Gearbox vibration analysis method

Abstract

The invention discloses a gearbox vibration analysis method, which belongs to the technical field of mechanical fault diagnosis, and comprises the following steps of: respectively constructing a gear local damage model, a bearing fault evaluation model, a shaft system centering model and a lubricating oil evaluation model by acquiring time domain, frequency domain and time-frequency domain state parameters, and calculating evaluation coefficients of each part; and outputting a gear box motion synergy degree based on the synergy model under the lubricating oil influence condition, and comparing the synergy degree through a preset threshold value to realize state grading judgment. According to the method, multi-domain parameters are fused, fault features are quantized, the defect of single-dimension analysis of a traditional method is overcome, the comprehensiveness and accuracy of fault diagnosis are remarkably improved, and a reliable basis is provided for intelligent maintenance of the gearbox.

Description

Technical Field

[0001] The invention belongs to the technical field of mechanical fault diagnosis, and in particular relates to a gear box vibration analysis method. Background Art

[0002] As the core component of mechanical transmission, the vibration characteristics of the gearbox are an important basis for fault diagnosis. Existing technologies mostly perform local analysis based on parameters of a single dimension in the time domain or frequency domain (such as vibration amplitude, spectrum peak), which makes it difficult to comprehensively reflect the synergistic effects of gears, bearings, shaft alignment and lubrication status. For example, traditional methods for diagnosing local gear damage often rely on time domain kurtosis, but do not combine frequency domain sideband characteristics and time-frequency domain ridge smoothness, resulting in a high misjudgment rate; bearing fault assessment often uses envelope spectrum analysis, but does not correlate time domain waveform factor and HHT transient jump number, resulting in insufficient sensitivity. In addition, there is a lack of modeling of the coupling relationship between lubricating oil status and mechanical component failure, making it impossible to achieve a quantitative assessment of the overall health status. Therefore, there is an urgent need for a vibration diagnosis method that integrates multi-domain fusion and collaborative analysis to improve the accuracy of gearbox fault warning and maintenance efficiency. Summary of the Invention

[0003] In view of the deficiencies in the prior art, the present invention provides a gearbox vibration analysis method to solve the above problems.

[0004] To achieve the above objectives, the present invention is implemented through the following technical solutions: A gearbox vibration analysis method comprises the following steps:

[0005] Obtaining time domain state parameters, frequency domain state parameters, and time-frequency domain state parameters;

[0006] According to the time domain kurtosis, the frequency domain meshing frequency sideband interval and number, and the time and frequency domain ridge smoothness, a gear local damage model is constructed to output the gear local damage evaluation coefficient;

[0007] A bearing fault assessment model is constructed based on the time domain waveform factor, the outer race fault frequency amplitude in the frequency domain envelope spectrum, and the number of HHT transient jumps in the time and frequency domains to output the bearing fault evaluation coefficient.

[0008] The shaft alignment model is constructed based on the waveform symmetry in the time domain, the meshing frequency harmonic amplitude in the frequency domain, and the energy density of the HHT marginal spectrum at 2× the rotation frequency in the time-frequency domain to output the shaft alignment evaluation coefficient.

[0009] Perform weighted average processing based on the current lubricating oil temperature and cleanliness to obtain the lubricating oil evaluation coefficient;

[0010] According to the shafting alignment evaluation coefficient, bearing fault evaluation coefficient and gear local damage evaluation coefficient under the condition of lubricating oil influence coefficient, a gearbox motion coordination model is constructed to output the gearbox motion coordination degree.

[0011] On the basis of the above technical solutions, the present invention also provides the following optional technical solutions:

[0012] Further technical solution: the time domain state parameters include kurtosis, waveform symmetry and waveform factor, the frequency domain state parameters include meshing frequency sideband interval, the number of meshing frequency sidebands, the outer ring fault frequency amplitude in the envelope spectrum and the meshing frequency harmonic amplitude, the time-frequency domain state parameters include ridge smoothness, the number of HHT instantaneous jumps, and the energy density of the HHT marginal spectrum at 2× the rotation frequency.

