A method and system for monitoring the thread accuracy of pipe fittings processed by an automatic threading machine.

Abstract

This invention relates to the field of pipe thread accuracy measurement, specifically to a method and system for monitoring the thread accuracy of pipe fittings processed by an automatic threading machine. First, a laser profile sensor is used to collect surface data of the pipe fitting and remove baseline offset. Then, an impact singularity index based on physical parameters is obtained to locate signal abrupt changes, and cross-correlation analysis is combined to calculate the aliasing risk index of spectral coupling. Based on this, logarithmic scaling logic is used to generate an adaptive decoupling mask amplitude, and sinusoidal mask signal-assisted empirical mode decomposition (EMD) is used to eliminate mode aliasing and reconstruct a clean thread profile. Finally, the major diameter, minor diameter, and pitch diameter of the thread are calculated by extracting the peaks and troughs of the reconstructed profile, and compared with national standard tolerances to achieve automatic judgment of processing quality.

Description

Technical Field

[0001] This invention relates to the field of pipe thread accuracy measurement, and specifically to a method and system for monitoring the thread accuracy of pipe fittings processed by an automatic threading machine. Background Technology

[0002] Automatic threading machines are commonly used in the threading process of pipe fittings such as galvanized steel pipes. The quality of this processing directly affects the sealing and safety of the pipe connections. In modern production, to ensure product qualification rates, high-precision non-contact sensors are typically used to collect real-time data on the pipe fitting surface. The geometric features within this data are then analyzed to achieve online monitoring of thread accuracy.

[0003] However, the cutting environment of automatic threading machines is extremely complex. The acquired raw signals not only contain the ideal thread profile of the pipe fitting itself, but also inevitably contain various interference components, including high-frequency burrs generated by machine tool motor vibration, periodic fluctuations caused by spindle eccentricity, and random sharp protrusions caused by metal filings or tool breakage. To extract the true thread waveform from these non-stationary, nonlinear, and noisy signals, the Empirical Mode Decomposition (EMD) algorithm, as a classic time-frequency analysis tool, is often used for noise reduction and separation of such signals. The EMD algorithm can adaptively decompose the signal into several intrinsic mode functions based on its own time-scale characteristics.

[0004] Applying traditional EMD algorithms directly in automated threading processes has its limitations. Because machining signals are often accompanied by discontinuous, strong impact noise or periodic interference similar to the thread frequency, EMD algorithms are highly susceptible to mode aliasing. This aliasing causes signal components at different time scales to be incorrectly assigned to the same IMF component, or the same component to be dispersed across different IMFs, making it impossible for the algorithm to effectively separate high-frequency noise from the low-frequency thread profile. This not only easily leads to signal distortion after reconstruction but also severely interferes with the subsequent accurate calculation of key geometric parameters such as the major diameter, minor diameter, and pitch diameter of the thread. Summary of the Invention

[0005] To address the aforementioned signal distortion problem after reconstruction, this invention proposes a method for monitoring the thread accuracy of pipe fittings processed by an automatic threading machine in a first aspect. The method includes: acquiring a contour signal of the pipe fitting surface; integrating the contour signal within a set integration interval including the current moment to obtain an impact singularity index; the integration object includes the amplitude at the current moment, the absolute value of the second derivative, and a weighted term inversely proportional to the global standard deviation and directly proportional to the difference between the amplitude and the global mean; calculating a mask signal based on an aliasing risk index and injecting it into the contour signal to obtain an auxiliary signal; performing empirical mode decomposition and reconstruction on the auxiliary signal to obtain a pure thread contour; extracting geometric parameters based on the pure thread contour and comparing them with a set standard to determine whether the thread accuracy is qualified; the amplitude of the mask signal is positively correlated with the standard deviation of the noise component in the contour signal and the logarithm of the aliasing risk index; the aliasing risk index is positively correlated with the impact singularity index, the theoretical thread frequency, and the product of the main vibration frequency, and negatively correlated with the absolute value of the difference between the main vibration frequency and the theoretical thread frequency.

