medium, and high noise intensity levels. Each noise suppression strategy is designed with specific processing methods tailored to the characteristics of the corresponding noise level. The suppression strategy for the low noise intensity level focuses on preserving the original data features, the suppression strategy for the medium noise intensity level focuses on smoothing data fluctuations, and the suppression strategy for the high noise intensity level focuses on filtering out core trends. After determining the noise intensity level of each data segment in the data segment sequence, the one that perfectly matches the noise intensity level of each data segment is precisely selected from the three preset noise suppression strategies, ensuring that each data segment corresponds to a specific and targeted noise suppression strategy, ultimately forming a noise suppression strategy sequence consistent with the order of the data segment sequence.

[0083] Based on the matched noise suppression strategy sequence, differentiated targeted filtering is performed on each data segment in the data segment sequence. For the denoised data segment corresponding to the low noise intensity level, each response intensity data point within the data segment is traversed one by one. The response intensity of each data point is compared with the response intensity of the two adjacent data points before and after it. If the deviation of the response intensity of the data point from the average response intensity of the adjacent data points exceeds a fixed range, it is determined to be a noise data point, and the noise data point is replaced with the average response intensity of the four adjacent data points. The remaining data points remain unchanged. For the denoised data segment corresponding to the noise intensity level, five consecutive data points within the data segment are divided into a local unit in sequence. The average response intensity of the five data points in each local unit is calculated, and the data point in the middle position of the unit is replaced with the average response intensity. All local units are traversed in sequence to complete the smoothing processing of data fluctuations and preserve the overall trend of the data segment. For a denoised data segment corresponding to a high noise intensity level, first connect the first and last data points within the data segment to form an initial trend line. Then calculate the distance from each intermediate data point to the initial trend line. Retain data points whose distance is within a fixed range. For noise data points that exceed the range, replace them with the theoretical response intensity at the corresponding wavelength position on the initial trend line. This ensures that each data segment effectively removes its own noise while retaining the core spectral features. Finally, the denoised data segment corresponding to each data segment is obtained and combined to form a denoised data segment sequence.

[0084] All denoised data segments in the denoised data segment sequence undergo boundary alignment. First, the boundary data of each denoised data segment is extracted, namely the response intensity data with the smallest wavelength and the response intensity data with the largest wavelength in each segment. The connection boundaries of adjacent denoised data segments are compared sequentially according to wavelength. The maximum wavelength response intensity data of the preceding segment and the minimum wavelength response intensity data of the following segment form a pair of connection data. The numerical difference of each pair of connection data is observed. If the difference is within a preset reasonable range, the boundary data of the two data segments remain unchanged. If the difference exceeds a reasonable range, the two connection data are fine-tuned by referring to the changing trends of the last five data points of the preceding segment and the changing trends of the first five data points of the following segment. This ensures that the last data of the preceding segment naturally transitions to the starting data of the following segment, without changing other data or the overall trend within the two data segments. After all adjacent denoised data segments have been adjusted to a smooth connection state, all denoised data segments are sequentially spliced ​​together according to the original wavelength order to form complete, boundary-free, and noise-removed spectral data. This data is the clean spectral data of the titanium alloy sample.

[0085] The beneficial effects are as follows: by dividing the raw spectral data of titanium alloy samples into data segments according to wavelength order, the noise fluctuation characteristics of each data segment are accurately extracted and the corresponding noise intensity level is determined. Then, a dedicated noise suppression strategy is matched for differentiated targeted filtering, which not only removes noise of different intensities but also preserves the core spectral characteristics of each data segment to the greatest extent. Subsequently, boundary alignment processing is used to smoothly connect the denoised data segments, and finally clean spectral data with no noise interference, coherent boundaries, and true reflection of the spectral characteristics of titanium alloy samples are obtained, providing an accurate and reliable spectral basis for the subsequent analysis and detection of titanium alloy samples.

[0086] S2. Detect local minima in the clean spectral data to construct the baseline profile of the titanium alloy sample, and perform drift calibration on the baseline profile to obtain the optimized spectral data of the titanium alloy sample.

[0087] In this embodiment of the invention, the step of detecting local minima in the clean spectral data to construct the baseline profile of the titanium alloy sample, and performing drift calibration on the baseline profile to obtain optimized spectral data of the titanium alloy sample, includes:

[0088] The clean spectral data is scanned point by point according to a sliding window of fixed width, and the point with the lowest signal intensity within the sliding window is taken as a candidate local minimum point to form a candidate point set of the clean spectral data.

[0089] By removing outliers from the candidate point set that deviate from the main distribution area, the reference points of the clean spectral data are obtained;

[0090] Curve fitting is performed on the reference points according to the wavelength coordinate order to generate the initial baseline profile of the clean spectral data;

[0091] Assess the degree of deviation of the characteristic wavelength positions of the initial baseline profile relative to the overall trend;

[0092] Based on the degree of offset, the initial baseline profile is adjusted in segments to obtain optimized spectral data of the titanium alloy sample.

[0093] The step of performing segmented compensation adjustment on the initial baseline profile based on the degree of offset to obtain optimized spectral data of the titanium alloy sample includes:

[0094] Based on the degree of deviation of the characteristic wavelength position, the initial baseline profile is divided into continuous adjustment segments;

[0095] The compensation direction and magnitude of the adjustment segment are determined based on the average deviation direction and magnitude of the characteristic wavelength points in the adjustment segment.

[0096] Based on the compensation direction and the amplitude, directional amplitude compensation is performed on the baseline contour points in the adjustment section to obtain the compensation section of the initial baseline contour;

[0097] At the boundary of adjacent adjustment sections, the contour points of the compensation section are smoothed to obtain optimized spectral data of the titanium alloy sample.

[0098] Prepare a fixed-width sliding window containing a fixed number of consecutive spectral data points. Align the sliding window with the first data point with the shortest wavelength in the clean spectral data, using this data point as the starting position, and ensure the window completely covers the starting position and the corresponding number of subsequent consecutive data points. Then, move the sliding window point by point, advancing only one data point at a time, until the sliding window covers the last data point with the longest wavelength in the clean spectral data, completing a comprehensive scan of the clean spectral data. At each stop point in the window, compare the signal intensity of all data points within the window range, selecting the data point with the lowest signal intensity. Mark this data point as a candidate local minimum and record its corresponding wavelength coordinates and signal intensity value. After the entire scan process is complete, collect all marked candidate local minima and integrate them to form a candidate point set of clean spectral data containing the wavelength coordinates and signal intensity information of each point.

