A method for extracting and analyzing spectral signal features of PDC drill bit cladding process

Abstract

This invention belongs to the field of laser cladding detection technology, specifically relating to a method for extracting and analyzing spectral signal features during the cladding process of PDC drill bits. The method includes: acquiring a real-time plasma spectral sequence of the PDC drill bit cladding process using a fiber optic probe and spectrometer; obtaining spectral line effectiveness weights based on the intensity and signal-to-noise ratio of characteristic spectral lines in the plasma spectral sequence; obtaining elemental fusion state indices based on the spectral line effectiveness weights, the intensity of hard phase spectral lines, and the intensity of binder phase spectral lines; obtaining cladding quality stability based on the elemental fusion state index and plasma electron density fluctuations; and determining cladding defects in the PDC drill bit based on the cladding quality stability. This invention effectively suppresses the interference of strong background noise on the monitoring results by integrating spectral signal-to-noise ratio weights and electron density fluctuation characteristics, thus improving the accuracy of detecting component segregation and thermodynamic instability in the cladding layer.

Description

Technical Field

[0001] This invention relates to the field of laser cladding detection technology. More specifically, this invention relates to a method for extracting and analyzing spectral signal features during the cladding process of PDC drill bits. Background Technology

[0002] Polycrystalline diamond compact (PDC) drill bits are core rock-breaking tools in oil and gas drilling and geological exploration. Their surface wear resistance directly determines drilling efficiency and cost. In the manufacturing and repair of PDC drill bits, laser cladding technology is often used to prepare high-performance cemented carbide coatings on the drill bit surface to enhance its erosion and wear resistance. The quality of the cladding layer, especially the fusion state of the hard and binder phases and the uniformity of the microstructure, is crucial to drill bit performance. Therefore, real-time and precise quality monitoring of the cladding process is essential.

[0003] In related technologies, for example, Chinese patent document CN106841177A discloses an online defect diagnosis method in the laser cladding process. This method includes: acquiring photo-induced plasma spectral signals through an optical fiber probe, combining elemental characteristic peak parameters, calculating the photo-induced plasma temperature in real time using Boltzmann plotting or multi-spectral line methods, and constructing a time-domain graph of temperature change over time. This method identifies macroscopic manufacturing defects such as depressions, protrusions, or severe oxidation during the cladding process by judging whether there are sharp fluctuations such as steep drops or rises in the temperature time-domain curve.

[0004] However, related technologies mainly rely on a single thermodynamic parameter and its macroscopic fluctuations to characterize cladding quality, which has significant limitations for composite cladding layers such as PDC drill bits, which consist of hard and binder phases. On the one hand, the macroscopic stability of plasma temperature cannot directly reflect the proportion of microscopic components inside the cladding layer, making it difficult to identify component segregation, local agglomeration, or incomplete fusion defects caused by uneven mixing of the hard and binder phases. On the other hand, relying solely on abrupt changes in the temperature curve is insufficient to capture thermodynamic disturbances inside the molten pool, such as keyhole instability or transient changes in electron density, resulting in the inability to effectively detect potential quality hazards that have not yet formed macroscopic morphological defects. Summary of the Invention

[0005] To address the technical problems mentioned above, such as the reliance on a single monitoring indicator, the inability to effectively monitor microscopic compositional segregation and thermodynamic instability of the cladding layer, and the susceptibility to strong background noise interference leading to low detection accuracy, this invention provides a method for extracting and analyzing spectral signal features during the cladding process of PDC drill bits. The method includes: acquiring a real-time plasma spectral sequence of the PDC drill bit cladding process using an optical fiber probe and a spectrometer; obtaining the signal-to-noise ratio (SNR) based on the peak intensity of characteristic spectral lines and the intensity of surrounding background noise in the real-time plasma spectral sequence; obtaining the spectral line effectiveness weight based on the SNR and a preset SNR benchmark threshold; and obtaining the hard surface features based on the real-time plasma spectral sequence. The net characteristic spectral intensity of the main phase element tungsten, the net characteristic spectral intensity of the binder phase element nickel, and the total intensity of the full-band spectrum are used to obtain the elemental fusion state index based on the net characteristic spectral intensity of the main phase element tungsten, the net characteristic spectral intensity of the binder phase element nickel, the total intensity of the full-band spectrum, and the spectral line effectiveness weight. The matrix element spectral lines are determined, and the plasma electron density is obtained based on the full width at half maximum (FWHM) of the matrix element spectral lines. An electron density sequence is constructed, and the cladding quality stability is obtained based on the standard deviation and mean of the electron density sequence and the elemental fusion state index at historical times. Defects in the cladding process are determined based on the cladding quality stability.