[0006] Existing technologies for processing thread profile signals often rely on filtering or traditional empirical mode decomposition (EMD) methods, which are easily affected by machine tool spindle vibration and cutting impact noise. This leads to mode aliasing between the actual thread frequency and the vibration frequency, making it difficult to extract accurate thread profiles. This invention calculates an impact singularity index, including amplitude, second derivative (curvature), and statistical characteristics, and combines this with the difference between the theoretical frequency and the dominant vibration frequency to obtain an aliasing risk index. Based on this, a mask signal is dynamically generated and injected for auxiliary decomposition. This method can adaptively improve signal decomposition results for specific interference risks under the current machining state, effectively solving the mode aliasing problem. This allows for accurate separation of pure thread profiles even in high-noise environments, improving the robustness and accuracy of thread precision monitoring.

[0007] Furthermore, the specific method for calculating the mask signal is as follows:

[0008] ;

[0009] in This represents the mask signal; This represents the amplitude of the mask signal; and ;in The theoretical thread frequency is determined by the machine tool's set rotational speed. The term "primary vibration frequency" refers to the non-fundamental primary vibration frequency extracted after spectral analysis of the contour signal.

[0010] By setting the frequency of the mask signal to the arithmetic mean of the theoretical thread frequency and the non-fundamental main vibration frequency, unlike the use of white noise or fixed frequency masks in the prior art, this targeted frequency selection can force the EMD algorithm to separate the thread signal and the vibration interference signal into different intrinsic mode functions (IMFs), minimizing the risk of decomposition failure caused by frequency proximity.

[0011] Furthermore, the method for calculating the amplitude of the mask signal is as follows:

[0012] ;

[0013] in This indicates the amplitude of the mask signal; The residual noise standard deviation is represented by the profile signal. The base level of background noise in the current signal is obtained by subtracting its smoothed moving average. Represents the natural constant; This represents the aliasing risk index.

[0014] By combining the residual noise standard deviation and the aliasing risk index to determine the amplitude, it is ensured that the strength of the mask signal is sufficient to help separate high-risk aliasing modes without destroying the original signal characteristics due to excessive amplitude. The introduction of the logarithmic function makes the amplitude adjustment smoother, preventing signal distortion caused by drastic fluctuations in the risk index. Compared with the fixed amplitude masking method, it is more adaptable to the variable noise environment during the processing.

[0015] Furthermore, the method for obtaining the pure thread profile specifically involves: calculating the mask signal. And acquire auxiliary signal pairs: ; ; respectively and Empirical mode decomposition was performed, resulting in two sets of eigenmode functions. , The mean of the eigenmode functions of the corresponding order is taken as the reconstructed component. Adding all the reconstructed components together yields the pure thread profile.

[0016] Furthermore, the calculation method for the aliasing risk index is as follows:

[0017] ;

[0018] in The value represents the aliasing risk index; the higher the value, the higher the probability of error in EMD decomposition. This represents the shock singularity index; This refers to the theoretical thread; Indicates the dominant vibration frequency; Represents phase, with a value range of 1000. ; It represents a very small positive number and is used to prevent the denominator from being zero.

[0019] By comprehensively considering the impact singularity and the proximity of the interference frequency to the target frequency, this formula can assess the probability of errors in EMD decomposition. When the vibration frequency is extremely close to the thread frequency or the signal impact characteristics are significant, the exponent increases significantly, thereby guiding the system to enhance the mask signal strength. This provides a precise mathematical basis for adaptive signal processing, overcoming the blindness of traditional methods that rely on empirical parameter settings.

[0020] Furthermore, the specific method for calculating the impact singularity index is as follows:

[0021] ;

[0022] in The shock singularity index is used to characterize time. The amplitude variation of nearby signal waveforms; Indicates the physical thread period, which is a known physical parameter; Represents the integral variable; The second derivative of the signal, i.e., curvature; and These represent the entire signal segment. The global arithmetic mean and global standard deviation.

[0023] By integrating the second derivative of the signal and weighting it using the degree to which local amplitudes deviate from the global mean, this index can sensitively capture abrupt changes in the thread surface. Compared to simple amplitude detection, this method can more effectively identify high-frequency non-stationary impact components that cause instability in EMD decomposition.