[0099] The signal intensity values ​​of all candidate local minima in the candidate point set are organized and arranged in ascending order of signal intensity. The distribution of these values ​​is observed to determine the range of signal intensity concentrated in the middle, which is the main distribution area of ​​the candidate point set. Then, each candidate local minima in the candidate point set is checked one by one to see if its signal intensity value falls within the determined main distribution area. If the signal intensity value of a candidate local minima exceeds the main distribution area, the point is removed from the candidate point set and identified as an outlier. After all candidate local minima have been checked and outliers removed, the remaining candidate local minima collectively constitute the reference points for the clean spectral data.

[0100] All reference points are arranged in ascending order of wavelength coordinates, ensuring that the order of the reference points matches the wavelength order of the cleanroom spectral data. Based on this arrangement, a curve is constructed using a continuous connection method. Starting with the reference point with the shortest wavelength, a smooth line segment connects this point to the next adjacent reference point, ensuring the line segment accurately passes through the coordinates of the two reference points. This process continues sequentially, connecting all subsequent adjacent reference points in the same way, until the last reference point with the longest wavelength is reached, forming a continuous, smooth curve that perfectly matches the distribution trend of all reference points. This curve, formed by the orderly connection of reference points, constitutes the initial baseline profile of the cleanroom spectral data.

[0101] Carefully observe the overall trend of the initial baseline profile, analyzing the signal intensity variation from the shortest to the longest wavelength to determine the overall trend of the initial baseline profile, which may be a gradual increase, a gradual decrease, or a relatively stable state. Then, identify the characteristic wavelength positions on the initial baseline profile. These positions are usually key points where the signal intensity trend changes, such as a transition from an increasing to a decreasing trend, a transition from a decreasing to an increasing trend, or a sudden and significant fluctuation in signal intensity. For each identified characteristic wavelength position, compare the actual signal intensity at that position with the expected signal intensity for that wavelength under the overall trend of the initial baseline profile, analyzing the magnitude of the difference to determine the direction and extent of deviation of the characteristic wavelength position relative to the overall trend, thus comprehensively assessing the degree of offset for each characteristic wavelength position.

[0102] Based on the distribution of characteristic wavelength positions on the initial baseline profile, the initial baseline profile is divided into multiple continuous segments, with two adjacent characteristic wavelength positions as the dividing points. Each segment contains a complete wavelength range and a corresponding baseline curve. For each segment, the degree of offset of all characteristic wavelength positions within that segment is summarized to determine the overall offset of the segment. If the offset indicates that the signal strength of the segment is generally higher than the overall trend, the signal strength of all data points within that segment is reduced by a uniform and appropriate amount. If the offset indicates that the signal strength of the segment is generally lower than the overall trend, the signal strength of all data points within that segment is increased by a uniform and appropriate amount. During the adjustment process, the relative change trend between data points within the segment is strictly maintained, and the overall direction of the baseline within the segment is not changed. After all segments have been compensated and adjusted, a precisely corrected baseline profile is formed. This corrected baseline profile is then superimposed on the original clean spectral data to remove background interference caused by the baseline, ultimately obtaining optimized spectral data of the titanium alloy sample with clearer signal characteristics and removed baseline effects.

[0103] The deviation of all characteristic wavelength positions on the initial baseline profile is analyzed to clarify the specific deviation of each characteristic wavelength position from the overall trend. The characteristic wavelength positions where the deviation changes significantly are identified, and these positions are used as the boundary points for adjustment segments. Based on these boundary points, the complete range of the initial baseline profile from the shortest to the longest wavelength is divided into multiple continuous parts. Each part contains a continuous wavelength interval and a corresponding baseline profile segment, ensuring that adjacent adjustment segments are connected end-to-end in terms of wavelength, without overlap or gaps. The deviation of characteristic wavelength positions within each adjustment segment maintains a consistent trend, ultimately forming multiple continuous and clearly defined adjustment segments.

[0104] For each adjustment segment, the deviation direction of all characteristic wavelength points within the segment is statistically analyzed. It is determined whether each characteristic wavelength point is above or below the overall trend. The direction with the greater prevalence of either deviation direction is identified as the average deviation direction for that adjustment segment. The deviation magnitudes of all characteristic wavelength points within the adjustment segment are collected, summed, and then divided by the total number of characteristic wavelength points to obtain the average deviation magnitude for that adjustment segment. The compensation direction is determined based on the average deviation direction. If the average deviation direction is above the overall trend, the compensation direction is to reduce signal strength; if the average deviation direction is below the overall trend, the compensation direction is to increase signal strength. The average deviation magnitude is directly determined as the compensation amplitude for that adjustment segment, ultimately clarifying the compensation direction and amplitude for each adjustment segment.

[0105] According to the compensation direction and amplitude determined for each adjustment segment, all baseline contour points within the segment are oriented. If the compensation direction is to reduce signal strength, the signal strength value of each baseline contour point within the adjustment segment is reduced by an equal amount according to the compensation amplitude, ensuring that the reduction amplitude of each point is completely consistent. If the compensation direction is to increase signal strength, the signal strength value of each baseline contour point within the adjustment segment is increased by an equal amount according to the compensation amplitude, similarly ensuring that the increase amplitude of each point is the same. During the compensation process, the overall orientation and relative change relationship of the baseline contour within each adjustment segment remain unchanged; only the overall level of signal strength is adjusted. After the compensation of all baseline contour points is completed, each adjustment segment forms a corresponding compensation segment, and all compensation segments together constitute the overall baseline contour after directional amplitude compensation.