[0006] This invention acquires plasma spectra in real time using fiber optic probes, constructs dynamic spectral effectiveness weights based on the signal-to-noise ratio of characteristic spectral lines, and calculates the elemental fusion state index by combining the spectral intensity ratio of the hard phase and the binder phase with the total intensity of the full-band spectrum. Simultaneously, it uses Stark broadening to invert plasma electron density and statistically analyzes the fluctuation characteristics of the electron density sequence. Finally, it combines the elemental fusion state index at historical moments to generate cladding quality stability to determine defects. This invention reduces the interference of strong background noise and stray light on spectral feature extraction by utilizing spectral effectiveness weights. It quantifies the uniformity of the cladding layer composition by using the net characteristic spectral intensity ratio of the main element tungsten in the hard phase to the main element nickel in the binder phase. It sensitively captures thermodynamic anomalies caused by keyhole effect instability or powder flow disturbance by using the standard deviation and mean of the electron density sequence. Therefore, it achieves the detection of defects such as component segregation and thermodynamic instability in the cladding layer of PDC drill bits under strong noise and complex operating conditions, improving the monitoring accuracy and product quality of the PDC drill bit production process.

[0007] Preferably, the real-time plasma spectral sequence of the PDC drill bit cladding process obtained by the fiber optic probe and spectrometer includes: performing baseline removal processing on the acquired raw spectral data to eliminate ambient stray light and blackbody radiation background, and extracting effective spectral data containing characteristic bands of tungsten, nickel, iron and carbon elements to construct a plasma spectral sequence.

[0008] This invention removes the baseline from the acquired raw spectral data and extracts characteristic bands containing elements such as tungsten and nickel. This eliminates the masking effect of broadband background noise generated by ambient stray light and high-temperature blackbody radiation from the molten pool on characteristic spectral lines. At the same time, it focuses on the effective spectral data of key elements, reduces the computational redundancy of invalid data, and improves the signal-to-noise ratio of real-time plasma spectral sequences and the accuracy of subsequent feature extraction.

[0009] Preferably, the step of obtaining the signal-to-noise ratio based on the peak intensity of the characteristic spectral lines and the intensity of the neighborhood background noise in the real-time plasma spectral sequence includes: obtaining the peak intensity of each characteristic spectral line; selecting a flat spectral region on both sides of the peak of the characteristic spectral line that does not contain other characteristic peaks; taking the average spectral intensity in this region as the neighborhood background noise intensity of the characteristic spectral line; and determining the signal-to-noise ratio of each characteristic spectral line based on the ratio of the peak intensity to the neighborhood background noise intensity.

[0010] Preferably, the spectral line effectiveness weights satisfy the following relationship: In the formula, For a moment Spectral line validity weights, This represents the total number of characteristic spectral lines selected. For the first The characteristic spectral lines at time The signal-to-noise ratio, The preset signal-to-noise ratio (SNR) benchmark threshold. It is the hyperbolic tangent function.

[0011] This invention assigns a positive weight close to 1 to high-quality signals with a signal-to-noise ratio (SNR) higher than the SNR reference threshold, while rapidly reducing the weight or even assigning negative values ​​to low-quality signals with an SNR lower than the SNR reference threshold. This achieves the preservation of high-quality spectral features and the automatic suppression of noise-overwhelmed spectral lines, avoiding the misleading influence of low-quality data on subsequent fusion state assessment.

[0012] Preferably, the step of obtaining the net characteristic spectral intensity of tungsten, the main element of the hard phase, and the net characteristic spectral intensity of nickel, the main element of the binder phase, based on the real-time plasma spectral sequence includes: constructing a search window centered on a pre-calibrated characteristic wavelength, taking the maximum spectral intensity within the search window as the peak intensity, and subtracting the average spectral intensity at the edge of the search window from the peak intensity to complete background subtraction, thereby obtaining the net characteristic spectral intensity of tungsten, the main element of the hard phase, and the net characteristic spectral intensity of nickel, the main element of the binder phase.