[0024] Furthermore, the specific method for calculating the physical thread period is as follows:

[0025] ;

[0026] in Indicates the physical thread period; This indicates the set spindle speed of the automatic threading machine; This indicates the number of scan lines per unit length of the laser profile sensor.

[0027] Furthermore, extracting geometric parameters also includes: identifying all peaks and troughs in the reconstructed pure thread profile; calculating the average height of all peaks as the major diameter of the thread; and calculating the average height of all troughs as the minor diameter of the thread.

[0028] Furthermore, it also includes comparing the major diameter and minor diameter of the thread with the tolerance zone specified in GB / T 7306 standard. If both are within the tolerance zone, the thread accuracy is deemed qualified.

[0029] In a second aspect, the present invention provides a thread accuracy monitoring system for pipe fittings processed by an automatic threading machine, comprising a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the present invention implements a thread accuracy monitoring method for pipe fittings processed by an automatic threading machine.

[0030] The technical effects of this invention are as follows:

[0031] This invention proposes an adaptive mask signal-assisted EMD thread monitoring method based on an aliasing risk index. Addressing the signal mode aliasing problem caused by vibration and cutting impact during threading, the invention calculates an aliasing risk index by analyzing impact singularity and frequency proximity, and dynamically adjusts the frequency and amplitude of the mask signal accordingly. By injecting this mask signal to assist empirical mode decomposition, machine tool vibration interference and the actual thread profile are effectively separated, solving the problem of low extraction accuracy under strong noise conditions in traditional methods, and achieving high-precision automatic monitoring of pipe thread parameters. Attached Figure Description

[0032] Figure 1 This is a schematic flowchart illustrating a method for monitoring the thread accuracy of pipe fittings processed by an automatic threading machine according to an embodiment of the present invention;

[0033] Figure 2 This is a schematic diagram illustrating the structure of an automatic threading machine for monitoring the thread accuracy of pipe fittings according to an embodiment of the present invention. Detailed Implementation

[0034] 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.

[0035] The specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0036] An example of a method for monitoring the thread accuracy of pipe fittings processed by an automatic threading machine:

[0037] like Figure 1 As shown, the thread accuracy monitoring method for pipe fittings processed by the automatic threading machine of the present invention includes:

[0038] S1. A laser profile sensor is used to collect surface topography data of the pipe fitting and remove baseline offset, thereby obtaining the profile signal to be analyzed, which includes thread periodic features and multi-source noise interference.

[0039] In this embodiment, firstly, during the cutting operation of the automatic threading machine, high-precision non-contact sensors are used to collect the morphology data of the pipe surface and perform basic signal conditioning.

[0040] Specifically, this can be achieved using a laser contour sensor at a high sampling rate, for example... Real-time acquisition of height data from the surface of the pipe fitting. Assume the object being processed is... Galvanized steel pipe, spindle speed Set as The collected raw data includes the geometric profile of the pipe fitting itself and the dynamic fluctuations during processing. To focus on processing quality analysis, further processing of the raw data is also required. Baseline removal is performed to eliminate macroscopic offsets inherent in the pipe fittings, such as low-frequency trends caused by installation eccentricity or pipe bending, thereby obtaining the profile signal to be analyzed. This signal Typically, it is a time series of fixed length (e.g., 2 seconds), whose main component is the frequency of the periodic changes in the corresponding thread profile. The sine wave is superimposed with high-frequency glitch (frequency approximately) generated by the vibration of the machine tool motor. ) and random sharp protrusions caused by flying iron filings or impacts from hard points.

[0041] S2. By integrating the amplitude, second derivative curvature, and statistical characteristics of the contour signal, an impact singularity index is obtained that characterizes the waveform variation and can accurately locate high-risk areas of modal aliasing.

[0042] This step aims to accurately locate the regions causing EMD (Empirical Mode Decomposition) mode aliasing, i.e., the discontinuous impact points in the signal. Therefore, this embodiment no longer relies on manually set thresholds, but proposes an integral calculation logic based on physical parameters.