[0106] Identify the boundary points of the compensation segments corresponding to two adjacent adjustment segments, namely the longest wavelength point of the previous compensation segment and the shortest wavelength point of the subsequent compensation segment. Simultaneously, extract multiple baseline contour points adjacent to each boundary point to form a target point set for boundary transition processing. Calculate the signal intensity change gradient at each point in the target point set to determine the magnitude of signal intensity change between adjacent points. Fine-tune the signal intensity of target points near the boundary points. If the signal intensity of the boundary point of the previous compensation segment is higher than that of the boundary point of the subsequent compensation segment, gradually decrease the signal intensity of points near the boundary of the previous segment while gradually increasing the signal intensity of points near the boundary of the subsequent segment. If the signal intensity of the boundary point of the previous compensation segment is lower than that of the boundary point of the subsequent compensation segment, reverse the fine-tuning to make the signal intensity change gradient at the boundary of adjacent compensation segments tend to be gradual, avoiding abrupt changes. After the boundaries of all adjacent compensation sections have been smoothly transitioned, all compensation sections are seamlessly connected in wavelength order to form a complete and smooth correction baseline. This correction baseline is then subtracted from the original clean spectral data to remove the interference of the baseline background, and finally, optimized spectral data of the titanium alloy sample with prominent signal characteristics and accurate data are obtained.

[0107] The beneficial effects are as follows: by scanning clean spectral data point by point through a fixed-width sliding window to screen candidate local minima, after forming a candidate point set, outliers that deviate from the main distribution area are removed to obtain benchmark points. The benchmark points are fitted in order of wavelength coordinates to form an initial baseline profile. The degree of deviation of the characteristic wavelength position relative to the overall trend is accurately evaluated and segmented compensation adjustment is performed, effectively removing baseline background interference, making the optimized spectral data signal characteristics clearer, and truly reflecting the spectral essence of titanium alloy samples, providing accurate and reliable data support for subsequent sample-related analysis and detection.

[0108] S3. Based on the characteristic wavelengths of the elements in the optimized spectral data, compare the characteristic wavelength heights of adjacent elements to obtain the spectral peak data of the optimized spectral data, and integrate the spectral peak data into the characteristic spectral peaks of the optimized spectral data.

[0109] In this embodiment of the invention, the step of comparing the characteristic wavelength heights of adjacent elements based on the characteristic wavelengths of elements in the optimized spectral data to obtain the spectral peak data of the optimized spectral data, and integrating the spectral peak data into the characteristic spectral peaks of the optimized spectral data, includes:

[0110] Identify the spectral peak points of elements in the optimized spectral data to form an initial set of spectral peak points for the optimized spectral data;

[0111] Using the spectral peaks in the initial set of spectral peaks as centers, construct a neighborhood relation matrix for the optimized spectral data;

[0112] Based on the neighborhood relation matrix, the intensity gradient changes between adjacent spectral peak points are analyzed to establish the correlation between the spectral peak points;

[0113] Based on the aforementioned correlation, spectral peaks with continuously changing characteristics are merged into a set of characteristic spectral peaks of the optimized spectral data.

[0114] Eliminate the repetitive peak structures in the set of characteristic peaks to form the characteristic peaks of the optimized spectral data.

[0115] The process iterates through all data points in the optimized spectral data, starting with the data point with the shortest wavelength. The signal intensity of each data point is compared sequentially with the signal intensity of its immediate preceding and following data points. If the signal intensity of a data point is simultaneously higher than that of its preceding and following data points, and the wavelength corresponding to that data point falls within the characteristic wavelength range of the target element in the titanium alloy, then that data point can be identified as a spectral peak point of the element. The wavelength coordinates and signal intensity values ​​corresponding to each spectral peak point are recorded. All spectral peak points that meet the criteria are then aggregated and integrated to form an initial set of spectral peak points for the optimized spectral data containing complete peak information.

[0116] Each spectral peak point in the initial set of spectral peak points is extracted and designated as a core peak point. Using the wavelength coordinates of each core peak point as a reference, a fixed wavelength range is defined as its neighborhood. This neighborhood range covers the intervals before and after the core peak point where other spectral peak points may exist. Other spectral peak points within the neighborhood of each core peak point are identified one by one, clarifying the correspondence between each core peak point and its neighboring spectral peak points. Following the order of the spectral peak points in the initial set, all core peak points and their corresponding neighboring spectral peak point information are systematically organized to construct a neighborhood relation matrix for optimized spectral data, clearly presenting the neighboring associated objects of each spectral peak point.

[0117] Extract each pair of adjacent spectral peaks from the neighborhood relation matrix, i.e., the core peak and each spectral peak in its neighborhood. Calculate the difference between the signal intensity of the previous spectral peak and the signal intensity of the next spectral peak. Determine the direction of the intensity gradient change between adjacent spectral peaks based on the sign of the difference—whether the signal intensity is increasing or decreasing. Simultaneously observe the intensity gradient change pattern of multiple adjacent pairs of spectral peaks. If the direction of intensity gradient change is consistent and the amplitude is continuously stable among multiple consecutive adjacent spectral peaks, it can be determined that these spectral peaks have an interrelationship. By marking the correspondence of these spectral peaks, the correlation between spectral peaks is established.

[0118] Based on the established correlations between spectral peaks, all spectral peaks in the initial set are categorized and integrated. Peaks that are correlated and exhibit continuously varying intensity gradients are grouped together, with each group forming a complete characteristic peak. This characteristic peak contains all wavelength coordinates and signal intensity information of a set of continuously varying spectral peaks. All the resulting characteristic peaks are then consolidated and organized to ensure each peak is clearly distinguishable from the others, ultimately forming a set of optimized spectral peaks for the spectral data.

[0119] Each characteristic spectral peak in the characteristic peak set is compared and analyzed one by one. First, the wavelength range of each characteristic spectral peak is extracted to determine whether there is any overlap in the wavelength ranges of different characteristic spectral peaks. Then, the signal intensity distribution of the characteristic spectral peaks within the overlapping wavelength ranges is compared, including the maximum and minimum signal intensity values ​​and their trends. If the wavelength ranges of two or more characteristic spectral peaks completely overlap and the various characteristics of the signal intensity distribution are highly consistent, these characteristic spectral peaks are determined to be repetitive peak structures. The characteristic spectral peak with the most complete signal intensity data and the clearest peak characteristics in the repetitive peak structure is retained, and the remaining repetitive characteristic spectral peaks are removed from the characteristic peak set. After all repetitive peak structures have been identified and removed, the remaining characteristic spectral peaks in the characteristic peak set are the characteristic spectral peaks of the optimized spectral data.

