A visualization method for monitoring the operating status of a maze compressor

By using adaptive bandwidth adjustment and hybrid interpolation resampling, combined with polar coordinate state distribution maps, the problems of accuracy and comprehensiveness in monitoring the operating status of maze compressors are solved, achieving highly sensitive fault identification and intuitive visualization.

CN121024908BActive Publication Date: 2026-11-17JIANGSU XIYA PETROCHEM EQUIP
View PDF 2 Cites 0 Cited by

Patent Information

Application Number
CN202511322483.1
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2025-09-16
Publication Date
2026-11-17
Estimated Expiration
2045-09-16

AI Technical Summary

Technical Problem

Existing technologies are insufficient for accurately and comprehensively monitoring the operating status of labyrinth compressors. Traditional methods suffer from decreased fault diagnosis accuracy due to speed changes and noise interference. Traditional visualization methods cannot effectively integrate multi-dimensional features, increasing the complexity of fault location.

Method used

An adaptive bandwidth adjustment strategy is adopted to determine the phase-locked loop bandwidth by using the Hearst exponent and kurtosis. Combined with hybrid interpolation kernel resampling, multidimensional features are extracted and a polar coordinate state distribution map is generated for visualization.

Benefits of technology

It achieves more accurate instantaneous phase estimation under speed fluctuations and noise interference, retains fault impact characteristics, and integrates multi-dimensional features into a single view, thereby improving the accuracy and intuitiveness of fault identification.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN121024908B_ABST
    Figure CN121024908B_ABST
Patent Text Reader

Abstract

This invention belongs to the field of maze compressor condition monitoring technology, specifically relating to a visualization method for monitoring the operating status of maze compressors. The method includes: acquiring vibration and key phase signals of the maze compressor; calculating the Hearst exponent of the instantaneous frequency and the kurtosis of the vibration signal based on the key phase signals to obtain the bandwidth of the phase-locked loop; extracting the instantaneous phase of the rotating shaft and performing equal-angle resampling; using a hybrid interpolation kernel to resample the vibration signal in the angle domain; extracting order energy and biphase coherence spectrum features through synchronous compression transformation; constructing a feature tensor indexed by rotational speed and order; visualizing the signal through a polar coordinate distribution map; identifying different state clusters using density peak clustering; distinguishing states by color and using brightness to indicate proximity to the cluster center. This invention solves the technical problem of existing technologies' difficulty in accurately and comprehensively monitoring the operating status of maze compressors.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention belongs to the field of status monitoring technology for maze compressors, and particularly relates to a visualization method for monitoring the operating status of maze compressors. Background Technology

[0002] A labyrinth compressor is a reciprocating piston compressor that utilizes non-contact labyrinth seal technology. It is primarily used for the compression and transportation of flammable, explosive, toxic gases, and gases under high-pressure conditions, and is widely applied in key industrial sectors such as petrochemicals, natural gas transportation, and energy processing. As a core fluid transportation device, the stability of the labyrinth compressor's operating status directly determines the continuity and safety of the production process. Due to the complex operating environment, labyrinth compressors are prone to malfunctions such as shaft imbalance, abnormal sealing gaps, and component rubbing during long-term operation. Failure to identify these malfunctions in a timely manner may lead to equipment shutdown, production interruption, or even safety accidents. Therefore, real-time and accurate monitoring of the labyrinth compressor's operating status and fault early warning has become a core requirement in the field of industrial equipment operation and maintenance.

[0003] Vibration signal analysis is a primary means of monitoring the operating status of labyrinth compressors and diagnosing potential faults. Its core principle is that changes in the operating status of the labyrinth compressor are directly reflected in the frequency, amplitude, and phase characteristics of the vibration signal. Analysis of the vibration signal allows for the inverse deduction of the equipment's health status. However, the actual operation of a labyrinth compressor is not a steady state. During start-up and shutdown, the speed linearly increases from zero to the rated value; under variable load conditions, the speed needs to be dynamically adjusted according to production requirements, causing the frequency components of the vibration signal to continuously change over time. This characteristic leads to problems such as spectral ambiguity and energy leakage in traditional spectrum analysis methods based on Fast Fourier Transform (FFT), making it impossible to accurately extract fault features related to speed, resulting in a significant decrease in fault diagnosis accuracy. To address these issues, the industry has gradually adopted order tracking analysis technology. Order tracking analysis technology transforms the vibration signal from the time domain to the angular domain, eliminating the influence of speed changes on spectrum analysis. This makes the frequency components related to shaft rotation, i.e., the order, stable in the angular domain, facilitating fault diagnosis.