[0043] Calculate any time interval The shock singularity index is as follows:

[0044] ;

[0045] in The shock singularity index is a dimensionless value used to characterize time. The amplitude variation of nearby signal waveforms; The physical thread cycle (unit: seconds) is a known physical parameter, and its calculation method is as follows: ,in For rotational speed, For the number of sensor scan lines, use As an integration interval, it ensures that the analysis window exactly covers a complete thread profile, thus achieving adaptive window size. Represents the integral variable; The second derivative of a signal, i.e., curvature, is represented by the second derivative of a normal thread profile, which is a smooth arc and has a finite second derivative. However, impact noise is usually sharp and will cause the second derivative to be extremely large. and These represent the entire signal segment. The global arithmetic mean and global standard deviation.

[0046] The terms in the above formula This provides an adaptive weighting effect. The logic of this formula shows that if the vibration amplitude at a certain point far exceeds the global standard deviation, for example, exceeding... The weight term will automatically increase, thereby significantly amplifying the singularity value at that point.

[0047] An exemplary illustration, assuming in A blade struck at that location, creating an area with a radius of only [missing information]. The sharp protrusion, though its amplitude is small on a macroscopic scale, results in a maximum second derivative due to its sharp shape. Calculated using the above formula, The value may change from the normal state. surge to Based on this, it is possible to accurately locate The area is considered a high-risk zone.

[0048] S3. By comprehensively utilizing cross-correlation analysis to evaluate signal morphological similarity and combining it with frequency proximity factors, a spectral coupling aliasing risk index that can quantify the probability of error in empirical mode decomposition and the risk of resonance is calculated.

[0049] Impact alone does not necessarily lead to EMD failure; modal aliasing only occurs when both the impact frequency and the thread frequency become disordered. Therefore, in this embodiment, the risk of aliasing is calculated by combining cross-correlation analysis and frequency comparison.

[0050] The aliasing risk index for spectral coupling is calculated as follows:

[0051] ;

[0052] in This represents the aliasing risk index; the higher the value, the higher the probability of errors in EMD decomposition. Indicates the singularity index of shocks; The theoretical thread frequency (unit: Hz) is determined by the machine tool's set speed. Represents phase, with a value range of 1000. ; The dominant vibration frequency (unit: Hz) is obtained by extracting the frequency of the non-fundamental frequency component with the highest energy after performing an FFT transformation on the original signal. Represents extremely small positive numbers (e.g.) (), used to prevent the denominator from being zero.

[0053] The fractional term on the left side of the formula represents the phase. The ratio of the maximum value of the cross-correlation integral within the range to the energy normalization term is actually used to calculate the singularity distribution. With standard thread cycle The similarity of the shape. If the impact occurs periodically with the thread, such as periodic vibration caused by spindle eccentricity, this value will be large.

[0054] The fractional term on the right side of the formula It is the frequency proximity factor, when the interference frequency... Infinitely close to the thread frequency When the denominator approaches The entire factor tends towards infinity. This reflects the following physical fact: the closer the frequencies are, the more difficult it is to separate them, and the risk of resonance aliasing is extremely high.

[0055] In one possible example, if the rotational speed is ,Right now When the noise source is a high-frequency whine from the motor When the denominator is large, The frequency is very small, and the system determines it to be safe; however, when processing long pipe fittings, low-frequency oscillations occur, with an oscillation frequency of approximately... When, the denominator Frequency proximity factor amplification times, The surge triggered a high-risk resonance effect, which the system identified as a potential threat.

[0056] S4. Based on the residual noise standard deviation and aliasing risk index, the injection intensity is calculated through logarithmic scaling logic to obtain an adaptive decoupling mask amplitude that is adapted to the variable noise environment and proportional to the background noise.

[0057] For the risks identified in step S3, the injection intensity of the mask signal is calculated. In this embodiment, logarithmic scaling logic is used to accommodate a wide dynamic range from weak vibrations to severe flutter.

[0058] The calculation of the mask signal amplitude is as follows: ;in Indicates the intensity of the artificial mask signal to be injected (unit: mm); The residual noise standard deviation (unit: mm) is expressed by the original signal. The result is obtained by subtracting its smoothed moving average, which represents the baseline level of background noise in the current signal. Represents the natural constant, approximately ; This represents the risk gain coefficient.