[0120] The beneficial effects include accurately identifying and optimizing the spectral peaks of elements in the spectral data to form an initial set of spectral peaks, constructing a neighborhood relation matrix with each peak as the center, establishing correlations by analyzing the intensity gradient changes between adjacent peaks, merging peaks with continuous variation characteristics into a set of characteristic spectral peaks, effectively eliminating the repetitive spectral peak structures in the set, and finally obtaining clear and unique characteristic spectral peaks. This fully preserves the spectral characteristic information of titanium alloy elements, providing accurate and reliable spectral basis for subsequent work such as elemental composition analysis and content detection of titanium alloy samples.

[0121] S4. Match the characteristic spectral peaks with wavelengths in a preset element characteristic wavelength library to determine the intensity of the characteristic peaks of the elements in the optimized spectral data;

[0122] In this embodiment of the invention, the step of performing similarity matching between the characteristic spectral peak and wavelengths in a preset elemental characteristic wavelength library to determine the characteristic peak intensity of elements in the optimized spectral data includes:

[0123] Extract the reference wavelength ranges from the preset element feature wavelength library that correspond to the elements in the optimized spectral data;

[0124] The wavelength parameters of the characteristic spectral peaks are mapped to the preset element characteristic wavelength library to obtain the initial matching pairs of the optimized spectral data;

[0125] Based on the peak shape profile of the characteristic spectral peak, the consistency between the characteristic spectral peak in the initial matching pair and the reference wavelength range is evaluated, and candidate matching pairs with matching peak shapes are selected from the initial matching pairs.

[0126] The signal-to-noise ratio of the candidate matching pairs is evaluated, and the signal amplitude of the characteristic spectral peak in the high-confidence candidate matching pairs is extracted as the characteristic peak intensity of the element.

[0127] Extract a pre-defined elemental characteristic wavelength library, which stores the characteristic wavelength ranges of various elements determined through extensive experiments and standard data. Clearly identify and optimize the elements contained in the spectral data. For each element, precisely locate the corresponding entry in the pre-defined elemental characteristic wavelength library and extract the reference wavelength range marked under that entry. This range represents the regular occurrence interval of the element's characteristic spectral peaks, ensuring that a unique reference wavelength range can be extracted for each element.

[0128] All characteristic spectral peaks in the optimized spectral data are organized and their wavelength parameters are recorded. The wavelength parameters of each characteristic spectral peak are compared one by one with the reference wavelength range of each element in the preset element characteristic wavelength library. When the wavelength parameter of a characteristic spectral peak falls within the reference wavelength range of an element, a matching relationship is formed between the characteristic spectral peak and the corresponding element. All such matching relationships are summarized to form the initial matching pairs of the optimized spectral data. Each initial matching pair is clearly associated with a characteristic spectral peak and an element.

[0129] For each initial matching pair, the peak shape profile of the characteristic spectral peak is extracted, including details such as peak width, height variation trend, and symmetry. Standard peak shape profiles within the reference wavelength range of the corresponding element are retrieved from a preset elemental characteristic wavelength library. A comprehensive comparison is made between the peak shape profile of the characteristic spectral peak and the standard peak shape profile, considering the positional relationship of the starting point, highest point, and ending point, as well as the similarity of the overall shape. If the two peak shape profiles are highly similar, the initial matching pair is determined as a candidate matching pair with matching peak shapes and is retained; if the peak shape differences are significant, they are excluded. Finally, all candidate matching pairs with matching peak shapes are selected.

[0130] For each candidate matching pair, the signal and noise components of its characteristic spectral peaks are determined, and the ratio of the average intensity of the signal component to the average intensity of the noise component is calculated to obtain the signal-to-noise ratio (SNR). A preset SNR threshold based on the reliability requirements of spectral detection is established. When the SNR of a candidate matching pair is higher than this threshold, it is determined to be a high-confidence candidate matching pair. The maximum signal intensity value of the characteristic spectral peak in each high-confidence candidate matching pair is extracted, i.e., the signal amplitude, as the characteristic peak intensity of the corresponding element. The characteristic peak intensities of all elements are summed to form a complete set of elemental characteristic peak intensities.

[0131] The beneficial effects are that the reference wavelength range corresponding to the elements in the spectral data is accurately extracted and optimized from the preset element characteristic wavelength library, providing a clear basis for element matching. The wavelength parameters of the characteristic spectral peaks are mapped to the library to form initial matching pairs. Candidate matching pairs with matching peak shapes are screened through peak shape profile consistency evaluation to ensure the accuracy of the correspondence between elements and characteristic spectral peaks. Then, high-confidence candidate matching pairs are screened through signal-to-noise ratio evaluation and characteristic peak intensities are extracted, effectively eliminating invalid matches and noise interference, and obtaining accurate and reliable element characteristic peak intensity data. This provides a solid foundation of data support for the subsequent establishment of concentration-intensity correspondence and calibration curve construction.

[0132] S5. Extract the elemental concentration data of the titanium alloy sample from the historical detection data, and correlate the elemental concentration data with the characteristic peak intensity to construct the calibration curve of the titanium alloy sample.

[0133] In this embodiment of the invention, the step of extracting elemental concentration data of the titanium alloy sample from historical detection data and correlating the elemental concentration data with the characteristic peak intensity to construct a calibration curve for the titanium alloy sample includes:

[0134] Based on the material grade of the titanium alloy sample, historical test records with the same material grade are retrieved from the historical database of the titanium alloy sample to obtain the initial historical data of the titanium alloy sample.

[0135] By removing historical data whose concentration values ​​exceed the conventional composition range of titanium alloys from the initial historical data, the effective historical data of the titanium alloy sample is obtained.

[0136] The effective historical data is quantified to determine the concentration values ​​of elements in the effective historical data, and the concentration values ​​are integrated into the elemental concentration data of the titanium alloy sample.

[0137] By binding the concentration value with the corresponding characteristic peak intensity, a concentration-intensity correspondence relationship for the titanium alloy sample is generated.

