A method for automatic diagnosis of gearbox failures
By calculating the gearbox's rotational frequency and meshing frequency, and combining peak outlier and spectral peak morphology significance indicators, an automatic diagnostic model is constructed. This solves the problems of convenience and robustness in gearbox fault diagnosis in existing technologies, and realizes automated fault identification and location under complex working conditions.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-08-31
- Publication Date
- 2026-03-24
AI Technical Summary
Existing technologies lack convenient and automated gearbox fault diagnosis technologies to cope with complex and variable operating conditions, resulting in high diagnosis time costs and poor robustness.
By calculating the rotational frequency and meshing frequency of the gear shaft, the spectrum of the vibration signal is obtained. Using the fault characteristic frequency identification method, combined with the peak outlier and spectral peak shape significance index, an automatic diagnostic model is constructed to realize the automatic identification and diagnosis of fault characteristic frequencies.
It enables automatic identification and location of gearbox faults under complex and variable operating conditions, improving the convenience and robustness of diagnosis and reducing reliance on professional knowledge and large amounts of sample data.
Smart Images
Figure CN115855483B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of gear fault diagnosis, specifically to an automatic fault diagnosis method for gearboxes. Background Technology
[0002] Gearboxes, with their advantages of fixed transmission ratio, large torque, and compact structure, have been widely used in mechanical transmission systems in industries such as wind power generation, aviation, shipbuilding, and transportation. During service, the transmission teeth of gearboxes are subjected to extremely high loads, leading to frequent damage such as tooth surface wear, tooth surface abrasion, and tooth root cracks. The occurrence and continuous deterioration of this damage severely impact and endanger the transmission accuracy and service safety of the gearbox. Therefore, developing automatic fault diagnosis for gearboxes to accurately identify and locate damage, and providing decision support for safe service and efficient maintenance, has significant engineering value.
[0003] Current research on gear fault mechanisms and feature extraction methods provides an important theoretical foundation and technical means for gear fault diagnosis. However, mechanism-based diagnosis requires users to have professional knowledge related to fault mechanisms and feature extraction. Conducting signal analysis and diagnostic decisions under the guidance of this professional knowledge is time-consuming and not conducive to widespread application in field maintenance and testing. Furthermore, due to the variability of the operating environment and conditions, it is difficult to pre-set the thresholds for feature indicators, resulting in weak robustness of automatic fault alarms and location. Therefore, even though mechanism-based gear fault diagnosis is relatively mature in principle and methodology, convenient and automated diagnostic technologies for handling complex and variable operating conditions are still lacking in engineering applications. Summary of the Invention
[0004] To address the problems existing in the prior art, this invention provides an automatic fault diagnosis method for gearboxes, which solves the problem that the prior art lacks convenient and automated diagnostic technology to cope with complex and changing working conditions.
[0005] This invention is achieved through the following technical solution:
[0006] An automatic fault diagnosis method for gearboxes includes the following steps:
[0007] Step 1: Calculate the rotational frequency of the gear shaft and the meshing frequency of the gearbox meshing pair;
[0008] Step 2: Acquire vibration signals from the drive end and non-drive end of various gearbox systems, perform signal processing on the vibration signals of various gearbox systems, and obtain the spectrum of the vibration signals; the spectrum of the vibration signals includes the meshing frequency and sidebands;
[0009] Step 3: Using the side frequencies of the first-order rotation frequency on the left and right sides of the meshing frequency, and the side frequencies of the second-order rotation frequency on the left and right sides of the meshing frequency, these four frequency converters are used as fault characteristic frequencies. The constructed fault characteristic frequency identification method is used to identify the four fault characteristic frequencies respectively, determine whether there is a fault component at the fault characteristic frequency, and obtain the correspondence between the fault characteristic frequency and the fault type.
[0010] Step 4: Construct an automatic diagnostic model based on the correspondence between fault characteristic frequencies and fault types. Using the automatic diagnostic model, analyze the vibration signals of various gearbox systems obtained from actual measurements and make fault diagnosis results.