[0059] use As a baseline, this ensures that the mask signal is always proportional to the current noise level, rather than a fixed value. Simultaneously, the introduction of the logarithmic function provides smooth gain control: when risk... hour, ,at this time That is, injecting a mask with a size equal to the background noise under low-risk conditions to provide basic stability; when the risk is high... Very high, for example, reaching hour, At this point, the injection is stronger than the background noise. A mask of multiples is used to force the aliased modes to separate.

[0060] S5. A sinusoidal mask signal of a specific frequency is injected into the original data to obtain auxiliary signal pairs. After empirical mode decomposition and mean cancellation processing, a pure thread profile that effectively separates machine tool vibration interference is finally reconstructed.

[0061] Based on the information obtained in step S4 The sinusoidal mask signal is determined and the original data is injected. Signal separation is achieved using the frequency filtering characteristics of the EMD. Specific operations include:

[0062] First, calculate the mask signal: ;in, The midpoint frequency between the two frequencies is chosen as the mask frequency to separate the aliased spectrum in the middle.

[0063] Then obtain the positive / negative auxiliary signal pairs: ; Subsequently, respectively and Standard EMD decomposition was performed to obtain two sets of IMF (Intrinsic Mode Function) components;

[0064] Further mask removal and reconstruction are performed: the average value of the corresponding order IMF is taken. That is, mask signal The positive and negative components cancel each other out during addition, preserving the pure component that has been freed from aliasing effects;

[0065] Finally, the filtered low-frequency principal components are superimposed to obtain the denoised, clean thread profile. .

[0066] S6. Identify and reconstruct the peaks and valleys of the profile to calculate the thread geometry parameters and compare them with the national standard tolerance zone, thereby realizing the automatic determination of the machining accuracy of pipe thread.

[0067] Based on the thread profile reconstructed in step S5 Geometric parameters are extracted and quality is assessed. Specifically, firstly... Identify all peak points and trough point Secondly, calculate all The average height is used as the major diameter of the thread, and all calculations are performed. The average height is used as the minor diameter; then, the mean diameter line is fitted based on the midpoint of the line connecting the crests and troughs; finally, the calculated dimensions are compared with the tolerance zone specified in national standards (such as GB / T 7306) to determine whether the pipe fitting is qualified.

[0068] An embodiment of a thread accuracy monitoring system for pipe fittings processed by an automatic threading machine:

[0069] On the other hand, the present invention also provides a thread accuracy monitoring system for pipe fittings processed by an automatic threading machine. For example... Figure 2 As shown, an automatic threading machine for monitoring the thread accuracy of pipe fittings includes a processor and a memory. The memory stores computer program instructions, which, when executed by the processor, implement the automatic threading machine for monitoring the thread accuracy of pipe fittings according to the first aspect of the present invention.

[0070] An automatic threading machine for monitoring the thread accuracy of pipe fittings also includes other components well known to those skilled in the art, such as communication interfaces. Their settings and functions are known in the art and will not be described in detail here.

[0071] In this invention, the aforementioned memory can be any tangible medium containing or storing a program that can be used or combined with an instruction execution system, apparatus, or device. For example, a computer-readable storage medium can be any suitable magnetic or magneto-optical storage medium, such as Resistive Random Access Memory (RRAM), Dynamic Random Access Memory (DRAM), Static Random Access Memory (SRAM), Enhanced Dynamic Random Access Memory (EDRAM), High-Bandwidth Memory (HBM), Hybrid Memory Cube (HMC), etc., or any other medium that can be used to store desired information and can be accessed by an application, module, or both. Any such computer storage medium can be part of a device or accessible to or connected to a device. Any application or module described in this invention can be implemented using computer-readable / executable instructions stored or otherwise maintained by such a computer-readable medium.