[0138] Based on the concentration-intensity correspondence, the variation trend between the concentration value and the corresponding characteristic peak intensity is analyzed to construct the calibration curve of the titanium alloy sample.

[0139] The formula for calculating the calibration curve is as follows:

[0140] ;

[0141] In the formula, Indicating the first in the optimized spectral data The final concentration of each element, Indicating the first in the optimized spectral data The preset linear fitting slope coefficients for each element. Indicating the first in the optimized spectral data Characteristic peak intensities of the elements Indicates the first of the valid historical data The average characteristic peak intensity of the elements Indicating the first in the optimized spectral data The preset linear fitting intercept coefficients for each element. Indicating the first in the optimized spectral data The preset baseline offset correction coefficient for each element. Indicating the first in the optimized spectral data The degree of shift in the characteristic wavelength of a species. Indicating the first in the optimized spectral data The standard characteristic wavelength of the element Indicating the first in the optimized spectral data The signal-to-noise ratio (SNR) of the characteristic spectral peaks of each element is compared to the reference SNR. Indicating the first in the optimized spectral data The reference signal-to-noise ratio of each element. Indicating the first in the optimized spectral data The preset signal-to-noise ratio weighting coefficients for each element Indicates the first of the valid historical data The average concentration of each element.

[0142] The material grade of the titanium alloy sample is clearly identified, as this is the core identifier distinguishing different types of titanium alloys. The historical database of titanium alloy samples stores past testing records for various material grades. These records contain complete information such as the elemental composition, concentration data of each element, and testing environment for each titanium alloy. Using the identified material grade as the unique search identifier, a comprehensive search is performed in the historical database. The material grade of each historical testing record in the database is compared with the material grade of the target titanium alloy sample. All historical testing records with completely identical material grades are extracted, and all relevant data in these records are summarized to form the initial historical data for titanium alloy samples containing testing information for multiple titanium alloys with the same material grade.

[0143] By consulting industry standards and technical specifications for titanium alloy materials, the conventional composition ranges of common elements in different grades of titanium alloys were determined. These ranges are reasonable intervals determined based on extensive practical testing and material property studies, covering the common concentration values ​​of each element during normal production and use of this type of titanium alloy. Each record in the initial historical data was reviewed one by one, extracting the concentration values ​​of each element in each record and accurately comparing each concentration value with the corresponding conventional composition range. If any element concentration in a record exceeded its corresponding conventional composition range, the entire record was removed from the initial historical data, retaining only historical records where all element concentrations were within their corresponding conventional composition ranges. These retained records together constitute the valid historical data for the titanium alloy samples.

[0144] Each record in the valid historical data is processed individually. Each record contains concentration-related information for multiple elements. The recording formats of this information may differ, so it needs to be uniformly converted into a standardized concentration representation format to ensure consistency in the concentration information of the same element across different records. For each valid historical record, the specific concentration information of each element is extracted sequentially by element type, clarifying the accurate concentration value of each element in that record to avoid omissions or misinterpretations. The concentration values ​​of the same element extracted from all valid historical records are categorized and arranged sequentially according to the order of the detection records, while also distinguishing between the concentration data of different elements. Finally, the concentration values ​​of all elements are systematically integrated to form a clearly structured and well-categorized elemental concentration data for titanium alloy samples.

[0145] The characteristic spectral peaks of the previously obtained optimized spectral data are retrieved to identify the specific element corresponding to each peak. This is achieved by matching the wavelength position of the characteristic spectral peak with the characteristic spectral wavelengths of known elements to determine the elemental type of each peak. Each concentration value of each element in the elemental concentration data is then mapped one-to-one with the intensity data of its corresponding characteristic spectral peak. This means that a specific concentration value of each element corresponds to the specific intensity of its characteristic peak under the given detection conditions. A dedicated correspondence record is established for each element, clearly indicating the concentration value and the corresponding characteristic peak intensity. All element correspondence records are then compiled and integrated to generate the concentration-intensity correspondence for the titanium alloy sample.

[0146] The concentration-intensity relationship is broken down by element type, and the variation between the concentration value and the corresponding characteristic peak intensity is analyzed separately for each element. Starting from the characteristic peak intensity corresponding to the lowest concentration value of the element, the corresponding changes in characteristic peak intensity are observed as the concentration value gradually changes. It is determined whether the characteristic peak intensity increases, decreases, or remains stable as the concentration value increases, while also paying attention to the smoothness and regularity of the change process to clarify the overall trend between the two. All concentration values ​​of the element are used as the x-axis data, and the corresponding characteristic peak intensities as the y-axis data. The coordinate points of each concentration value and corresponding intensity are marked on the coordinate plane in ascending order of concentration value. All coordinate points are connected sequentially with a smooth line, ensuring that the line accurately reflects the distribution pattern of each point and fully reflects the variation relationship between concentration value and characteristic peak intensity. This line is the calibration curve for the corresponding element of the titanium alloy sample. The calibration curves of all elements together constitute the calibration curve for the titanium alloy sample.

[0147] By analyzing the first effective historical data The characteristic peak intensities and concentrations of the elements are linearly fitted, and the preset linear fitting slope coefficients are calculated. These are the preset linear fitting intercept coefficients obtained through the same linear fitting process. Based on the optimized spectral data The offset of the characteristic wavelengths of each element, combined with the baseline compensation adjustment results, determines the preset baseline offset correction coefficient; By comparing and optimizing the spectral data, the first Characteristic wavelengths and standard characteristic wavelengths of various elements The degree of characteristic wavelength shift obtained from the analysis; The standard database of elemental spectra of titanium alloys is the first The standard characteristic wavelength of a certain element. It is the first in the measurement and optimization of spectral data The signal-to-noise ratio of the characteristic spectral peaks of each element is calculated after removing the signal and noise, and the reference signal-to-noise ratio is obtained. The first in the valid historical data The reference signal-to-noise ratio is obtained by averaging the signal-to-noise ratios of the characteristic peaks of the elements. It is based on the first The importance of each element in titanium alloys and the stability of spectral detection are determined by the preset signal-to-noise ratio weighting coefficients. The first in the valid historical data The average concentration is obtained by averaging the concentration values ​​of the elements. The first peak extracted from the characteristic spectral peaks of optimized spectral data Characteristic peak intensities of each element; The first in the valid historical data The average characteristic peak intensity is obtained by averaging the characteristic peak intensities of the elements.