[0011] Furthermore, in step 3, by observing the peak size and spectral peak at the fault characteristic frequency in the spectrum, a peak outlier index and a spectral peak shape significance index are constructed based on the two elements of peak size and peak shape. The peak outlier index and the spectral peak shape significance index are then fused to obtain a fault characteristic frequency identification method.
[0012] Furthermore, by fusing the peak outlier index and the spectral peak morphology significance index, a fault feature frequency identification method is obtained, expressed as follows:
[0013] CI i =αPI i +βSI i ≥λ;
[0014] Where PI i SI represents the peak outlier index at the i-th fault characteristic frequency. i The index representing the significance of the spectral peak shape at the i-th fault characteristic frequency;
[0015] When the fusion index CI i If the value is greater than λ, then it is determined that there is a fault characteristic component at the fault characteristic frequency; α, β and λ are constant coefficients.
[0016] Furthermore, the peak outlier index is constructed as follows: It evaluates whether the peak value is greater than the amplitude within the sideband range of the third-order turning frequencies to the left and right of the meshing frequency by using the ratio of the peak value at the fault characteristic frequency to the mean spectral value within the sideband range of the third-order turning frequencies to the left and right of the meshing frequency. The specific expression for the peak outlier index is as follows:
[0017]
[0018] Where PI i A represents the peak outlier index at the i-th fault characteristic frequency. is represents the spectral peak at the i fault characteristic frequencies, N represents the number of spectral lines in the third-order turn-frequency sidebands to the left and right of the meshing frequency, and s represents the number of spectral lines in the sidebands to the left and right of the meshing frequency. j This indicates the amplitude of the spectral line.
[0019] Furthermore, the significance index for spectral peak shape is constructed as follows: the significance of the spectral peak shape is evaluated by the ratio of the peak value at the fault characteristic frequency to the peak values at two discrete frequency points to the left and right of the peak value. The specific expression for the significance index of spectral peak shape is as follows:
[0020]
[0021] SI i A represents the significance index of the spectral peak shape at the i-th fault characteristic frequency. i s represents the spectral peaks at the i fault characteristic frequencies. i-1 This represents the amplitude at the frequency point to the left of the meshing frequency, s. i+1 This indicates the amplitude of the frequency point to the right of the meshing frequency.
[0022] Furthermore, in step 4, when it is determined that a fault feature component exists at a certain fault feature frequency, the identification result is recorded as 1; when it is determined that a fault feature component does not exist at a certain fault feature frequency, the identification result is recorded as 0.
[0023] When the identification result is 1, the fault feature component count is accumulated once;
[0024] Set the warning threshold for the measuring points, accumulate the counts of four fault characteristic components at the driving end and non-driving end respectively, and output the diagnostic results based on the fault characteristic component counts, specifically:
[0025] If the cumulative counts of the four fault feature components on the driving end and the non-driving end are less than the warning threshold, then the "no fault" diagnosis result is output.
[0026] If the count of fault characteristic components exceeds the warning threshold at either the driving end or the non-driving end, a diagnostic result of "fault reminder" will be output.
[0027] If the count of fault characteristic components exceeds the warning threshold at both the drive end and the non-drive end, the faulty gear is located based on the characteristic frequency value and the fault mechanism, and the "gear fault location" diagnostic result is output.
[0028] Furthermore, in step 1, the meshing frequency of each gearbox system is calculated using the following formula:
[0029] f m =Z1·f n1 =Z2·f n2
[0030] In the formula, fm f is the meshing frequency. n1 f n2 Z1 and Z2 are the rotational frequencies of the driving gear and driven gear of the gearbox, respectively, and the number of teeth of the driving gear and driven gear of the gearbox, respectively.
[0031] The formula for calculating the rotational frequency of the gear shaft is as follows:
[0032] f r =n r / 60
[0033] In the formula, n r This represents the rotational speed of the gear shaft.