[0013] Preferably, the method for obtaining the total intensity of the full-band spectrum includes: using the integral value of the spectral intensities of all bands in the plasma spectral sequence at each time moment as the total intensity of the full-band spectrum.

[0014] Preferably, the element fusion state index satisfies the following relationship: In the formula, For a moment The fusion state index of elements, For a moment The net characteristic spectral intensity of tungsten, the main element in the hard phase. For a moment The net characteristic spectral intensity of nickel, the main element of the binder phase. The preset standard melt ratio value, The allowable standard deviation of the ratio fluctuation, For a moment The total intensity of the full-band spectrum, For a moment Spectral line validity weights, It is an exponential function with the natural constant as the base.

[0015] This invention calculates the absolute value of the difference between the net characteristic spectral line intensity ratio of the hard phase main element tungsten and the binder phase main element nickel and the standard fusion ratio. This results in a high output value when the intensity ratio is close to the standard fusion ratio, and a rapid decrease in the output value once segregation causes the intensity ratio to drift. At the same time, it combines the spectral line effectiveness weight and the total intensity of the full-band spectrum for correction, thereby achieving the characterization of the microstructure uniformity and composition ratio of the cladding layer.

[0016] Preferably, obtaining the plasma electron density based on the full width at half maximum (FWHM) of the matrix element spectral lines includes: fitting the spectral profile of the matrix element spectral lines with a Lorentz function to obtain the full WHM of the fitted curve; and obtaining the plasma electron density based on the Stark broadening principle using the power function relationship between electron density and full WHM.

[0017] Preferably, the stability of the cladding quality satisfies the following relationship: In the formula, For a moment The stability of cladding quality For a moment The standard deviation of the electron density sequence, For a moment The mean of the plasma electron density sequence, For a historic moment The fusion state index of elements, The electron density fluctuation penalty coefficient. The time window length, The time decay factor, It is an exponential function with the natural constant as the base.

[0018] This invention constructs a penalty term reflecting the severity of thermodynamic fluctuations by calculating the standard deviation and mean of the electron density sequence. This allows the numerical value of cladding quality stability to be directly suppressed when there are severe fluctuations in the thermodynamic environment. At the same time, it uses a time decay factor to perform a weighted summation of the element fusion state index at historical moments, giving greater weight to the element fusion state index that is closer to the current moment. This enables a rapid response to current keyhole instability or powder flow disturbance. Furthermore, by smoothing historical data, the influence of instantaneous noise is reduced, accurately reflecting the comprehensive stability of the cladding process.

[0019] Preferably, the defect determination in the cladding process based on the stability of cladding quality includes: when the stability of cladding quality is lower than the quality determination threshold, initiating abnormal duration monitoring; in response to the cladding quality stability being lower than the quality determination threshold for a period exceeding a preset anti-disturbance duration, a defect exists in the current cladding area, and an alarm signal is generated.

[0020] The beneficial effects of this invention are as follows: Utilizing the signal-to-noise ratio characteristics of characteristic spectral lines, this invention identifies and suppresses low-quality spectral data in laser processing environments with high background radiation and dust interference. This blocks the interference of environmental stray light and molten pool blackbody radiation on monitoring accuracy, ensuring the reliability of subsequent characteristic analysis data. By analyzing the intensity ratio of characteristic spectral lines of the hard phase principal elements and the binder phase principal elements, and introducing the total radiation energy across the entire wavelength band as a constraint, this invention achieves the assessment of the uniformity of micro-component distribution and energy coupling state within the cladding layer. This enables the identification of micro-defects such as component segregation, local agglomeration, or lack of fusion. Furthermore, this invention combines the plasma electron density fluctuation characteristics inverted based on the Stark broadening principle to capture the transient thermodynamic changes in molten pool keyhole behavior and powder flow state. This compensates for the shortcomings of single-component monitoring in judging process stability, effectively solving the problem of missed defect detection and false alarms caused by large operating condition fluctuations and strong background noise during PDC drill bit laser cladding. This improves the detection capability for the quality of cemented carbide composite cladding layers and the level of intelligent monitoring of the process. Attached Figure Description

[0021] Figure 1 This is a flowchart illustrating a method for extracting and analyzing spectral signal features during the cladding process of a PDC drill bit according to the present invention;

[0022] Figure 2 This is a schematic diagram illustrating the changes in spectral line effectiveness weight and elemental fusion state index in this invention;

[0023] Figure 3 This is a schematic diagram illustrating the variation in cladding quality stability in this invention. Detailed Implementation

[0024] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0025] The specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0026] This invention discloses a method for extracting and analyzing spectral signal features during the cladding process of PDC drill bits, referring to... Figure 1 This includes steps S1 to S5:

[0027] S1. Obtain the real-time plasma spectral sequence of the PDC drill bit cladding process using an optical fiber probe and spectrometer.