[0013] Further technical solution: The specific steps of constructing a gear local damage model and outputting a gear local damage evaluation coefficient based on the time domain kurtosis, the frequency domain meshing frequency sideband interval and number, and the time-frequency domain ridge smoothness are as follows:

[0014] Perform maximum-minimum normalization processing on the obtained time domain kurtosis, frequency domain meshing frequency sideband interval and number, and time-frequency domain ridge smoothness to obtain the kurtosis index, sideband interval index, sideband number index, and ridge smoothness index;

[0015] According to the kurtosis index, sideband spacing index, sideband number index and ridge smoothness index, a gear local loss damage model is constructed and the gear local damage coefficient is imported and output;

[0016] Compare the obtained gear local damage coefficient with the preset damage coefficient threshold. If the gear local damage coefficient is not within the damage coefficient threshold, it becomes maintenance information;

[0017] The gear local damage model is expressed as:

[0018]

[0019] Among them, C gea Indicates the local damage coefficient of the gear, K norm represents the time domain kurtosis index, N norm represents the sideband number index, Δf norm represents the sideband spacing index, S norm Represents the ridge smoothness index.

[0020] Further technical solution: The specific steps of constructing a bearing fault assessment model and outputting a bearing fault evaluation coefficient based on the time domain waveform factor, the outer race fault frequency in the frequency domain envelope spectrum, and the number of HHT transient jumps in the time and frequency domain are as follows:

[0021] Perform maximum-minimum normalization processing on the time domain waveform factor, the outer race fault frequency in the frequency domain envelope spectrum, the number of HHT transient jumps in the time and frequency domains, and the lubricant viscosity to obtain the waveform factor index, the outer race fault frequency amplitude index, the transient jump number index, and the lubricant viscosity index;

[0022] A bearing fault assessment model is constructed based on the form factor index, outer race fault frequency amplitude index, transient jump number index, and lubricating oil viscosity index, and the bearing fault evaluation coefficient is imported and output;

[0023] Compare the acquired bearing fault evaluation coefficient with the preset fault evaluation coefficient threshold value. If the fault evaluation coefficient is not within the fault evaluation coefficient threshold value, maintenance information is generated.

[0024] The bearing fault assessment model is expressed as:

[0025]

[0026] Among them, C bea Indicates the bearing fault evaluation coefficient, F norm Represents the waveform factor index, A norm Indicates the frequency amplitude index of the outer race fault, J HHT Indicates the HHT transient jump number index, μ norm represents the viscosity index of the lubricating oil, and w1, w2, w3, and w4 represent weight coefficients.

[0027] Further technical solution: Based on the waveform symmetry in the time domain, the meshing frequency harmonic amplitude in the frequency domain, and the energy density of the HHT marginal spectrum at 2× the rotation frequency in the time-frequency domain, the specific steps for constructing a shaft alignment model and outputting the shaft alignment evaluation coefficient are as follows:

[0028] The waveform symmetry in the time domain, the meshing frequency harmonic amplitude in the frequency domain, and the energy density of the HHT marginal spectrum at 2× the rotation frequency in the time-frequency domain are normalized to obtain the waveform symmetry index, meshing frequency harmonic amplitude index, and rotation frequency energy density index.

[0029] A shaft alignment model is constructed based on the waveform symmetry index, meshing frequency harmonic amplitude index, and rotational frequency energy density index, and the output shaft alignment evaluation coefficient is imported.

[0030] Compare the obtained alignment evaluation coefficient with the preset alignment evaluation coefficient threshold. If the alignment evaluation coefficient is not within the alignment evaluation coefficient threshold, maintenance information is generated.

[0031] The shaft alignment model is expressed as:

[0032] C ali =A1(1-D sym )+A2H norm +A3E sHHT

[0033] Among them, C ali Indicates the shaft alignment factor, D symRepresents the waveform symmetry index, H norm Indicates the meshing frequency harmonic amplitude index, E sHHT represents the rotational energy density index, A1, A2, and A3 represent weight coefficients, and A1+A2+A3=1.

[0034] Further technical solutions:.

[0035] Further technical solution: The gearbox motion coordination model is expressed as:

[0036]

[0037] Among them, K p Indicates the gearbox motion coordination, C gea Indicates the local damage coefficient of the gear, C bea Indicates the bearing fault evaluation coefficient, C ali represents the shafting alignment factor, α represents the lubricating oil evaluation factor, and β represents the deterioration acceleration factor.