[0004] However, the accuracy of order tracking analysis largely depends on two key aspects: instantaneous phase estimation and equal-angle resampling. Currently, instantaneous phase acquisition primarily relies on a key phase signal synchronized with the shaft, processed using phase-locked loop (PLL) technology. However, the loop filter bandwidth of traditional PLLs is fixed, making it unsuitable for the complex operating conditions of compressors. While wideband filters can quickly track sudden speed changes, their noise immunity is weak, easily misinterpreting environmental interference signals as speed changes, leading to phase estimation errors. Narrowband filters, while suppressing noise, have slow response times and cannot keep up with rapid speed fluctuations, also causing phase calculation errors and ultimately affecting the accuracy of order tracking. In the equal-angle resampling stage, commonly used methods such as linear interpolation and cubic spline interpolation, although simple to implement, may smooth out key transient features when the original vibration signal contains impact components caused by faults, introducing amplitude and phase errors and affecting the accuracy of subsequent feature extraction. Traditional visualization methods (such as time-domain waveforms and frequency-domain spectrograms) can only display features in a single dimension and cannot effectively integrate multi-dimensional features. This makes it difficult for maintenance personnel to intuitively and comprehensively judge the overall status of the equipment, increasing the complexity and time cost of fault location. Summary of the Invention

[0005] Therefore, the purpose of this invention is to provide a visualization method for monitoring the operating status of a maze compressor, so as to solve the technical problem that the existing technology is unable to accurately and comprehensively monitor the operating status of a maze compressor.

[0006] This invention proposes a visualization method for monitoring the operating status of a labyrinth compressor, comprising the following steps: The vibration signal and the key phase signal synchronized with the shaft during the operation of the labyrinth compressor are acquired. Based on the key phase signal, the Hearst exponent of the instantaneous frequency within the time window and the kurtosis of the vibration signal are calculated. The loop filter bandwidth of the phase-locked loop is obtained by combining the Hearst exponent and the kurtosis. The loop filter bandwidth is negatively correlated with the Hearst exponent and has a preset piecewise function relationship with the kurtosis. The key phase signal is processed using the phase-locked loop configured with the loop filter bandwidth to obtain the instantaneous phase information of the shaft. The equiangular sampling time is determined based on the instantaneous phase information; for the vibration signal, the peak factor in the neighborhood of each sampling time is calculated, and the mixed weight of the Gaussian kernel function and the Lagrange basis function is obtained based on the peak factor, wherein the peak factor is negatively correlated with the weight of the Gaussian kernel function and positively correlated with the weight of the Lagrange basis function; the vibration signal is resampled using a hybrid interpolation kernel composed of the linear superposition of the weighted kernel functions to obtain the angular domain signal; The angle domain signal is synchronously compressed and transformed to extract the energy of the preset fundamental frequency order and harmonic order, as well as the biphase spectrum values ​​between each order. The features extracted at different speeds are constructed into feature tensors indexed by speed and order. A polar coordinate state distribution map is generated for visualization, with the polar radial axis and polar angle axis representing the rotational speed and order, respectively; density peak clustering is performed on the feature vectors in the feature tensor to divide the state clusters; the color of the coordinate point in the polar coordinate state distribution map is determined by the index of the state cluster to which the corresponding feature vector belongs, and the brightness is determined by the reciprocal of the distance from the feature vector to the center of the state cluster to which the vector belongs.

[0007] Further, the step of calculating the Hearst exponent of the instantaneous frequency within the time window and the kurtosis of the vibration signal based on the key phase signal, and obtaining the loop filter bandwidth of the phase-locked loop based on the Hearst exponent and the kurtosis, includes: Obtain a first kurtosis threshold and a second kurtosis threshold, wherein the second kurtosis threshold is greater than the first kurtosis threshold; Calculate the Hearst exponent H of the instantaneous frequency within the time window and the kurtosis K of the vibration signal; Based on the range of kurtosis K, the corresponding bandwidth coefficient C is selected: when K is less than or equal to the first kurtosis threshold, the first bandwidth coefficient C1 is selected; when K is greater than the first kurtosis threshold and less than or equal to the second kurtosis threshold, the second bandwidth coefficient C2 is selected; when K is greater than the second kurtosis threshold, the third bandwidth coefficient C3 is selected; where C3... <C2<C1; The bandwidth B of the loop filter is calculated according to the formula. Calculations show that This is the baseline bandwidth.

[0008] Furthermore, the hybrid interpolation kernel determines whether there is an impact in the vibration signal by calculating the peak factor; when the vibration signal is stable and the peak factor is low, the weight of the Gaussian kernel function is increased to smooth the noise; when the vibration signal contains an impact and the peak factor is high, the weight of the Lagrange basis function is increased to preserve the transient waveform characteristics.