[0148] This formula is used to calculate the first [value] in optimized spectral data. final concentration of the elements It first passed through , for characteristic peak intensity Linear fitting is performed to obtain an intensity-based concentration estimate; then... , and The baseline offset is corrected to adjust the concentration deviation caused by the wavelength shift; then, through... , and The influence of signal-to-noise ratio is adjusted by weighting, and the impact of signal quality on concentration calculation is considered; finally, the average concentration of the element in the valid historical data is added. By comprehensively considering the average level of historical data, the first [number] can be accurately obtained. The final concentration of each element.

[0149] When the characteristic peak intensity When it increases, Partial increase, final concentration Consequently, the degree of characteristic wavelength shift increases; When it increases, Partial enlargement, It also increases accordingly; when the signal-to-noise ratio Relative to reference signal-to-noise ratio When it increases, Partial enlargement, The same increase; while the average concentration of effective historical data... As a base value, it is superimposed on the preceding correction term, and by taking into account the combined effect of all factors, the final concentration is... It can accurately and stably reflect the actual concentration of elements in optimized spectral data.

[0150] The beneficial effects are as follows: based on the material grade of the titanium alloy sample, historical test records of the same grade are retrieved to obtain initial historical data; invalid data with concentration values ​​exceeding the conventional composition range of titanium alloys are removed to obtain valid historical data; the concentration values ​​of each element are clearly defined through concentration quantification and integrated into elemental concentration data; the concentration values ​​are precisely bound to the corresponding characteristic peak intensities to generate a concentration-intensity correspondence; based on this correspondence, the changing trends of the two are analyzed to construct a calibration curve, providing a scientific and reliable reference standard for the accurate detection of the concentration of each element in the titanium alloy sample, ensuring the accuracy and consistency of subsequent test results.

[0151] S6. Map the optimized spectral data to the calibration curve to generate the composition ratio data of the titanium alloy sample.

[0152] In this embodiment of the invention, mapping the optimized spectral data to the calibration curve to generate the compositional ratio data of the titanium alloy sample includes:

[0153] The optimized spectral data is input into the effective mapping range of the calibration curve as the mapping data to be processed for the calibration curve;

[0154] Abnormal mapping points are removed from the mapping data to be processed to obtain a valid mapping subset of the mapping data to be processed;

[0155] The elemental composition data of the effective mapping subset are fitted using the calibration curve to form the initial composition ratio dataset of the titanium alloy sample.

[0156] The initial composition ratio dataset is normalized to generate the composition ratio data of the titanium alloy sample.

[0157] The effective mapping range of the calibration curve is clearly defined. This range is determined based on the actual distribution range of element concentration values ​​and characteristic peak intensities in valid historical data when constructing the calibration curve, covering the entire range where the two are stably correlated and without abnormal deviations. The characteristic peak intensities corresponding to all characteristic spectral peaks in the optimized spectral data are extracted, and these characteristic peak intensities are mapped one by one to the mapping range corresponding to each element in the calibration curve. This ensures that each characteristic peak intensity falls within the effective mapping range of its corresponding element's calibration curve. All the characteristic peak intensities that meet the requirements and their corresponding element correlation data together constitute the unprocessed mapping data for the calibration curve.

[0158] Each mapping point in the mapping data to be processed is analyzed one by one. Each mapping point contains an element identifier, a characteristic peak intensity, and the corresponding expected concentration mapping value. The characteristic peak intensity of each mapping point is substituted into the calibration curve of the corresponding element to obtain the theoretical concentration value of that mapping point. The difference between the actual expected concentration mapping value and the theoretical concentration value is compared. If the difference exceeds the reasonable range allowed by the calibration curve, the mapping point is determined to be an abnormal mapping point and is directly removed from the mapping data to be processed. All mapping points whose actual expected concentration mapping value differs from the theoretical concentration value within a reasonable range are retained. These mapping points are integrated to form the effective mapping subset of the mapping data to be processed.

[0159] The calibration curve of the previously constructed titanium alloy sample is retrieved, which clearly defines the correspondence between the characteristic peak intensity and concentration of each element. For each mapping point in the effective mapping subset, the corresponding calibration curve is found based on its element identifier. Using the characteristic peak intensity of the mapping point as input information, the unique concentration value corresponding to that intensity is found along the trend of the calibration curve. This concentration value is the composition data of the element corresponding to that mapping point. All mapping points in the effective mapping subset are processed sequentially, collecting all composition data for each element. The data is then categorized and organized by element type to ensure that the composition data for each element is complete and without omission, ultimately forming the initial composition ratio dataset of the titanium alloy sample.

[0160] The initial composition ratio dataset is calculated by summing the composition data of all elements. This sum represents the total content level of all detected elements in the titanium alloy sample. For each element, its composition data is divided by the sum of the composition data of all elements to obtain the proportion of that element in the total content of all detected elements. This proportion calculation is performed for all elements sequentially to ensure that the proportion of each element accurately reflects its relative content in the sample. All element proportion data are then arranged in order of element type to form a clearly structured composition ratio dataset for the titanium alloy sample with well-defined proportions. The sum of the proportions of all elements represents the overall composition.

[0161] The beneficial effects are that by inputting optimized spectral data into the effective mapping range of the calibration curve to obtain the mapping data to be processed, removing abnormal mapping points to obtain the effective mapping subset, fitting the initial component ratio dataset through the calibration curve, and generating component ratio data through normalization, the accuracy of data mapping is ensured, abnormal interference is removed, and the component ratio intuitively reflects the relative content of each element in the titanium alloy sample, providing accurate, reliable and easy-to-read results for the component analysis of titanium alloy samples.

[0162] like Figure 2 The diagram shown is a functional block diagram of a titanium alloy composition determination system based on a spectrometer provided in an embodiment of the present invention.