[0038] Further technical solution: compare the gearbox movement coordination with the preset first coordination threshold, second coordination threshold and third coordination threshold. If the gearbox movement coordination is within the first coordination threshold, it indicates that the status is normal. If the gearbox movement coordination is within the second coordination threshold, it becomes a warning message. If the gearbox movement coordination is within the third threshold, it becomes a shutdown for maintenance.

[0039] Further technical solution: obtain shutdown and maintenance information and generate control instructions to shut down and repair the device where the gearbox is located.

[0040] The present invention provides a gearbox vibration analysis method, which has the following advantages compared with the prior art:

[0041] 1. The present invention constructs a comprehensive evaluation model by integrating multi-dimensional parameters in the time domain, frequency domain, and time-frequency domain, and establishes a motion coordination evaluation system based on the lubricating oil status. It effectively solves the problems of traditional methods with a single diagnostic dimension and insufficient coupled analysis of component status, and has the advantage of improving the accuracy of fault warning and maintenance efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 It is a schematic diagram of the process of the present invention. DETAILED DESCRIPTION

[0043] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0044] The specific implementation of the present invention is described in detail below with reference to specific embodiments.

[0045] See also Figure 1 , according to an embodiment of the present invention, a gearbox vibration analysis method is provided, comprising the following steps:

[0046] Obtaining time domain state parameters, frequency domain state parameters, and time-frequency domain state parameters;

[0047] According to the time domain kurtosis, the frequency domain meshing frequency sideband interval and number, and the time and frequency domain ridge smoothness, a gear local damage model is constructed to output the gear local damage evaluation coefficient;

[0048] A bearing fault assessment model is constructed based on the time domain waveform factor, the outer race fault frequency amplitude in the frequency domain envelope spectrum, and the number of HHT transient jumps in the time and frequency domains to output the bearing fault evaluation coefficient.

[0049] The shaft alignment model is constructed based on the waveform symmetry in the time domain, the meshing frequency harmonic amplitude in the frequency domain, and the energy density of the HHT marginal spectrum at 2× the rotation frequency in the time-frequency domain to output the shaft alignment evaluation coefficient.

[0050] Perform weighted average processing based on the current lubricating oil temperature and cleanliness to obtain the lubricating oil evaluation coefficient;

[0051] Based on the shaft alignment evaluation coefficient, bearing fault evaluation coefficient, and gear local damage evaluation coefficient under the condition of lubricating oil influence coefficient, a gearbox motion coordination model is constructed to output the gearbox motion coordination degree;

[0052] The gearbox motion coordination is compared with the preset first, second, and third thresholds. If the gearbox motion coordination is within the first threshold, it indicates normal status. If the gearbox motion coordination is within the second threshold, it becomes a warning message (monitoring is required). If the gearbox motion coordination is within the third threshold, it is a shutdown for maintenance. By establishing three progressive threshold intervals, the gearbox motion coordination is divided into three states: normal, warning, and fault. When the coordination is within the first threshold range, the gearbox is judged to be in normal working condition and routine monitoring is continued. When the coordination exceeds the first threshold but does not reach the third threshold, it is judged that there is a potential risk, triggering a warning message and recommending preventive maintenance inspections. When the coordination exceeds the third threshold, it is judged that there is a serious failure risk, and a shutdown instruction is directly triggered to cut off the equipment operation. This process achieves a graded response from status monitoring to emergency intervention by dynamically comparing the real-time coordination with the numerical range of the preset threshold.

[0053] Among them, time-domain state parameters include kurtosis, waveform symmetry, and form factor. These can be calculated using statistical calculations of vibration signals and are used to characterize the impact strength, symmetry, and shape characteristics of the vibration waveform. Frequency-domain state parameters include the meshing frequency sideband spacing, the number of meshing frequency sidebands, the outer race fault frequency amplitude in the envelope spectrum, and the meshing frequency harmonic amplitude. These can be calculated using fast Fourier transforms and envelope demodulation analysis to detect modulation phenomena and fault frequency characteristics. Time-frequency domain state parameters include ridge smoothness, the number of HHT transient jumps, and the energy density of the HHT marginal spectrum at 2× the rotation frequency. These can be calculated using the Hilbert-Huang transform and a time-frequency ridge extraction algorithm to capture the transient characteristics and energy distribution patterns of non-stationary signals. The lubricant evaluation coefficient, obtained from data from a temperature sensor and a particle counter and calculated using a weighted average algorithm, is used to quantify the impact of lubrication conditions on mechanical component wear.