[0009] Further, for the vibration signal, calculating the peak factor in the neighborhood at each sampling time, and obtaining the mixed weight of the Gaussian kernel function and the Lagrange basis function based on the peak factor, includes: For each equiangular sampling moment, a preset number of time sampling points before and after that moment are taken to form the neighborhood of that moment; Calculate the peak factor CF of the vibration signal within the neighborhood; Using the preset slope parameter k and center threshold Through function The weights of the Lagrange basis function are calculated. ; The weights of the Gaussian kernel function According to the formula Calculated.

[0010] Further, the synchronous compression transformation of the angle domain signal to extract the energy of the preset fundamental frequency order and harmonic orders, as well as the biphase coherence spectrum values ​​between each order, includes: From the synchronous compression transformation result of the angle domain signal, extract the signal components of the first, second, and third harmonics of the fundamental frequency; Integrating the squared amplitudes of the signal components at the first, second, and third harmonics orders respectively yields the energy of each order. , , ; Calculate the biphase spectrum values ​​between the first and second harmonic order signal components. ; Energy , , and bicoherence spectral values These are combined into a feature vector to represent the operating state at the current speed.

[0011] Further, performing density peak clustering on the feature vectors in the feature tensor to divide the state clusters includes: Set a cutoff distance As the analysis radius, the Euclidean distance between any two eigenvectors in the feature tensor is calculated. Calculate the local density of each eigenvector Local density This means the Euclidean distance to the current feature vector is less than the cutoff distance. The number of other feature vectors; Calculate the minimum distance δ for each feature vector. The minimum distance δ represents the minimum Euclidean distance to the current feature vector among other feature vectors with a local density greater than the current feature vector. Selecting those with large local density The feature vector with the largest minimum distance δ is used as the center point of the state cluster. The remaining feature vectors are assigned to the state clusters to which the feature vectors closest to the center point and whose local density is greater than that of the center point belong, thus completing the state cluster partitioning.

[0012] Furthermore, the state clusters include normal state clusters, unbalanced state clusters, and rubbing state clusters.

[0013] Furthermore, the vibration signal and the key phase signal synchronized with the shaft during the operation of the labyrinth compressor are acquired by means of an accelerometer and an eddy current sensor aligned with the keyway or protrusion on the shaft of the labyrinth compressor.

[0014] Furthermore, the Hearst exponent is calculated using the rescaled range analysis method.

[0015] Furthermore, the method for determining the equal-angle sampling time is as follows: based on the instantaneous phase information, the time corresponding to the phase reaching the equal-interval angle is calculated by linear interpolation inverse solution, and this time is the equal-angle sampling time.

[0016] The beneficial effects of this invention are: This invention employs an adaptive bandwidth adjustment strategy. The Hearst exponent is used to determine the trend of rotational speed changes, and the kurtosis is used to determine the intensity of impulse interference in the signal. These two factors jointly determine the loop filter bandwidth of the phase-locked loop (PLL). This dynamic adjustment allows the PLL to narrow its bandwidth to enhance interference immunity when the rotational speed is stable and noise is high, and to widen its bandwidth to achieve rapid tracking when the rotational speed changes drastically, thereby obtaining more accurate instantaneous phase information of the rotating shaft. In the equal-angle resampling stage, an adaptive hybrid interpolation kernel is used. The presence of an impulse is determined by calculating the peak factor in the neighborhood of each sampling point. When the peak factor is high, indicating an impulse, the weight of the Lagrange basis function is increased to preserve the transient waveform; when the peak factor is low, indicating a stable signal, the weight of the Gaussian kernel function is increased to smooth noise. This ensures that the impulse characteristics caused by faults are not incorrectly filtered out or passivated during the critical process of converting the signal from the time domain to the angle domain. Furthermore, this invention, based on the extraction of energy at each order, introduces biphasic coherence spectrum values ​​as features, enabling the capture of deep-seated fault information that cannot be discovered by energy analysis alone. This achieves multi-dimensional, highly sensitive feature extraction and characterization of the labyrinth compressor's operating status, improving the accuracy of fault identification. Finally, this invention, through density peak clustering algorithm and polar coordinate state distribution map, can automatically identify and classify equipment operating states. It also integrates multi-dimensional feature information into a single view through a novel visualization method, enhancing the comprehensiveness and intuitiveness of labyrinth compressor status monitoring. Attached Figure Description

[0017] Figure 1 This is a flowchart illustrating the steps of a visualization method for monitoring the operating status of a labyrinth compressor according to the present invention. Figure 2 This is a schematic diagram of signal acquisition. Figure 3 Schematic diagram for adaptive setting of phase-locked loop bandwidth; Figure 4 A schematic diagram illustrating the setting of weights for the hybrid interpolation kernel; Figure 5A schematic diagram of the construction of the characteristic tensor; Figure 6 This is a polar coordinate state distribution diagram. Detailed Implementation

[0018] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention.