[0028] Specifically, a fiber optic probe is deployed next to the laser cladding station of the PDC drill bit, and the probe angle is adjusted so that its field of view covers the plasma plume region where the laser and powder interact. A spectrometer connected to the fiber optic probe continuously acquires spectral data during the cladding process at a preset sampling frequency. The acquired raw spectral data undergoes baseline removal processing to eliminate ambient stray light and blackbody radiation background, and effective spectral data containing characteristic bands of tungsten, nickel, iron, and carbon are extracted to construct a plasma spectral sequence for each moment.

[0029] For example, the preset sampling frequency is 100Hz.

[0030] S2. Obtain the spectral line effectiveness weight based on the intensity and signal-to-noise ratio of the characteristic spectral lines in the plasma spectral sequence.

[0031] It should be noted that the concentration of metal vapor generated during the PDC drill bit cladding process varies drastically with the processing position, which may cause characteristic spectral lines to be submerged in strong background noise at certain moments. Directly using low-quality spectral lines for subsequent analysis would introduce certain calculation errors. To ensure the accuracy of subsequent fusion state assessment, this invention obtains spectral line validity weights based on the intensity and signal-to-noise ratio of characteristic spectral lines in the plasma spectral sequence.

[0032] Specifically, in the plasma spectral sequence at each time step, characteristic spectral lines of tungsten and nickel are extracted according to a pre-defined list of characteristic wavelengths. The peak intensity of each characteristic spectral line is obtained, and a flat spectral region of a preset width, containing no other characteristic peaks, is selected on both sides of the peak. The average spectral intensity within this region is taken as the neighborhood background noise intensity of that spectral line. The signal-to-noise ratio (SNR) of each characteristic spectral line is determined based on the ratio of the peak intensity to the neighborhood background noise intensity. The spectral line validity weight at each time step is obtained based on the SNR of all characteristic spectral lines of the same element and a preset SNR benchmark threshold.

[0033] For example, the list of characteristic wavelengths includes: tungsten atom spectral line wavelengths of 400.88 nm and 429.46 nm, and nickel atom spectral line wavelengths of 341.48 nm and 352.45 nm, with a preset width of 1 nm and a signal-to-noise ratio reference threshold of 5.

[0034] Specifically, the spectral line validity weights satisfy the following relationship:

[0035] ;

[0036] In the formula, For a moment Spectral line validity weights, This represents the total number of characteristic spectral lines selected. For the first The characteristic spectral lines at time The signal-to-noise ratio, The preset signal-to-noise ratio (SNR) threshold is set to 5 in this embodiment. It is the hyperbolic tangent function.

[0037] in, This represents the quality of the spectral line signal relative to a reference standard. A larger value indicates a clearer signal, more reliable feature information, and a greater weighting for spectral line validity; a smaller value indicates a less clear signal, less reliable feature information, and a smaller weighting for spectral line validity. It plays a role in adjusting the weight values ​​based on signal reliability; when Significantly higher than This indicates that the current spectral features are clear and the background interference is low, leading to... The value is positive and approaches 1, thus maintaining the high weight value calculated by the preceding terms, allowing high-quality signals to participate in subsequent calculations; when Significantly lower than This indicates that the current spectral features are weak or have been drowned out by noise. The value quickly drops to 0 or even turns negative, thus lowering the final spectral validity weight and eliminating the influence of unreliable signals.

[0038] S3. Obtain the element fusion state index based on the spectral line effectiveness weight, the intensity of the hard phase spectral line, and the intensity of the binder phase spectral line.

[0039] It should be noted that the quality of the cladding layer of the PDC drill bit largely depends on the fusion ratio and distribution uniformity of the hard phase and the binder phase. When the cladding process is normal, the ratio of the two phases entering the plasma state should remain relatively stable. If agglomeration or lack of fusion occurs, the ratio of their spectral intensities will drift drastically. Therefore, this invention obtains the elemental fusion state index based on the spectral line effectiveness weight, the spectral line intensity of the hard phase, and the spectral line intensity of the binder phase.