[0054] Specifically, the gear local damage model uses time-domain kurtosis to detect gear impact vibration intensity, combines frequency-domain sideband spacing and number to identify modulation phenomena, and uses time-frequency domain ridge smoothness to assess energy distribution anomalies during gear meshing. These three complementary methods can avoid misjudgments caused by single parameters. The bearing fault assessment model uses time-domain waveform factors to detect waveform distortion caused by bearing raceway damage, frequency-domain outer race fault frequency amplitude to locate the fault location, and time-frequency domain HHT transient jump count to capture transient impacts caused by bearing defects, comprehensively improving fault detection sensitivity. The shaft alignment model uses time-domain waveform symmetry to identify asymmetric vibration caused by shaft misalignment, frequency-domain meshing frequency harmonic amplitude to assess gear meshing anomalies, and time-frequency domain rotational frequency energy density to analyze the coupling strength of shaft vibration, enabling accurate assessment of misalignment. The lubricant evaluation coefficient is calculated by weighting temperature and cleanliness to reflect the acceleration or reduction effect of lubrication conditions on the wear rate of mechanical components. Finally, a gearbox motion coordination model dynamically correlates component status parameters with the lubricant influence coefficient, establishing a comprehensive health assessment mechanism based on multi-domain parameter fusion.

[0055] Compared with existing technologies, traditional methods use single-domain parameters for isolated analysis. For example, using only time-domain kurtosis to determine gear damage is incapable of distinguishing between normal impacts and fault impacts. However, this solution accurately distinguishes between normal gear meshing impacts and local damage impacts through the multi-dimensional fusion of time-domain kurtosis, frequency-domain sideband characteristics, and time-frequency-domain ridge smoothness. Existing bearing fault diagnosis relies on envelope spectrum analysis, which makes it difficult to detect early, weak faults. This solution combines the time-domain waveform factor with the number of HHT transient jumps to capture waveform distortion and transient impacts caused by early bearing damage. Existing shaft alignment diagnosis lacks multi-domain parameter correlation. This solution identifies the coupling effect of shaft misalignment and gear meshing anomalies through the coordinated analysis of time-domain waveform symmetry, frequency-domain harmonic amplitude, and time-frequency-domain rotational energy density. In addition, existing technologies do not consider the dynamic impact of lubricant oil status on component degradation. This solution, by introducing a lubricant oil evaluation coefficient, can quantify the accelerating effect of lubrication conditions on the evolution of each component's faults.

[0056] Through the above technical solutions, this application can solve the problems of high misjudgment rate, insufficient sensitivity and lack of overall evaluation caused by single parameters in traditional methods. The local damage evaluation coefficient of gears can distinguish the impact characteristics of normal working conditions and fault conditions through multi-domain feature fusion; the bearing fault evaluation coefficient combines time domain waveform distortion and transient impact detection to improve the ability to identify early faults; the shaft alignment evaluation coefficient can accurately diagnose shaft deflection faults through vibration symmetry and energy coupling analysis; the lubricating oil evaluation coefficient dynamically reflects the weight of the influence of lubrication conditions on component wear. Finally, the status of each component and the lubrication conditions are integrated through a collaborative model to achieve a quantitative assessment of the overall health status of the gearbox, avoiding chain failures caused by local faults not being detected in time.

[0057] Preferably, the time domain state parameters include kurtosis, waveform symmetry and form factor, the frequency domain state parameters include meshing frequency sideband spacing, the number of meshing frequency sidebands, the outer ring fault frequency amplitude in the envelope spectrum and the meshing frequency harmonic amplitude, and the time-frequency domain state parameters include ridge smoothness, the number of HHT instantaneous jumps, and the energy density of the HHT marginal spectrum at 2× the rotation frequency.

[0058] Preferably, the specific steps of constructing a gear local damage model and outputting a gear local damage evaluation coefficient based on the time domain kurtosis, the frequency domain meshing frequency sideband interval and number, and the time-frequency domain ridge smoothness are:

[0059] Perform maximum-minimum normalization processing on the obtained time domain kurtosis, frequency domain meshing frequency sideband interval and number, and time-frequency domain ridge smoothness to obtain the kurtosis index, sideband interval index, sideband number index, and ridge smoothness index;

[0060] According to the kurtosis index, sideband spacing index, sideband number index and ridge smoothness index, a gear local loss damage model is constructed and the gear local damage coefficient is imported and output;

[0061] Compare the obtained gear local damage coefficient with the preset damage coefficient threshold. If the gear local damage coefficient is not within the damage coefficient threshold, it becomes maintenance information;

[0062] The gear local damage model is expressed as:

[0063]

[0064] Among them, C gea Indicates the local damage coefficient of the gear, K norm represents the time domain kurtosis index, N norm represents the sideband number index, Δf norm represents the sideband spacing index, S norm Represents the ridge smoothness index.