[0034] Furthermore, in step 2, the vibration signals of various gearbox systems are subjected to fast Fourier transform to obtain the spectrum of the vibration signals.
[0035] Furthermore, in step 2, the meshing frequency is f m The rotational frequency of the gear shaft is denoted as f. r The sideband of the first-order turn frequency to the left of the meshing frequency is f. m- f r The sideband of the first-order turn frequency to the right of the meshing frequency is f. m+ f r ;
[0036] The sideband of the second-order turn frequency to the left of the meshing frequency is f. m-2 f r The sideband of the second-order turn frequency to the right of the meshing frequency is f. m+2 f r .
[0037] Compared with the prior art, the present invention has the following beneficial technical effects:
[0038] This invention discloses an automatic fault diagnosis method for gearboxes. Based on the vibration mechanism of gearbox faults, it proposes a fault characteristic frequency identification method, constructs peak outlier indices and spectral peak morphology significance indices, and achieves automatic identification of fault characteristic frequencies. It also constructs a threshold-free automatic diagnosis technology framework based on fault characteristic frequency identification and automatic fault diagnosis logic, achieving automatic gear fault identification and location without relying on a large number of fault samples, and overcoming interference from identification errors, amplitude fluctuations caused by environmental and operating condition changes. This method aims to improve diagnostic convenience and robustness to adapt to complex and changing service environments and meet engineering application needs. Attached Figure Description
[0039] Figure 1 This is the vibration data acquisition system described in the example of the present invention.
[0040] Figure 2The images shown are of small gears with different fault types as described in the examples of this invention; Figure a represents a normal gear, Figure b represents a gear with minor cracks, Figure c represents a gear with root cracks, and Figure d represents a gear with missing tooth tips.
[0041] Figure 3a This is the vibration response spectrum of the measuring point of the drive end axle box rotating arm in the example of the present invention;
[0042] Figure 3b This is the vibration response spectrum of the non-drive end axle box arm measuring point described in the example of the present invention;
[0043] Figure 3c This is a partially enlarged view of the vibration response spectrum of the measuring point of the drive end axle box rotating arm in the example of the present invention;
[0044] Figure 3d This is a partial magnification of the vibration response spectrum of the non-drive end axle box arm measuring point in the example of the present invention;
[0045] Figure 4 This is the conceptual framework of the fault feature frequency identification method described in the examples of this invention;
[0046] Figure 5 This is the threshold-free automatic diagnosis technology framework described in the examples of this invention. Detailed Implementation
[0047] The present invention will be further described in detail below with reference to specific embodiments. These descriptions are for explanation purposes only and are not intended to limit the scope of the invention.
[0048] The automatic fault diagnosis method for gearboxes is based on the vibration response mechanism of gearbox faults to construct an anti-interference automatic diagnosis method. Under the conditions of limited fault samples and incomplete fault modes, it can realize the automatic diagnosis of gearbox faults, providing equipment maintenance personnel with convenient and robust fault diagnosis application services.
[0049] In this preferred embodiment, a bogie-fidelity fault simulation test bench is used as the object to verify the effectiveness of the automatic fault diagnosis method for gearboxes.
[0050] The vibration data acquisition system is used to collect vibration signals from the bogie gears, such as... Figure 1 As shown, it mainly includes a vibration acceleration sensor, a data acquisition unit, and a data acquisition control and analysis system. The data acquisition control and analysis system can realize the setting of acquisition parameters, acquisition control, data display, data storage, and data analysis and visualization.
[0051] This invention discloses an automatic fault diagnosis method for gearboxes, comprising the following steps:
[0052] Step 1: Calculate the rotational frequency of the gear shaft and the meshing frequency of the gearbox meshing pair; the formula is as follows:
[0053] f m =Z1·f n1 =Z2·f n2
[0054] In the formula, f m f is the meshing frequency. n1 f n2 Z1 and Z2 are the rotational frequencies of the driving gear and driven gear of the gearbox, respectively, and the number of teeth of the driving gear and driven gear of the gearbox, respectively.