[0019] A specific embodiment of the visualization method for monitoring the operating status of a labyrinth compressor according to the present invention: like Figure 1 As shown, a visualization method for monitoring the operating status of a labyrinth compressor according to the present invention includes the following steps: S1: Acquire the vibration signal and the key phase signal synchronized with the shaft during the operation of the labyrinth compressor; based on the key phase signal, calculate the Hearst exponent of the instantaneous frequency within the time window and the kurtosis of the vibration signal; obtain the loop filter bandwidth of the phase-locked loop (PLL) based on the Hearst exponent and the kurtosis, wherein the loop filter bandwidth is negatively correlated with the Hearst exponent and has a preset piecewise function relationship with the kurtosis; process the key phase signal using the PLL configured with the loop filter bandwidth to obtain the instantaneous phase information of the shaft; Specifically, in this step, an accelerometer installed at a designated location on the labyrinth compressor collects radial or axial vibration signals. Simultaneously, a keyway or protrusion is created on the rotating shaft. By aligning an eddy current sensor with a mark on the shaft, a pulse is generated each rotation of the shaft; this pulse signal is the key phase signal. The two signals are acquired synchronously to obtain a discrete time-series signal, such as... Figure 2 As shown.

[0020] The instantaneous frequency sequence of the shaft is obtained by calculating the time interval between adjacent pulses of the key phase signal. The Hearst exponent H of the instantaneous frequency sequence within a time window is calculated using the rescaled range analysis method. The Hearst exponent reflects the trend of rotational speed change. Simultaneously, the kurtosis value of the original vibration signal within the same time window is calculated. The kurtosis value reflects the intensity of the impact component in the signal. The bandwidth of the phase-locked loop (PLL) can be determined based on the Hearst exponent, kurtosis, and reference bandwidth. The calculated bandwidth is configured in the digital PLL, which includes a phase detector, a loop filter, and a voltage-controlled oscillator (VCO). The phase detector compares the phase difference between the input key phase pulse signal and the internal oscillation signal generated by the VCO, generating an error signal. This error signal is sent to the loop filter for processing. The bandwidth of this loop filter is configured according to the Hearst exponent and kurtosis, and it can determine whether to smooth noise or respond quickly to changes based on the rotational speed stability and vibration impact. The filtered error signal adjusts the frequency and phase of the VCO, and the VCO output is the high-resolution instantaneous phase information of the shaft. Figure 3 As shown.

[0021] S2: Determine the equal-angle sampling time based on the instantaneous phase information; for the vibration signal, calculate the peak factor in the neighborhood of each sampling time, and obtain the mixed weight of the Gaussian kernel function and the Lagrange basis function based on the peak factor, wherein the peak factor is negatively correlated with the weight of the Gaussian kernel function and positively correlated with the weight of the Lagrange basis function; resample the vibration signal using a mixed interpolation kernel composed of the linear superposition of the weighted kernel functions to obtain the angle domain signal; Based on the instantaneous phase information obtained in step S1, the phase reaches the equally spaced angles through linear interpolation, such as the time corresponding to every 1024 points of rotation. These times are the equal-angle sampling times. For each equal-angle sampling time, a neighborhood window containing dozens of original sampling points is taken before and after it. The peak factor of the vibration signal within the neighborhood window is calculated, which is the ratio of the maximum absolute value of the signal to the effective value of the signal. The weights of the Gaussian kernel function and the Lagrange basis function are obtained, and their sum is 1. Preferably, the weight of the Lagrange basis function is determined by the peak factor through a sigmoid function. When the vibration signal is stable and the peak factor is small, the weight of the Lagrange basis function is close to 0, and the weight of the Gaussian kernel function is close to 1; when the vibration signal contains impact and the peak factor is large, the weight of the Lagrange basis function is close to 1, and the weight of the Gaussian kernel function is close to 0. The amplitude of the original time-domain vibration signal at each equal-angle sampling moment is interpolated using a hybrid interpolation kernel composed of a Gaussian kernel multiplied by a weighted Gaussian kernel multiplied by a weighted Lagrange base function multiplied by a Lagrange base function. Specifically, the vibration amplitude at the target moment is calculated using the hybrid interpolation kernel, and the calculated amplitudes at all equal-angle time points are arranged sequentially to obtain the angle-domain signal, such as... Figure 4 As shown. For example, for the equiangular sampling points that need to be calculated, such as 10°, 20°, etc., find their positions on the time axis, call the hybrid interpolation kernel, take the original time sampling values ​​around the target time point as input, and use the hybrid interpolation kernel to calculate the vibration amplitude at the target time point; arrange the amplitudes calculated for all angle points in order of angle to obtain the vibration signal sequence with angle as the horizontal axis, that is, the angle domain signal.