[0065] Specifically, the multi-dimensional characteristic parameters are normalized to eliminate the influence of different physical dimensions on the evaluation results. The kurtosis index and the sideband number index serve as positive factors, reflecting the impact energy and modulation effect intensity caused by gear damage, respectively. The sideband spacing index and the ridge smoothness index serve as suppression factors in the denominator, characterizing the discreteness of the fault modulation frequency distribution and the degree of fluctuation of the time-frequency characteristics, respectively. The four types of indices are coupled and calculated using a nonlinear formula to achieve a comprehensive assessment of the gear health status. When the local damage coefficient exceeds the preset threshold, it indicates abnormal wear or cracks on the gear, triggering a maintenance instruction.

[0066] Compared with existing technologies, traditional methods rely solely on a single parameter, time-domain kurtosis, to determine gear damage, without considering frequency-domain sideband characteristics or time-frequency stability indicators. This method is prone to misjudgment due to environmental noise interference or fluctuations in normal operating conditions. This solution constructs a multidimensional evaluation model by integrating time-domain, frequency-domain, and time-frequency domain characteristics. This model not only captures the characteristics of gear impact vibration, but also identifies fault modulation effects and time-frequency distribution anomalies, significantly improving the accuracy of damage determination.

[0067] Through the above technical solution, this application effectively solves the misjudgment problem caused by isolated parameter analysis in traditional gear damage assessment. Through multi-domain feature fusion and nonlinear coupling calculation, it realizes accurate quantitative assessment of local gear damage, providing a reliable basis for gearbox maintenance decision-making.

[0068] Preferably, the specific steps of constructing a bearing fault assessment model and outputting a bearing fault evaluation coefficient based on the time domain waveform factor, the outer race fault frequency in the frequency domain envelope spectrum, and the number of HHT transient jumps in the time and frequency domains are:

[0069] Perform maximum - minimum normalization on the time - domain waveform factor, the outer - race fault frequency in the frequency - domain envelope spectrum, the number of instantaneous jumps in the HHT of the time - frequency domain, and the lubricant viscosity, and obtain the waveform factor index, the outer - race fault frequency amplitude index, the instantaneous jump number index, and the lubricant viscosity index;

[0070] Construct a bearing fault assessment model based on the waveform factor index, the outer - race fault frequency amplitude index, the instantaneous jump number index, and the lubricant viscosity index, and import and output the bearing fault evaluation coefficient;

[0071] Compare the obtained bearing fault evaluation coefficient with the preset fault evaluation coefficient threshold. If the fault evaluation coefficient is not within the fault evaluation coefficient threshold, generate maintenance information;

[0072] The bearing fault assessment model is expressed as:

[0073]

[0074] where, C bea represents the bearing fault evaluation coefficient, F norm represents the waveform factor index, A norm represents the outer - race fault frequency amplitude index, J HHT represents the HHT instantaneous jump number index, μ norm represents the lubricant viscosity index, and w1, w2, w3, w4 represent weight coefficients.

[0075] Specifically, first perform multi - dimensional feature extraction on the original vibration signal in the time - domain, frequency - domain, and time - frequency domain. Eliminate the dimension difference through maximum - minimum normalization and highlight the fault - sensitive parameters. The time - domain waveform factor reflects the overall distortion of the bearing vibration waveform. The outer - race fault frequency amplitude in the frequency - domain envelope spectrum is directly related to the degree of outer - race damage. The number of instantaneous jumps in the time - frequency domain HHT captures the transient impact characteristics. The lubricant viscosity index is introduced into the denominator term of the model to reflect the acceleration effect of lubrication deterioration on faults. The weight coefficients can adjust the contribution degree of each parameter according to different bearing types. For example, in rolling bearings, w1 < w2 can be set to strengthen the outer - race fault frequency characteristics. The model aggregates multi - dimensional fault characteristics through the numerator term and introduces lubrication state compensation through the denominator term. The finally output evaluation coefficient realizes the coupled evaluation of mechanical damage and lubrication deterioration. When the evaluation coefficient exceeds the preset threshold, the maintenance instruction is automatically triggered.