[0055] The formula for calculating the rotational frequency of the gear shaft is as follows:
[0056] f r =n r / 60
[0057] In the formula, n r This represents the rotational speed of the gear shaft.
[0058] Step 2: Acquire vibration signals from the drive end and non-drive end of various gearbox systems, perform signal processing on the vibration signals of various gearbox systems, and obtain the spectrum of the vibration signals; the spectrum of the vibration signals includes the meshing frequency and sidebands;
[0059] The meshing frequency is f m The rotational frequency of the gear shaft is f r Then the sideband of the first-order turning frequency to the left of the meshing frequency is f. m- f r The sideband of the first-order turn frequency to the right of the meshing frequency is f. m+ f r The sideband of the second-order turning frequency to the left of the meshing frequency is f. m-2 f r The sideband of the second-order turn frequency to the right of the meshing frequency is f. m+2 f r .
[0060] Step 3: Using the side frequencies of the first-order rotation frequency on the left and right sides of the meshing frequency, and the side frequencies of the second-order rotation frequency on the left and right sides of the meshing frequency, these four frequency converters are used as fault characteristic frequencies. The constructed fault characteristic frequency identification method is used to identify the four fault characteristic frequencies respectively, determine whether there is a fault component at the fault characteristic frequency, and obtain the correspondence between the fault characteristic frequency and the fault type.
[0061] Step 4: Construct an automatic diagnostic model based on the correspondence between fault characteristic frequencies and fault types. Using the automatic diagnostic model, analyze the vibration signals of various gearbox systems obtained from actual measurements and make fault diagnosis results.
[0062] Specifically, in step 3, by observing the peak size and spectral peak at the fault characteristic frequency in the spectrum, a peak outlier index and a spectral peak shape significance index are constructed based on the two elements of peak size and peak shape. The peak outlier index and the spectral peak shape significance index are then fused to obtain the fault characteristic frequency identification method.
[0063] By fusing the peak outlier index and the spectral peak shape significance index, a fault feature frequency identification method is obtained, expressed as follows:
[0064] CI i =αPI i +βSI i ≥λ;
[0065] Where PI i SI represents the peak outlier index at the i-th fault characteristic frequency. i The index representing the significance of the spectral peak shape at the i-th fault characteristic frequency;
[0066] When the fusion index CI i When the value is greater than λ, it is determined that there is a fault characteristic component at the fault characteristic frequency; α, β and λ are constant coefficients, which are determined based on the specific values and experience of several typical fault spectra.
[0067] The peak outlier index is constructed as follows: It evaluates whether the peak value at the fault characteristic frequency is greater than the amplitude within the sideband range of the third-order turning frequencies to the left and right of the meshing frequency by comparing the ratio of the peak value at that frequency to the mean of the spectral lines within the sideband range of the third-order turning frequencies to the left and right of the meshing frequency. The specific expression for the peak outlier index is as follows:
[0068]
[0069] Where PI i A represents the peak outlier index at the i-th fault characteristic frequency. i s represents the spectral peak at the i fault characteristic frequencies, N represents the number of spectral lines in the third-order turn-frequency sidebands to the left and right of the meshing frequency, and s represents the number of spectral lines in the sidebands to the left and right of the meshing frequency. j This indicates the amplitude of the spectral line.
[0070] The significance index for spectral peak shape is constructed as follows: the ratio of the peak value at the fault characteristic frequency to the peak values at two discrete frequency points to the left and right of the peak value is used to evaluate the significance of the spectral peak shape. The specific expression for the significance index of spectral peak shape is as follows:
[0071]
[0072] SI iA represents the significance index of the spectral peak shape at the i-th fault characteristic frequency. i s represents the spectral peaks at the i fault characteristic frequencies. i-1 This represents the amplitude at the frequency point to the left of the meshing frequency, s. i+1 This indicates the amplitude of the frequency point to the right of the meshing frequency.