[0022] S3: Perform synchronous compression transformation on the angle domain signal to extract the energy of the preset fundamental frequency order and harmonic order, as well as the biphase spectrum values ​​between each order, and construct the extracted features at different speeds into feature tensors indexed by speed and order.

[0023] In this step, the resampled angle-domain signal undergoes synchronous compression transformation to obtain an angle-order spectrum. From this angle-order spectrum, energy amplitude is calculated by integrating along specific order lines, such as the 1st, 2nd, and 3rd harmonics, to obtain the energy characteristics of each order. Simultaneously, the angle-domain signal is segmented and subjected to Fourier transform to calculate the biphase coherence spectrum values ​​between preset order pairs. These biphase coherence spectrum values ​​reflect the nonlinear correlation between orders. All order energies and biphase coherence spectrum values ​​extracted at a specific speed are combined into a feature vector. By repeating this process at multiple different speed points during the acceleration and deceleration of the maze compressor, a three-dimensional feature tensor is formed. The three dimensions represent speed, order, and the features composed of energy and biphase coherence spectrum values, respectively. Figure 5 As shown.

[0024] S4: Generate a polar coordinate state distribution map for visualization, with the polar radial axis and polar angle axis representing the rotational speed and order, respectively; perform density peak clustering on the feature vectors in the feature tensor to divide the state clusters; the color of the coordinate point in the polar coordinate state distribution map is determined by the index of the state cluster to which the corresponding feature vector belongs, and the brightness is determined by the reciprocal of the distance from the feature vector to the center of the state cluster to which the vector belongs.

[0025] In this step, all eigenvectors in the feature tensor are used as input. A density peak clustering algorithm is applied to identify cluster centers by calculating the local density of each eigenvector and its distance to points with higher local densities, thus classifying all eigenvectors into different state clusters. Examples include normal state clusters, unbalanced state clusters, and rubbing state clusters. A polar coordinate state distribution map is created. Preferably, the polar radius from the center to the outer edge represents the rotational speed from low to high, and the polar angle is evenly distributed 360 degrees to the analyzed orders. Each coordinate point in the polar coordinate state distribution map corresponds to a specific rotational speed and order, with its polar radius and polar angle corresponding to the specific rotational speed and order, respectively. The color of the coordinate point is obtained based on the index of the state cluster to which the corresponding eigenvector belongs; for example, green for normal, yellow for unbalanced, and red for rubbing. The Euclidean distance from the eigenvector to the center of its respective state cluster is calculated, and the reciprocal of the Euclidean distance is mapped to the brightness of the point. The closer the distance, the more representative the state of the maze compressor, and the higher the brightness; the farther the distance, the darker the brightness, indicating a possible state boundary or abnormal operating condition, such as... Figure 6 As shown.

[0026] To address variations in the rotational speed stability of the equipment under different conditions, in an optional embodiment, the calculation of the Hearst exponent of the instantaneous frequency within a time window and the kurtosis of the vibration signal based on the key phase signal, and the acquisition of the loop filter bandwidth of the phase-locked loop based on the Hearst exponent and the kurtosis, includes: Obtain a first kurtosis threshold and a second kurtosis threshold, wherein the second kurtosis threshold is greater than the first kurtosis threshold; Calculate the Hearst exponent H of the instantaneous frequency within the time window and the kurtosis K of the vibration signal; Based on the range of kurtosis K, the corresponding bandwidth coefficient C is selected: when K is less than or equal to the first kurtosis threshold, the first bandwidth coefficient C1 is selected; when K is greater than the first kurtosis threshold and less than or equal to the second kurtosis threshold, the second bandwidth coefficient C2 is selected; when K is greater than the second kurtosis threshold, the third bandwidth coefficient C3 is selected; where C3... <C2<C1; The bandwidth B of the loop filter is calculated according to the formula. Calculations show that This is the baseline bandwidth.

[0027] For example, the first kurtosis threshold is set to 3.5, and the second kurtosis threshold is set to 6. When the equipment operates smoothly and there is no obvious impact in the vibration signal, the calculated kurtosis K may be 2.8, which is less than 3.5. This indicates that the speed fluctuation is small, and a larger first bandwidth coefficient C1, such as 0.8, can be selected to enhance the noise immunity of the phase-locked loop and avoid misinterpreting random noise as speed changes.