[0076] Compared with existing technologies, traditional methods rely solely on envelope spectrum analysis to analyze outer race fault frequencies, failing to correlate time-domain waveform distortion characteristics with transient impact information in the time-frequency domain, leading to missed detection of early, weak faults. Existing technologies also ignore the nonlinear effects of lubricant viscosity changes on bearing vibration characteristics, making it impossible to accurately distinguish between mechanical damage and lubrication failure. This solution constructs a weighted assessment model by fusing multi-domain features. The HHT transient jump count enhances the detection of intermittent impacts, and the introduction of the lubricant viscosity index establishes a quantitative correlation between lubrication status and fault severity.

[0077] Through the above technical solution, this application effectively addresses the lack of sensitivity inherent in existing technologies due to single-dimensional analysis. By collaboratively analyzing the time-domain waveform factor and the frequency-domain fault amplitude, the accuracy of outer ring crack detection is improved. The HHT transient jump count enables timely detection of early pitting corrosion faults. Furthermore, a dynamic compensation mechanism for the lubricant viscosity index avoids misjudgments due to lubrication degradation, making bearing fault assessment results more accurate in line with actual operating conditions.

[0078] Preferably, the specific steps of constructing a shaft alignment model and outputting a shaft alignment evaluation coefficient based on the waveform symmetry in the time domain, the meshing frequency harmonic amplitude in the frequency domain, and the energy density of the HHT marginal spectrum at 2× the rotation frequency in the time-frequency domain are as follows:

[0079] The waveform symmetry in the time domain, the meshing frequency harmonic amplitude in the frequency domain, and the energy density of the HHT marginal spectrum at 2× the rotation frequency in the time-frequency domain are normalized to obtain the waveform symmetry index, meshing frequency harmonic amplitude index, and rotation frequency energy density index.

[0080] A shaft alignment model is constructed based on the waveform symmetry index, meshing frequency harmonic amplitude index, and rotational frequency energy density index, and the output shaft alignment evaluation coefficient is imported.

[0081] Compare the obtained alignment evaluation coefficient with the preset alignment evaluation coefficient threshold. If the alignment evaluation coefficient is not within the alignment evaluation coefficient threshold, maintenance information is generated.

[0082] The shaft alignment model is expressed as:

[0083] C ali =A1(1-D sym )+A2H norm +A3E sHHT

[0084] Among them, C ali Indicates the shaft alignment factor, D sym Represents the waveform symmetry index, H norm Indicates the meshing frequency harmonic amplitude index, E sHHTrepresents the rotational energy density index, A1, A2, and A3 represent weight coefficients, and A1+A2+A3=1.

[0085] Specifically, after the waveform symmetry calculation of the time-domain vibration signal, a decrease in its index value indicates that the shaft system has asymmetric vibration caused by radial or angular deviation; the abnormal increase in the amplitude of the meshing frequency harmonics in the frequency domain reflects the gear meshing phase misalignment caused by the shaft system displacement; the increase in energy density at 2 times the rotation frequency in the time-frequency domain represents the enhancement of periodic shock during the shaft system rotation. After eliminating dimensional differences through maximum-minimum normalization processing, the three indices are assigned dynamic weights and linearly combined. The weight coefficients are adjusted according to the specific working conditions and the sum is maintained at 1. When the shaft system alignment evaluation coefficient exceeds the preset threshold, it indicates that at least two dimensions of parameters are abnormal at the same time, effectively avoiding the misjudgment of a single parameter.

[0086] Compared with existing technologies, traditional shaft alignment diagnosis typically only detects time-domain vibration amplitude or frequency-domain rotational frequency components, cannot distinguish between misalignment types, and is susceptible to interference from load fluctuations. This solution constructs a multidimensional evaluation system by integrating time-domain waveform symmetry to identify static deviations, frequency-domain harmonic amplitude to detect dynamic displacements, and time-frequency domain energy density to capture transient impacts. This solves the problem of single-dimensional analysis being insufficiently sensitive to complex shaft alignment deviations. Furthermore, a dynamic weight allocation mechanism automatically adjusts parameter contributions based on different equipment characteristics, offering greater adaptability to operating conditions than fixed threshold methods.