[0073] The process of building an automatic diagnostic model is as follows: when it is determined that a fault feature component exists at a certain fault feature frequency, the identification result is recorded as 1; when it is determined that a fault feature component does not exist at a certain fault feature frequency, the identification result is recorded as 0.
[0074] When the identification result is 1, the fault feature component count is accumulated once;
[0075] Set the warning threshold for the measuring points, accumulate the counts of four fault characteristic components at the driving end and non-driving end respectively, and output the diagnostic results based on the fault characteristic component counts, specifically:
[0076] If the cumulative counts of the four fault feature components on the driving end and the non-driving end are less than the warning threshold, then the "no fault" diagnosis result is output.
[0077] If the count of fault characteristic components exceeds the warning threshold at either the driving end or the non-driving end, a diagnostic result of "fault reminder" will be output.
[0078] If the count of fault characteristic components exceeds the warning threshold at both the drive end and the non-drive end, the faulty gear is located based on the characteristic frequency value and the fault mechanism, and the "gear fault location" diagnostic result is output.
[0079] The features and performance of the present invention will be further described in detail below with reference to embodiments.
[0080] Using the pinion as the test object, typical gear damage of different fault types was fabricated, such as... Figure 2 Figures 2a-2d show four typical gear conditions: normal, minor scratches, tooth root cracks, and tooth tip chipping, respectively. Vibration response data for different health conditions were acquired using vibration accelerometers. Considering vibration transmission and the feasibility of measurement points, vibration accelerometers were placed at two locations on the drive-end axle box arm and the non-drive-end axle box arm, as shown in Figure 3, to acquire vibration response data at each measurement point.
[0081] The motor speed is set to 1000 RPM using the controller, but the actual output speed differs from the controlled value. The pinion gear has 21 teeth, and the gear has 75 teeth. Therefore, the rotational frequency of the pinion shaft is approximately f... r =16.6Hz, the gear meshing frequency is approximately f m=348Hz. According to the mechanism analysis, the fault characteristic of the pinion is the appearance of sideband components of the pinion's rotational frequency on both sides of the meshing frequency.
[0082] The acquired data underwent spectral analysis, as shown in Figure 3. Figure 3a Corresponding drive end axle box swing arm measuring point, Figure 3b Corresponding to the measuring point of the non-drive end axle box swing arm, Figure 3c and Figure 3d They are respectively Figure 3a and Figure 3b A local magnification.
[0083] As can be seen from Figure 3, the meshing frequency f of a normal gear m There are very few peak values at the engagement frequency, and there are no significant sidebands on either side of the engagement frequency, i.e., f m+ f r f m+2 f r f m- f r f m-2 f r The peak value at the same frequency is not significant. Conversely, for the faulty gears in the other three states, it can be seen from the figure that they all have significant meshing frequencies, and the peak value is at the fault characteristic frequency f of the pinion. m+ f r f m+2 f r f m- f r f m-2 f r There are intuitive and significant frequency components present.
[0084] Based on the rotational frequency of the pinion, at four fault characteristic frequency components (f) from low to high on both sides of the meshing frequency. m- 2f r f m- f r f m+ f r f m+2 f r (i) refers to Fre1, Fre2, Fre3 and Fre4 in Tables 1 and 2, and the feature indices PI and SI for frequency component identification are extracted.
[0085] The proposed fault feature frequency identification method has the following conceptual framework: Figure 4 As shown, the peak outlier index PI and the peak shape significance index SI are calculated for the normal gear drive end and the scuffed gear drive end.
[0086] In this example, we take α = 0.8 and β = 0.2, and use the formula CI. i =αPIi +βSI i The fusion index CI corresponding to each characteristic frequency was calculated and is shown in Table 1 and Table 2, respectively. When a fault characteristic component is determined to exist at a certain fault characteristic frequency, the identification result is recorded as 1; when a fault characteristic component is determined not to exist at a certain fault characteristic frequency, the identification result is recorded as 0.