[0163] The spectrometer-based titanium alloy composition determination system 100 of this invention can be installed in an electronic device. Depending on the functions implemented, the spectrometer-based titanium alloy composition determination system 100 may include a spectral data denoising module 101, a baseline calibration module 102, a characteristic peak extraction module 103, a peak intensity determination module 104, a calibration curve construction module 105, and a composition ratio generation module 106. The module described in this invention can also be referred to as a unit, which refers to a series of computer program segments that can be executed by the processor of an electronic device and can perform a fixed function, and are stored in the memory of the electronic device.

[0164] In this embodiment, the functions of each module / unit are as follows:

[0165] The spectral data denoising module 101 is used to acquire the original spectral data of the titanium alloy sample using a spectrometer, and filter the high-frequency noise of the original spectral data to obtain clean spectral data of the titanium alloy sample.

[0166] The baseline calibration module 102 is used to detect local minima in the clean spectral data to construct the baseline profile of the titanium alloy sample, and to perform drift calibration on the baseline profile to obtain optimized spectral data of the titanium alloy sample.

[0167] The feature peak extraction module 103 is used to compare the feature wavelength height of adjacent elements according to the feature wavelength of the elements in the optimized spectral data, obtain the spectral peak data of the optimized spectral data, and integrate the spectral peak data into the feature spectral peaks of the optimized spectral data.

[0168] The peak intensity determination module 104 is used to perform similarity matching between the characteristic spectral peak and the wavelength in the preset element characteristic wavelength library to determine the characteristic peak intensity of the element in the optimized spectral data.

[0169] The calibration curve construction module 105 is used to extract the element concentration data of the titanium alloy sample from historical detection data, and associate the element concentration data with the characteristic peak intensity to construct the calibration curve of the titanium alloy sample.

[0170] The component ratio generation module 106 is used to map the optimized spectral data to the calibration curve to generate the component ratio data of the titanium alloy sample.

[0171] In the several embodiments provided by this invention, it should be understood that the disclosed methods and systems can be implemented in other ways. For example, the system embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and other division methods may be used in actual implementation.

[0172] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0173] Furthermore, the functional modules in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or in the form of hardware plus software functional modules.

[0174] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0175] This application embodiment can acquire and process relevant data based on artificial intelligence technology. Artificial intelligence is the theory, method, technology, and application system that uses digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use that knowledge to obtain optimal results.

[0176] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for determining the composition of titanium alloys based on a spectrometer, characterized in that, The method includes: S1. Collect the raw spectral data of the titanium alloy sample using a spectrometer, and filter the high-frequency noise of the raw spectral data to obtain the clean spectral data of the titanium alloy sample. S2. Detect local minima in the clean spectral data to construct the baseline profile of the titanium alloy sample, and perform drift calibration on the baseline profile to obtain the optimized spectral data of the titanium alloy sample. S3. Based on the characteristic wavelengths of the elements in the optimized spectral data, compare the characteristic wavelength heights of adjacent elements to obtain the spectral peak data of the optimized spectral data, and integrate the spectral peak data into the characteristic spectral peaks of the optimized spectral data. S4. Match the characteristic spectral peaks with wavelengths in a preset element characteristic wavelength library to determine the intensity of the characteristic peaks of the elements in the optimized spectral data; S5. Extract the elemental concentration data of the titanium alloy sample from the historical detection data, and correlate the elemental concentration data with the characteristic peak intensity to construct the calibration curve of the titanium alloy sample. S6. Map the optimized spectral data to the calibration curve to generate the composition ratio data of the titanium alloy sample.

2. The method for determining the composition of titanium alloys based on a spectrometer as described in claim 1, characterized in that, The process of acquiring raw spectral data of the titanium alloy sample using a spectrometer and filtering high-frequency noise from the raw spectral data to obtain clean spectral data of the titanium alloy sample includes: Raw spectral data of titanium alloy samples were acquired using a spectrometer; The original spectral data is divided into data segments according to wavelength order to obtain a data segment sequence of the original spectral data; Extract the noise fluctuation characteristics of the data segment sequence, and determine the noise intensity level corresponding to the data segment sequence based on the noise fluctuation characteristics; Select the noise suppression strategy corresponding to the noise intensity level from the preset filtering strategies; Based on the noise suppression strategy, data segments with different noise intensity levels in the data segment sequence are differentially targeted and filtered out to obtain the denoised data segments of the data segment sequence. The denoised data segments are then boundary aligned to obtain clean spectral data of the titanium alloy sample.

3. The method for determining the composition of titanium alloys based on a spectrometer as described in claim 1, characterized in that, The process of detecting local minima in the clean spectral data to construct the baseline profile of the titanium alloy sample, and performing drift calibration on the baseline profile to obtain optimized spectral data of the titanium alloy sample includes: The clean spectral data is scanned point by point according to a sliding window of fixed width, and the point with the lowest signal intensity within the sliding window is taken as a candidate local minimum point to form a candidate point set of the clean spectral data. By removing outliers from the candidate point set that deviate from the main distribution area, the reference points of the clean spectral data are obtained; Curve fitting is performed on the reference points according to the wavelength coordinate order to generate the initial baseline profile of the clean spectral data; Assess the degree of deviation of the characteristic wavelength positions of the initial baseline profile relative to the overall trend; Based on the degree of offset, the initial baseline profile is adjusted in segments to obtain optimized spectral data of the titanium alloy sample.

4. The method for determining the composition of titanium alloys based on a spectrometer as described in claim 3, characterized in that, The step of performing segmented compensation adjustment on the initial baseline profile based on the degree of offset to obtain optimized spectral data of the titanium alloy sample includes: Based on the degree of deviation of the characteristic wavelength position, the initial baseline profile is divided into continuous adjustment segments; The compensation direction and magnitude of the adjustment segment are determined based on the average deviation direction and magnitude of the characteristic wavelength points in the adjustment segment. Based on the compensation direction and the amplitude, directional amplitude compensation is performed on the baseline contour points in the adjustment section to obtain the compensation section of the initial baseline contour; At the boundary of adjacent adjustment sections, the contour points of the compensation section are smoothed to obtain optimized spectral data of the titanium alloy sample.