[0028] If the equipment experiences an early fault, resulting in a minor impact, the kurtosis K is 5.0, falling between 3.5 and 6.0. In this case, a moderate second bandwidth coefficient C2, such as 0.5, is selected to ensure the PLL can respond promptly to the actual speed fluctuations caused by the fault impact. When the equipment fault is severe, generating a violent impact, the kurtosis K > 6, for example, K = 8.1. In this case, a smaller third bandwidth coefficient C3, such as 0.2, is selected to ensure the PLL can quickly track speed changes. Simultaneously, the Hurst exponent H reflects the stability of the instantaneous frequency. If H is 0.9, it indicates a very smooth frequency sequence, and the bandwidth B will decrease accordingly; if H is 0.6, it indicates violent fluctuations in the frequency sequence, and the bandwidth B will increase, achieving fine-grained dual adjustment of the bandwidth.

[0029] In an optional embodiment, the step of calculating the peak factor in the neighborhood of each sampling time for the vibration signal, and obtaining the mixed weight of the Gaussian kernel function and the Lagrange basis function based on the peak factor, includes: For each equiangular sampling moment, a preset number of time sampling points before and after that moment are taken to form the neighborhood of that moment; Calculate the peak factor CF of the vibration signal within the neighborhood; Using the preset slope parameter k and center threshold Through function The weights of the Lagrange basis function are calculated. ; The weights of the Gaussian kernel function According to the formula Calculated.

[0030] For example, suppose interpolation is performed on a sampling point at a constant angle, defining the neighborhood of the sampling point as 16 time sampling points before and after it. If the vibration signal waveform within the neighborhood of the sampling point is smooth and there is no fault impact, the calculated peak factor CF may be 2.5. A preset center threshold for the Logistic function is used. The slope parameter k is 5, and the peak factor of 2.5 is less than the center threshold of 4.0. Since the peak factor of 2.5 is less than the center threshold of 4.0, the calculated result of the Logistic function is... It is very close to 0. Therefore, the weights of the Gaussian kernel function are... The value will be very close to 1, in which case the Gaussian kernel function will be mainly used for interpolation, and its smoothing properties can filter out high-frequency noise in the signal. Conversely, if the neighborhood contains a strong impact caused by bearing pitting, the peak factor CF is 7.0, which is greater than the center threshold of 4.0. In this case, the calculated value will be... It will be very close to 1, and If the value is close to 0, then the Lagrange basis function will be mainly used for interpolation, which can reconstruct the transient waveform of the impact signal and avoid the fault feature passivation that may be caused by Gaussian smoothing.

[0031] In an optional embodiment, the step of synchronously compressing and transforming the angle domain signal to extract the energy of the preset fundamental frequency order and harmonic orders, as well as the biphase spectral values ​​between each order, includes: From the synchronous compression transformation result of the angle domain signal, extract the signal components of the first, second, and third harmonics of the fundamental frequency; Integrating the squared amplitudes of the signal components at the first, second, and third harmonics orders respectively yields the energy of each order. , , ; Calculate the biphase spectrum values ​​between the first and second harmonic order signal components. ; Energy , , and bicoherence spectral values These are combined into a feature vector to represent the operating state at the current speed.

[0032] Assuming the fundamental frequency of the rotating shaft is a first harmonic, and the gear meshing frequency exhibits higher harmonics, the analysis involves performing a synchronous compression transform on a segment of the angle domain signal to obtain its order spectrum. The first harmonic signal is then extracted from this spectrum, and its energy is calculated to obtain... The value is 12.5 units. Similarly, the second and third harmonic orders of the signal are extracted, and the energy is calculated. It is 4.8 units. The value is 2.1 units. The three energy values ​​reflect the intensity of the fundamental frequency and harmonic components. The biphase spectral values ​​between the first and second harmonic orders are calculated, yielding... A value of 0.78 indicates the strength of the nonlinear correlation between these two order components. Faults such as tooth surface wear or cracks often lead to an enhanced correlation. These four values ​​are combined into a feature vector, namely [12.5, 4.8, 2.1, 0.78], which serves as a fingerprint of the gearbox's operating state at the current speed.

[0033] In an optional embodiment, performing density peak clustering on the feature vectors in the feature tensor to partition state clusters includes: Set a cutoff distance As the analysis radius, the Euclidean distance between any two eigenvectors in the feature tensor is calculated. Calculate the local density of each eigenvector Local density This means the Euclidean distance to the current feature vector is less than the cutoff distance. The number of other feature vectors; Calculate the minimum distance δ for each feature vector. The minimum distance δ represents the minimum Euclidean distance to the current feature vector among other feature vectors with a local density greater than the current feature vector. Selecting those with large local density The feature vector with the largest minimum distance δ is used as the center point of the state cluster. The remaining feature vectors are assigned to the state clusters to which the feature vectors closest to the center point and whose local density is greater than that of the center point belong, thus completing the state cluster partitioning.