[0040] Specifically, a search window of a preset length is constructed centered on a pre-calibrated characteristic wavelength. The maximum spectral intensity within the window is taken as the peak intensity, and the average spectral intensity at the edge of the window is subtracted from the peak intensity to complete background subtraction, thereby obtaining the net characteristic spectral line intensity of tungsten, the main element of the hard phase, and the net characteristic spectral line intensity of nickel, the main element of the binder phase. At the same time, the integral value of the spectral intensity of all bands in the plasma spectral sequence at each time moment is taken as the total intensity of the full-band spectrum. The elemental fusion state index at each time moment is obtained based on the net characteristic spectral line intensity of tungsten, the main element of the hard phase, the net characteristic spectral line intensity of nickel, the total intensity of the full-band spectrum, and the spectral line effectiveness weight.

[0041] For example, the calibrated characteristic wavelength is 400.88 nm and the preset length is 0.5 nm.

[0042] Specifically, the fusion state index of elements satisfies the following relationship:

[0043] ;

[0044] In the formula, For a moment The fusion state index of elements, For a moment The net characteristic spectral intensity of tungsten, the main element in the hard phase. For a moment The net characteristic spectral intensity of nickel, the main element of the binder phase. The preset standard melt ratio value, The allowable standard deviation of the ratio fluctuation, For a moment The total intensity of the full-band spectrum, For a moment Spectral line validity weights, In this embodiment, the function is an exponential function with the natural constant as its base. It is 0.8. It is 0.1.

[0045] in, This represents the instantaneous deviation of the ratio of the spectral intensities of the hard phase to the binder phase from the ideal process reference. The smaller the size, the more likely the tungsten carbide particles are to mix evenly with the nickel-based alloy. The value approaches 1, thus the larger the element fusion state index becomes; The larger the value, the more likely the mixing ratio of tungsten carbide particles to nickel-based alloys is to be unbalanced or segregated, leading to... It approaches 0, thus significantly reducing the element fusion state index. This represents the energy contribution of elements in the cladding material to the overall plasma radiation. A larger value indicates stronger elemental radiation from the cladding material, higher signal confidence, and a higher level of elemental fusion state index. A smaller value indicates that the spectral signal is more likely to contain background radiation or impurity radiation from non-cladding materials, lower signal confidence, and a smaller elemental fusion state index, thus preventing misleadingly high assessment values ​​when the effective signal is weak.

[0046] For example, Figure 2 The graph shows the changes in spectral line effectiveness weight and element fusion state index in this invention. As can be seen from the graph, the spectral line effectiveness weight remains high, close to 1, during the stable processing stage, but fluctuates significantly and drops when the signal quality is disturbed. This indicates that the invention can sense changes in the signal-to-noise ratio and automatically reduce the weight of inferior signals. The element fusion state index oscillates within a relatively high normal range during normal cladding, but once the fusion ratio is imbalanced, the element fusion state index quickly drops to a low value, indicating an abnormality in the metallurgical state of the molten pool and providing a reliable basis for subsequent judgment.

[0047] S4. Obtain the stability of cladding quality based on the element fusion state index and plasma electron density fluctuation.

[0048] It should be noted that drastic fluctuations in plasma electron density usually indicate instability of the keyhole effect (where the recoil pressure generated by the violent evaporation of material on the surface of the molten pool under continuous irradiation by a high-energy laser beam creates vapor pores that enhance energy absorption) or disturbances in the powder flow. Therefore, this invention obtains the stability of cladding quality by combining the temporal elemental fusion state index with fluctuations in plasma electron density.

[0049] Specifically, a matrix element spectral line with an isolated spectral profile, no obvious self-absorption effect, and a known Stark broadening parameter is selected. Its spectral profile is fitted with a Lorentz function to obtain the full width at half maximum (FWHM) of the fitted curve. Based on the Stark broadening principle, the plasma electron density at each time step is obtained using the power function relationship between electron density and FWHM. An electron density sequence is constructed, containing the plasma electron density at each time step and the plasma electron density at each time step within the corresponding time window. The standard deviation and mean of the electron density sequence are obtained. Combined with the elemental fusion state index at each time step, the cladding quality stability at each time step is obtained; where, at time step... The data within the corresponding time window is the time. Previous data.