Claims

1. A method of monitoring the accuracy of the threading of a pipe by an automatic threading machine, characterized in that, The method comprises: acquiring a profile signal of a pipe; integrating the profile signal in a set integral interval containing a current time to obtain an impact singularity index; the object of the integration includes the amplitude of the current time, the absolute value of the second derivative, and a weighted item inversely proportional to the global standard deviation and proportional to the difference between the amplitude and the global mean; A mask signal is calculated based on the aliasing risk index, and the profile signal is injected to obtain an auxiliary signal; the auxiliary signal is subjected to empirical mode decomposition and reconstruction to obtain a pure thread profile; geometric parameters are extracted based on the pure thread profile and compared with a set standard to determine whether the thread precision is qualified; The amplitude of the mask signal is positively correlated with the standard deviation of the noise component in the profile signal and the logarithm value of the aliasing risk index; the aliasing risk index is positively correlated with the product of the impact singularity index, the theoretical thread frequency and the main vibration frequency, and negatively correlated with the absolute value of the difference between the main vibration frequency and the theoretical thread frequency.

2. A method of monitoring the accuracy of the threading of a pipe by an automatic threading machine according to claim 1, characterized in that, The calculation method of the mask signal is specifically: ； wherein represents the mask signal; represents the amplitude of the mask signal; and ; wherein represents the theoretical thread frequency, determined by the machine set speed; represents the main vibration frequency, which is the non- fundamental main vibration frequency extracted after performing a spectral analysis of the profile signal.

3. A method of monitoring the accuracy of the threading of a pipe by an automatic threading machine according to claim 2, wherein The amplitude calculation method of the mask signal is specifically: ； wherein represents the amplitude of the mask signal; represents the residual noise standard deviation, obtained by subtracting the moving average of the profile signal from its smoothed moving average, representing the base level of the background noise in the current signal; represents the natural constant; represents the aliasing risk index.

4. A method of monitoring the accuracy of the threading of a pipe by an automatic threading machine according to claim 2 or 3, characterised in that, The method for obtaining the pure thread profile is specifically: computing the mask signal and obtaining an auxiliary signal pair: ; ; The EMD is performed on the signals and respectively, to obtain two groups of intrinsic mode functions , ; Taking the mean value of the corresponding order intrinsic modal function as the reconstruction component: ; All the reconstructed components are added to obtain the pure thread profile.

5. A method of monitoring the accuracy of the threading of a pipe by an automatic threading machine according to claim 3, wherein The calculation method of the aliasing risk index is specifically: ； wherein represents the aliasing risk index, the larger the value, the higher the probability of EMD decomposition error; represents the impact singularity index; represents the theoretical thread frequency; represents the main vibration frequency; represents the phase, the value range is ; represents a very small positive number, used to prevent the denominator from being zero.

6. A method of monitoring the accuracy of the threading of a pipe by an automatic threading machine according to claim 5, wherein The calculation method of the impact singularity index is specifically: ； wherein represents the impact singularity indicator, used to characterize the signal waveform in the vicinity of the time instant ; represents the physical thread period, being a known physical parameter; represents the integral variable; represents the second derivative of the signal, i.e. the curvature; and represent the global arithmetic mean and the global standard deviation of the whole signal , respectively.

7. A method of monitoring the accuracy of the threading of a pipe by an automatic threading machine according to claim 6, wherein The calculation method of the physical thread period is specifically: ； wherein represents the physical thread cycle; represents the set spindle speed of the automatic tapping machine; represents the number of scanning lines per unit length of the laser profile sensor.

8. A method of monitoring the accuracy of the threading of a pipe by an automatic threading machine according to claim 1, characterized in that, The geometric parameters are extracted, including: All the peak points and valley points in the reconstructed pure thread profile are identified; The average value of the heights of all the peak points is calculated as the major diameter of the thread; The average value of the heights of all the valley points is calculated as the minor diameter of the thread.

9. A method of monitoring the accuracy of the threading of a pipe by an automatic threading machine according to claim 8, wherein, The major diameter and the minor diameter are compared with the tolerance band specified in the GB / T 7306 standard, and if both are within the tolerance band, the thread precision is determined to be qualified.

10. A system for monitoring the accuracy of the threading of a pipe by an automatic threading machine, characterized in that The device comprises a processor and a memory, and the memory stores computer program instructions which, when executed by the processor, implement the thread precision monitoring method of the machined pipe of the automatic threading machine according to any one of claims 1-9.

Images