[0034] For example, assuming 1000 feature vectors are collected, and a cutoff distance is set... The value is 0.8. For one feature vector point A, calculations show that 60 other vector points are less than 0.8 away from it. Therefore, the local density of point A is... The value is 60. Find all vector points with a local density greater than 60; these are high local density points. Calculate the distance from point A to these high local density points. The smallest distance is 2.5; this value is the minimum distance δ for point A. Perform this calculation on all 1000 points. You will find that some points, such as point X, A high value, such as 150, and a large δ value, such as 3.0, indicate that point X is located in a dense region and is far from the center of other, even denser regions. Therefore, it is likely the center of an independent state cluster, representing a typical operating state, such as the normal operation of equipment. Another... Points Y with both large values ​​of δ and δ may represent another state, such as an early fault state. After all the center points are determined, the remaining points are assigned to these state clusters according to the rules, thus completing the automatic division of the equipment's operating state.

[0035] Compared to existing technologies, this invention dynamically adjusts the phase-locked loop bandwidth by using the Hearst exponent of the instantaneous frequency of the bond phase signal and the kurtosis of the vibration signal, thus solving the problem of phase estimation deviation due to response lag or noise interference during speed fluctuations and improving the accuracy of instantaneous phase estimation. Addressing the drawback of traditional fixed interpolation kernel resampling easily losing fault impact features, this invention dynamically allocates the mixed weights of the Gaussian kernel and Lagrange basis function based on the peak factor of the neighborhood signal at the sampling time, effectively preventing the passivation of fault impact components and ensuring that early minor fault features are not lost. Finally, addressing the limitation of existing technologies in fusing multidimensional features, this invention extracts order energy and inter-order biphase coherence spectra through synchronous compression transform. The invention constructs a feature tensor indexed by rotational speed and order, fully preserving state information under different operating conditions and reducing false positives and false negatives caused by single-dimensional features. Addressing the lack of intuitiveness in traditional visualization methods, this invention combines polar coordinate state distribution maps with density peak clustering, using color to distinguish state types and brightness to reflect the distance to cluster centers. This integrates multi-dimensional features into a single intuitive view, significantly reducing the difficulty of judgment for maintenance personnel and shortening fault location time. The overall technical solution is more suitable for the complex operating requirements of labyrinth compressors in industrial scenarios, achieving significant breakthroughs in the accuracy, feature completeness, comprehensive state representation, and ease of maintenance of labyrinth compressor operating status monitoring.

[0036] While various embodiments of the invention have been shown and described in this specification, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Many modifications, alterations, and alternatives will occur to those skilled in the art without departing from the spirit and essence of the invention.

Claims

1. A method for visualization of labyrinth compressor operating condition monitoring, characterized in that, Includes the following steps: The vibration signal and the key phase signal synchronized with the shaft during the operation of the labyrinth compressor are acquired. Based on the key phase signal, the Hearst exponent of the instantaneous frequency within the time window and the kurtosis of the vibration signal are calculated. The loop filter bandwidth of the phase-locked loop is obtained by combining the Hearst exponent and the kurtosis. The loop filter bandwidth is negatively correlated with the Hearst exponent and has a step-decreasing function relationship with the kurtosis. The key phase signal is processed using the phase-locked loop configured with the loop filter bandwidth to obtain the instantaneous phase information of the shaft. The equiangular sampling time is determined based on the instantaneous phase information; for the vibration signal, the peak factor in the neighborhood of each sampling time is calculated, and the mixed weight of the Gaussian kernel function and the Lagrange basis function is obtained based on the peak factor, wherein the peak factor is negatively correlated with the weight of the Gaussian kernel function and positively correlated with the weight of the Lagrange basis function; the vibration signal is resampled using a hybrid interpolation kernel composed of the linear superposition of the weighted kernel functions to obtain the angular domain signal; The angle domain signal is synchronously compressed and transformed to extract the energy of the preset fundamental frequency order and harmonic order, as well as the biphase spectrum values ​​between each order. The features extracted at different speeds are constructed into feature tensors indexed by speed and order. A polar coordinate state distribution map is generated for visualization, with the polar radial axis and polar angle axis representing the rotational speed and order, respectively; density peak clustering is performed on the feature vectors in the feature tensor to divide the state clusters; the color of the coordinate point in the polar coordinate state distribution map is determined by the index of the state cluster to which the corresponding feature vector belongs, and the brightness is determined by the reciprocal of the distance from the feature vector to the center of the state cluster to which the vector belongs.