[0050] For example, nickel spectral lines are used as matrix element spectral lines.

[0051] It should be added that, under the high-temperature and high-energy environment of laser cladding, the microscopic electric field generated by charged particles in the plasma perturbs the energy level structure of the emitting atoms, producing the Stark effect and causing spectral line broadening. Since the line shape of Stark broadening mainly follows the Lorentz distribution, the interference of factors such as instrument broadening can be effectively eliminated by fitting the Lorentz function, accurately extracting the full width at half maximum (FWHM) caused by particle collisions. According to the Stark broadening theory, there is a significant positive power function relationship between the FWHM and the electron density in the plasma. Therefore, this physical model can be used to achieve quantitative inversion of the plasma electron density in the cladding region.

[0052] Specifically, the stability of cladding quality satisfies the following relationship:

[0053] ;

[0054] In the formula, For a moment The stability of cladding quality For a moment The standard deviation of the electron density sequence, For a moment The mean of the plasma electron density sequence, For a historic moment The fusion state index of elements, The electron density fluctuation penalty coefficient. The time window length, The time decay factor, In this embodiment, it is an exponential function with the natural constant as its base. It is 2. It is 10. It is 0.5.

[0055] in, This value represents the constraint of thermodynamic environment stability on quality evaluation. A larger value indicates smaller electron density fluctuations, a more stable cladding process, and thus higher cladding quality stability. Conversely, a smaller value indicates more drastic electron density fluctuations, increasing the likelihood of violent splashing or keyhole collapse. This, in turn, strongly suppresses the final cladding quality stability. In the formula, Divide by It is for adjustment The numerical range of, makes The changes are more flexible.

[0056] Represents the moment The elemental fusion state, which reflects the instantaneous mass of the cladding layer at that time, as well as the energy coupling. Represents a moment Distance Time The time span, when historical moments The closer to the moment Time span The smaller, the more The closer it is to 1, the greater the weight of the recent fusion state in the current stability evaluation, enabling the cladding quality stability to respond quickly to changes in current operating conditions; when historical moments The further away from the current moment Time span The larger it is, the more it leads to The closer the value is to 0, the more it suppresses the interference of long-term historical data on the current evaluation, and avoids misleading judgments on the current cladding quality from outdated status information. Therefore, by... right Adjustments are made to obtain the stability of the cladding quality.

[0057] S5. Determine the cladding defects of PDC drill bits based on the stability of cladding quality.

[0058] Specifically, the stability of the cladding quality is compared with the quality judgment threshold in real time. When the stability of the cladding quality is lower than the quality judgment threshold, monitoring of the abnormal duration is initiated. If the stability of the cladding quality remains lower than the quality judgment threshold for a period exceeding the preset anti-disturbance duration, it is determined that the thermodynamic equilibrium or compositional fusion state of the current cladding area has substantially deteriorated, confirming the existence of a defect in the current cladding area, and generating an alarm signal. The quality judgment threshold is 0.6, and the anti-disturbance duration is 0.5 seconds. Implementers can determine the quality judgment threshold and anti-disturbance duration according to the actual situation.

[0059] For example, Figure 3The graph shows the variation of cladding quality stability in this invention. As can be seen from the graph, during the normal processing stage, the cladding quality stability is always maintained above the judgment threshold. However, during the defect occurrence stage, the cladding quality stability can respond quickly and drop significantly below the threshold, thereby realizing the monitoring of abnormalities in the cladding process.