[0087] Through the above technical solution, this application effectively improves the diagnostic accuracy of the shaft alignment status and can accurately identify shaft offset faults caused by foundation settlement, thermal expansion or assembly errors during gearbox operation. Through the collaborative analysis of multi-dimensional parameters, it reduces misjudgments caused by load fluctuations or random interference, providing a reliable quantitative assessment basis for equipment preventive maintenance.

[0088] Preferably, the gearbox motion coordination model is expressed as:

[0089]

[0090] Among them, K p Indicates the gearbox motion coordination, C gea Indicates the local damage coefficient of the gear, C bea Indicates the bearing fault evaluation coefficient, C ali represents the shafting alignment coefficient, α represents the lubricating oil evaluation coefficient, and β represents the deterioration acceleration factor (usually 0.1-0.3).

[0091] By establishing a gearbox motion coordination model, the state parameters of gears, bearings, and shaft alignment are coupled with the influence of lubricating oil. In this model, the numerator is calculated by the product form (1-C gea )、(1-Cbea )、(1-C ali ) comprehensively reflects the negative impact of gear damage, bearing failure and shaft alignment deviation on the system coordination. The lubricant evaluation coefficient α is used as the numerator to adjust the regulation effect of lubrication status on the overall coordination. The denominator is calculated by the cumulative form (C gea +C bea +C ali ) multiplied by the degradation acceleration factor β, characterizes the cumulative effect of multiple component failures and the interaction of degradation rates. This collaborative calculation structure of the numerator and denominator not only reflects the independent impact of a single component failure, but also quantifies the degradation acceleration effect of multiple coupled faults through the β parameter. This enables dynamic correlation modeling of mechanical component status and lubrication conditions, resolving the limitation of traditional methods in being unable to quantitatively assess the synergistic impact of multiple factors.

[0092] Preferably, the shutdown and maintenance information is obtained and a control instruction is generated to shut down and repair the device where the gearbox is located.

[0093] It should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that includes a list of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus.

[0094] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A gearbox vibration analysis method, characterized in that: The following steps are involved: Obtaining time domain state parameters, frequency domain state parameters, and time-frequency domain state parameters; According to the time domain kurtosis, the frequency domain meshing frequency sideband interval and number, and the time and frequency domain ridge smoothness, a gear local damage model is constructed to output the gear local damage evaluation coefficient; A bearing fault assessment model is constructed based on the time domain waveform factor, the outer race fault frequency amplitude in the frequency domain envelope spectrum, and the number of HHT transient jumps in the time and frequency domains to output the bearing fault evaluation coefficient. The shaft alignment model is constructed based on the waveform symmetry in the time domain, the meshing frequency harmonic amplitude in the frequency domain, and the energy density of the HHT marginal spectrum at 2× the rotation frequency in the time-frequency domain to output the shaft alignment evaluation coefficient. Perform weighted average processing based on the current lubricating oil temperature and cleanliness to obtain the lubricating oil evaluation coefficient; According to the shafting alignment evaluation coefficient, bearing fault evaluation coefficient and gear local damage evaluation coefficient under the condition of lubricating oil influence coefficient, a gearbox motion coordination model is constructed to output the gearbox motion coordination degree.

2. The gearbox vibration analysis method according to claim 1, characterized in that: The time domain state parameters include kurtosis, waveform symmetry and form factor; the frequency domain state parameters include meshing frequency sideband spacing, the number of meshing frequency sidebands, the outer ring fault frequency amplitude in the envelope spectrum and the meshing frequency harmonic amplitude; the time-frequency domain state parameters include ridge smoothness, the number of HHT instantaneous jumps, and the energy density of the HHT marginal spectrum at 2× the rotation frequency.