2. The visualization method for monitoring the operating status of a labyrinth compressor according to claim 1, characterized in that, The step of calculating the Hearst exponent of the instantaneous frequency within the time window and the kurtosis of the vibration signal based on the key phase signal, and obtaining the loop filter bandwidth of the phase-locked loop based on the Hearst exponent and the kurtosis, includes: Obtain a first kurtosis threshold and a second kurtosis threshold, wherein the second kurtosis threshold is greater than the first kurtosis threshold; Calculate the Hearst exponent H of the instantaneous frequency within the time window and the kurtosis K of the vibration signal; Based on the range of kurtosis K, the corresponding bandwidth coefficient C is selected: when K is less than or equal to the first kurtosis threshold, the first bandwidth coefficient C1 is selected; when K is greater than the first kurtosis threshold and less than or equal to the second kurtosis threshold, the second bandwidth coefficient C2 is selected; when K is greater than the second kurtosis threshold, the third bandwidth coefficient C3 is selected; where C3... <C2<C1; The bandwidth B of the loop filter is calculated according to the formula. Calculations show that This is the baseline bandwidth.

3. The visualization method for monitoring the operating status of a labyrinth compressor according to claim 1, characterized in that, The hybrid interpolation kernel determines whether there is an impact in the vibration signal by calculating the peak factor; when the vibration signal is stable and the peak factor is low, the weight of the Gaussian kernel function is increased to smooth the noise; when the vibration signal contains an impact and the peak factor is high, the weight of the Lagrange basis function is increased to preserve the transient waveform characteristics.

4. The visualization method for monitoring the operating status of a labyrinth compressor according to claim 3, characterized in that, For the vibration signal, the peak factor in the neighborhood at each sampling time is calculated, and the mixed weight of the Gaussian kernel function and the Lagrange basis function is obtained based on the peak factor, including: For each equiangular sampling moment, a preset number of time sampling points before and after that moment are taken to form the neighborhood of that moment; Calculate the peak factor CF of the vibration signal within the neighborhood; Using the preset slope parameter k and center threshold Through function The weights of the Lagrange basis function are calculated. ; The weights of the Gaussian kernel function According to the formula Calculated.

5. The visualization method for monitoring the operating status of a labyrinth compressor according to claim 1, characterized in that, The synchronous compression transformation of the angle domain signal to extract the energy of the preset fundamental frequency order and harmonic orders, as well as the biphase coherence spectrum values ​​between each order, includes: From the synchronous compression transformation result of the angle domain signal, extract the signal components of the first, second, and third harmonics of the fundamental frequency; Integrating the squared amplitudes of the signal components at the first, second, and third harmonics orders respectively yields the energy of each order. , , ; Calculate the biphase spectrum values ​​between the first and second harmonic order signal components. ; Energy , , and bicoherence spectral values These are combined into a feature vector to represent the operating state at the current speed.

6. The visualization method for monitoring the operating status of a labyrinth compressor according to claim 5, characterized in that, Performing density peak clustering on the feature vectors in the feature tensor to divide state clusters includes: Set a cutoff distance As the analysis radius, the Euclidean distance between any two eigenvectors in the feature tensor is calculated. Calculate the local density of each eigenvector Local density This means the Euclidean distance to the current feature vector is less than the cutoff distance. The number of other feature vectors; Calculate the minimum distance δ for each feature vector. The minimum distance δ represents the minimum Euclidean distance to the current feature vector among other feature vectors with a local density greater than the current feature vector. Selecting those with large local density The feature vector with the largest minimum distance δ is used as the center point of the state cluster. The remaining feature vectors are assigned to the state clusters to which the feature vectors closest to the center point and whose local density is greater than that of the center point belong, thus completing the state cluster partitioning.

7. The visualization method for monitoring the operating status of a labyrinth compressor according to claim 6, characterized in that, The state clusters include normal state clusters, unbalanced state clusters, and rubbing state clusters.

8. The visualization method for monitoring the operating status of a labyrinth compressor according to claim 1, characterized in that, The vibration signal and the key phase signal synchronized with the shaft of the labyrinth compressor are obtained during operation. The vibration signal is acquired by an accelerometer, and the key phase signal is acquired by an eddy current sensor aligned with the keyway or protrusion on the shaft of the labyrinth compressor.

9. The visualization method for monitoring the operating status of a labyrinth compressor according to claim 1, characterized in that, The Hearst exponent was calculated using the rescaled range analysis method.

10. The visualization method for monitoring the operating status of a labyrinth compressor according to claim 1, characterized in that, The method for determining the equal-angle sampling time is as follows: based on the instantaneous phase information, the time corresponding to the phase reaching the equal-interval angle is calculated by linear interpolation inverse solution. This time is the equal-angle sampling time.

Citation Information

Patent Citations

  • Rolling bearing fault diagnosis method based on calculation order tracking and spectral kurtosis

    CN111307460A

  • Keyless phase order tracking method for fault diagnosis of rolling bearing under strong variable rotating speed

    CN116659866A