Claims

1. A method for spectral signal feature extraction and analysis of a PDC bit cladding process, characterized in that, include: Real-time plasma spectral sequences of the PDC drill bit cladding process are acquired using a fiber optic probe and spectrometer. The signal-to-noise ratio (SNR) is obtained based on the peak intensity of characteristic spectral lines and the surrounding background noise intensity. The spectral line validity weight at each moment is obtained based on the SNR of all characteristic spectral lines of the same element and a preset SNR benchmark threshold, including: ； In the formula, For a moment Spectral line validity weight, This represents the total number of characteristic spectral lines selected. For the first The characteristic spectral lines at time signal-to-noise ratio, The preset signal-to-noise ratio (SNR) benchmark threshold. It is the hyperbolic tangent function; The net characteristic line intensities of tungsten, the main element of the hard phase, and nickel, the main element of the binder phase, as well as the total intensity of the full-band spectrum, are obtained based on real-time plasma spectral sequences. The elemental fusion state index is then obtained based on the net characteristic line intensities of tungsten and nickel, the total intensity of the full-band spectrum, and the spectral effectiveness weights. This index includes: ； In the formula, For a moment The elemental melting state index, For a moment The net characteristic spectral intensity of tungsten, the main element in the hard phase. For a moment The net characteristic spectral intensity of nickel, the main element of the binder phase. The preset standard melt ratio value, The allowable standard deviation of the ratio fluctuation, For a moment The total intensity of the full-band spectrum, It is an exponential function with the natural constant as its base; Determine the matrix element spectral lines and obtain the plasma electron density based on the full width at half maximum (FWHM) of the matrix element spectral lines; An electron density sequence is constructed, and the cladding quality stability is obtained based on the standard deviation and mean of the electron density sequence and the elemental fusion state index at historical times, including: ； In the formula, For a moment The stability of cladding quality For a moment The standard deviation of the electron density sequence, For a moment The mean of the plasma electron density sequence, For a historic moment The elemental melting state index, The electron density fluctuation penalty coefficient. The time window length, This is the time decay factor; Defects in the cladding process are determined based on the stability of the cladding quality.

2. The method for extracting and analyzing spectral signal features of the PDC drill bit cladding process according to claim 1, characterized in that, The method of acquiring the real-time plasma spectral sequence of the PDC drill bit cladding process through fiber optic probe and spectrometer includes: performing baseline removal processing on the acquired raw spectral data to eliminate ambient stray light and blackbody radiation background, and extracting effective spectral data containing characteristic bands of tungsten, nickel, iron and carbon elements to construct the plasma spectral sequence.

3. The method for extracting and analyzing spectral signal features of the PDC drill bit cladding process according to claim 1, characterized in that, The step of obtaining the signal-to-noise ratio (SNR) based on the peak intensity of characteristic spectral lines and the intensity of neighborhood background noise in a real-time plasma spectral sequence includes: obtaining the peak intensity of each characteristic spectral line; selecting a flat spectral region on both sides of the peak of the characteristic spectral line that does not contain other characteristic peaks; taking the average spectral intensity in this region as the neighborhood background noise intensity of the characteristic spectral line; and determining the SNR of each characteristic spectral line based on the ratio of the peak intensity to the neighborhood background noise intensity.

4. The method for extracting and analyzing spectral signal features of the PDC drill bit cladding process according to claim 1, characterized in that, The step of obtaining the net characteristic spectral intensity of tungsten, the main element of the hard phase, and nickel, the main element of the binder phase, based on the real-time plasma spectral sequence includes: constructing a search window centered on a pre-calibrated characteristic wavelength; taking the maximum spectral intensity within the search window as the peak intensity; subtracting the average spectral intensity at the edge of the search window from the peak intensity to complete background subtraction, thereby obtaining the net characteristic spectral intensity of tungsten, the main element of the hard phase, and nickel, the main element of the binder phase.

5. The method for extracting and analyzing spectral signal features of the PDC drill bit cladding process according to claim 1, characterized in that, The method for obtaining the total intensity of the full-band spectrum includes: using the integral value of the spectral intensity of all bands in the plasma spectral sequence at each time moment as the total intensity of the full-band spectrum.

6. The method for extracting and analyzing spectral signal features of the PDC drill bit cladding process according to claim 1, characterized in that, The method of obtaining the plasma electron density based on the full width at half maximum (FWHM) of the matrix element spectral lines includes: fitting the spectral profile of the matrix element spectral lines with a Lorentz function to obtain the full WHM of the fitted curve; and obtaining the plasma electron density based on the Stark broadening principle using the power function relationship between electron density and full WHM.

7. The method for extracting and analyzing spectral signal features of the PDC drill bit cladding process according to claim 1, characterized in that, The method of determining defects in the cladding process based on the stability of cladding quality includes: when the stability of cladding quality is lower than the quality determination threshold, initiating abnormal duration monitoring; in response to the time when the stability of cladding quality is lower than the quality determination threshold for a longer than a preset anti-disturbance duration, a defect exists in the current cladding area, and an alarm signal is generated.

Images