5. The method for determining the composition of titanium alloys based on a spectrometer as described in claim 1, characterized in that, The step of comparing the characteristic wavelength heights of adjacent elements based on the characteristic wavelengths of elements in the optimized spectral data to obtain the spectral peak data of the optimized spectral data, and integrating the spectral peak data into the characteristic spectral peaks of the optimized spectral data, includes: Identify the spectral peak points of elements in the optimized spectral data to form an initial set of spectral peak points for the optimized spectral data; Using the spectral peaks in the initial set of spectral peaks as centers, construct a neighborhood relation matrix for the optimized spectral data; Based on the neighborhood relation matrix, the intensity gradient changes between adjacent spectral peak points are analyzed to establish the correlation between the spectral peak points; Based on the aforementioned correlation, spectral peaks with continuously changing characteristics are merged into a set of characteristic spectral peaks of the optimized spectral data. Eliminate the repetitive peak structures in the set of characteristic peaks to form the characteristic peaks of the optimized spectral data.

6. The method for determining the composition of titanium alloys based on a spectrometer as described in claim 1, characterized in that, The step of matching the characteristic spectral peaks with wavelengths in a preset elemental characteristic wavelength library to determine the characteristic peak intensities of elements in the optimized spectral data includes: Extract the reference wavelength ranges from the preset element feature wavelength library that correspond to the elements in the optimized spectral data; The wavelength parameters of the characteristic spectral peaks are mapped to the preset element characteristic wavelength library to obtain the initial matching pairs of the optimized spectral data; Based on the peak shape profile of the characteristic spectral peak, the consistency between the characteristic spectral peak in the initial matching pair and the reference wavelength range is evaluated, and candidate matching pairs with matching peak shapes are selected from the initial matching pairs. The signal-to-noise ratio of the candidate matching pairs is evaluated, and the signal amplitude of the characteristic spectral peak in the high-confidence candidate matching pairs is extracted as the characteristic peak intensity of the element.

7. The method for determining the composition of titanium alloys based on a spectrometer as described in claim 1, characterized in that, The step of extracting elemental concentration data from historical detection data of the titanium alloy sample and correlating the elemental concentration data with the characteristic peak intensity to construct a calibration curve for the titanium alloy sample includes: Based on the material grade of the titanium alloy sample, historical test records with the same material grade are retrieved from the historical database of the titanium alloy sample to obtain the initial historical data of the titanium alloy sample. By removing historical data whose concentration values ​​exceed the conventional composition range of titanium alloys from the initial historical data, the effective historical data of the titanium alloy sample is obtained. The effective historical data is quantified to determine the concentration values ​​of elements in the effective historical data, and the concentration values ​​are integrated into the elemental concentration data of the titanium alloy sample. By binding the concentration value with the corresponding characteristic peak intensity, a concentration-intensity correspondence relationship for the titanium alloy sample is generated. Based on the concentration-intensity correspondence, the variation trend between the concentration value and the corresponding characteristic peak intensity is analyzed to construct the calibration curve of the titanium alloy sample.

8. The method for determining the composition of titanium alloys based on a spectrometer as described in claim 7, characterized in that, The formula for calculating the calibration curve is as follows: ； In the formula, Indicating the optimized spectral data, the first... The final concentration of each element, Indicating the first in the optimized spectral data The preset linear fitting slope coefficients for each element. Indicating the optimized spectral data, the first... Characteristic peak intensities of the elements Indicates the first of the valid historical data The average characteristic peak intensity of the elements Indicating the optimized spectral data, the first... The preset linear fitting intercept coefficients for each element. Indicating the first in the optimized spectral data The preset baseline offset correction coefficient for each element. Indicating the optimized spectral data, the first... The degree of shift in the characteristic wavelength of a species. Indicating the first in the optimized spectral data The standard characteristic wavelength of the element Indicating the first in the optimized spectral data The signal-to-noise ratio (SNR) of the characteristic spectral peaks of each element is compared to the reference SNR. Indicating the optimized spectral data, the first... The reference signal-to-noise ratio of each element. Indicating the first in the optimized spectral data The preset signal-to-noise ratio weighting coefficients for each element Indicates the first of the valid historical data The average concentration of each element.

9. The method for determining the composition of titanium alloys based on a spectrometer as described in claim 1, characterized in that, The step of mapping the optimized spectral data to the calibration curve to generate the composition ratio data of the titanium alloy sample includes: The optimized spectral data is input into the effective mapping range of the calibration curve as the mapping data to be processed for the calibration curve; Abnormal mapping points are removed from the mapping data to be processed to obtain a valid mapping subset of the mapping data to be processed; The elemental composition data of the effective mapping subset are fitted using the calibration curve to form the initial composition ratio dataset of the titanium alloy sample. The initial composition ratio dataset is normalized to generate the composition ratio data of the titanium alloy sample.

10. A system for determining the composition of titanium alloys based on a spectrometer, characterized in that, For implementing the spectrometer-based method for determining the composition of titanium alloys according to any one of claims 1-9, the system comprises: The spectral data denoising module is used to acquire the original spectral data of the titanium alloy sample using a spectrometer, and filter the high-frequency noise of the original spectral data to obtain clean spectral data of the titanium alloy sample. The baseline calibration module is used to detect local minima in the clean spectral data to construct the baseline profile of the titanium alloy sample, and to perform drift calibration on the baseline profile to obtain the optimized spectral data of the titanium alloy sample. The feature peak extraction module is used to compare the feature wavelength heights of adjacent elements based on the feature wavelengths of the elements in the optimized spectral data, obtain the spectral peak data of the optimized spectral data, and integrate the spectral peak data into the feature spectral peaks of the optimized spectral data. The peak intensity determination module is used to perform similarity matching between the characteristic spectral peaks and wavelengths in a preset element characteristic wavelength library to determine the characteristic peak intensity of elements in the optimized spectral data. The calibration curve construction module is used to extract the elemental concentration data of the titanium alloy sample from historical detection data and correlate the elemental concentration data with the characteristic peak intensity to construct the calibration curve of the titanium alloy sample. The composition ratio generation module is used to map the optimized spectral data to the calibration curve to generate the composition ratio data of the titanium alloy sample.

Images