3. The gearbox vibration analysis method according to claim 2, characterized in that: The specific steps for constructing a gear local damage model and outputting a gear local damage evaluation coefficient based on the time domain kurtosis, the frequency domain meshing frequency sideband interval and number, and the time-frequency domain ridge smoothness are as follows: Perform maximum-minimum normalization processing on the obtained time domain kurtosis, frequency domain meshing frequency sideband interval and number, and time-frequency domain ridge smoothness to obtain the kurtosis index, sideband interval index, sideband number index, and ridge smoothness index; According to the kurtosis index, sideband spacing index, sideband number index and ridge smoothness index, a gear local loss damage model is constructed and the gear local damage coefficient is imported and output; Compare the obtained gear local damage coefficient with the preset damage coefficient threshold. If the gear local damage coefficient is not within the damage coefficient threshold, it becomes maintenance information; The gear local damage model is expressed as: Among them, C gea Indicates the local damage coefficient of the gear, K norm represents the time domain kurtosis index, N norm represents the sideband number index, Δf norm represents the sideband spacing index, S norm Represents the ridge smoothness index.

4. The gearbox vibration analysis method according to claim 2, characterized in that: The specific steps for constructing a bearing fault assessment model and outputting a bearing fault evaluation coefficient based on the time domain waveform factor, the outer race fault frequency in the frequency domain envelope spectrum, and the number of HHT transient jumps in the time and frequency domain are as follows: Perform maximum-minimum normalization processing on the time domain waveform factor, the outer race fault frequency in the frequency domain envelope spectrum, the number of HHT transient jumps in the time and frequency domains, and the lubricant viscosity to obtain the waveform factor index, the outer race fault frequency amplitude index, the transient jump number index, and the lubricant viscosity index; A bearing fault assessment model is constructed based on the form factor index, outer race fault frequency amplitude index, transient jump number index, and lubricating oil viscosity index, and the bearing fault evaluation coefficient is imported and output; Compare the acquired bearing fault evaluation coefficient with the preset fault evaluation coefficient threshold value. If the fault evaluation coefficient is not within the fault evaluation coefficient threshold value, maintenance information is generated. The bearing fault assessment model is expressed as: Among them, C bea Indicates the bearing fault evaluation coefficient, F norm Represents the waveform factor index, A norm Indicates the frequency amplitude index of the outer race fault, J HHT Indicates the HHT transient jump number index, μ norm represents the viscosity index of the lubricating oil, and w1, w2, w3, and w4 represent weight coefficients.

5. The gearbox vibration analysis method according to claim 2, characterized in that: The specific steps for constructing the shaft alignment model and outputting the shaft alignment evaluation coefficient based on the waveform symmetry in the time domain, the meshing frequency harmonic amplitude in the frequency domain, and the energy density of the HHT marginal spectrum at 2× the rotation frequency in the time-frequency domain are as follows: The waveform symmetry in the time domain, the meshing frequency harmonic amplitude in the frequency domain, and the energy density of the HHT marginal spectrum at 2× the rotation frequency in the time-frequency domain are normalized to obtain the waveform symmetry index, meshing frequency harmonic amplitude index, and rotation frequency energy density index. A shaft alignment model is constructed based on the waveform symmetry index, meshing frequency harmonic amplitude index, and rotational frequency energy density index, and the output shaft alignment evaluation coefficient is imported. Compare the obtained alignment evaluation coefficient with the preset alignment evaluation coefficient threshold. If the alignment evaluation coefficient is not within the alignment evaluation coefficient threshold, maintenance information is generated. The shaft alignment model is expressed as: C ali =A1(1-D sym )+A2H norm +A3E sHHT Among them, C ali Indicates the shaft alignment factor, D sym Represents the waveform symmetry index, H norm Indicates the meshing frequency harmonic amplitude index, E sHHT represents the rotational energy density index, A1, A2, and A3 represent weight coefficients, and A1+A2+A3=1.

6. The gearbox vibration analysis method according to claim 2, characterized in that: The gearbox motion coordination model is expressed as: Among them, K p Indicates the gearbox motion coordination, C gea Indicates the local damage coefficient of the gear, C bea Indicates the bearing fault evaluation coefficient, C ali represents the shafting alignment factor, α represents the lubricating oil evaluation factor, and β represents the deterioration acceleration factor.

7. The gearbox vibration analysis method according to claim 1, characterized in that: The gearbox motion coordination degree is compared with the preset first, second and third thresholds. If the gearbox motion coordination degree is within the first threshold, it indicates that the status is normal. If the gearbox motion coordination degree is within the second threshold, it becomes a warning message. If the gearbox motion coordination degree is within the third threshold, it requires shutdown for maintenance.

8. The gearbox vibration analysis method according to any one of claims 1 to 7, characterized in that: Obtain shutdown and maintenance information and generate control instructions to shut down and repair the device where the gearbox is located.

Images