[0087] Table 1 Characteristic Frequency Identification of Normal Gear Drive End
[0088]
[0089] Table 2 Identification of Characteristic Frequencies of Scratched Gear Drive End
[0090]
[0091] Based on the α, β, and λ values set for the normal and scratched data at the drive end, fault characteristic frequency components were identified in the vibration acceleration signals of the gear and two measuring points at the drive and non-drive ends for four healthy states: normal, scratched, chipped, and cracked. The characteristic frequency components with an identification result of "1" at each measuring point were statistically counted. The results are shown in Table 3. From the characteristic frequency component counts, it can be seen that the number of characteristic frequency components identified at the two measuring points of the gear in the normal state is only 1 (this is not always 0 due to the influence of noise, environment, and other factors), while the faulty gear has more than 7 characteristic frequency components counted at both measuring points. Based on the constructed automatic diagnostic model, such as... Figure 5 As shown, the warning threshold for both the driving end and the non-driving end is set to 2. That is, when the count of the fault characteristic frequency component at a certain end exceeds 2, the channel is determined to be abnormal. When an abnormality occurs at a certain end, a warning for that fault is issued; when both the driving end and the non-driving end are abnormal, it is determined that there is a fault corresponding to the fault characteristic frequency, and the gear fault is located according to the gear to which the characteristic frequency belongs.
[0092] In this example, the characteristic frequency component counts at both the driving and non-driving ends of a normal pinion are less than 2, therefore there is no "fault warning" state, and the final fault determination state is "no fault". However, for the three fault states of pinion scratches, chipped corners, and cracks, the characteristic frequency component counts at both the driving and non-driving ends are greater than 2, and "fault location" occurs at both measurement points. According to the automatic diagnostic model, a pinion fault can be determined. Therefore, this method does not require extensive manual data analysis or rely on a large amount of sample data, and can achieve interference-resistant automatic diagnosis of gear faults.
[0093] Table 3. Pinion Fault Diagnosis
[0094]
Claims
1. An automatic fault diagnosis method for gearboxes, characterized in that, Includes the following steps: Step 1: Calculate the rotational frequency of the gear shaft and the meshing frequency of the gearbox meshing pair; Step 2: Acquire vibration signals from the drive end and non-drive end of various gearbox systems, perform signal processing on the vibration signals of various gearbox systems, and obtain the spectrum of the vibration signals; the spectrum of the vibration signals includes the meshing frequency and sidebands; Step 3: Using the side frequencies of the first-order rotation frequency on the left and right sides of the meshing frequency, and the side frequencies of the second-order rotation frequency on the left and right sides of the meshing frequency, these four frequency converters are used as fault characteristic frequencies. The constructed fault characteristic frequency identification method is used to identify the four fault characteristic frequencies respectively, determine whether there is a fault component at the fault characteristic frequency, and obtain the correspondence between the fault characteristic frequency and the fault type. Step 4: Construct an automatic diagnostic model based on the correspondence between fault characteristic frequencies and fault types. Use the automatic diagnostic model to analyze the vibration signals of various gearbox systems obtained from actual measurements and make fault diagnosis results. In step 3, by observing the peak size and spectral peak at the fault characteristic frequency in the spectrum, the peak outlier index and the spectral peak shape significance index are constructed based on the two elements of peak size and peak shape. The peak outlier index and the spectral peak shape significance index are then fused to obtain the fault characteristic frequency identification method. The peak outlier index is constructed as follows: It evaluates whether the peak value at the fault characteristic frequency is greater than the amplitude within the sideband range of the third-order turning frequencies to the left and right of the meshing frequency by comparing the ratio of the peak value at that frequency to the mean of the spectral lines within the sideband range of the third-order turning frequencies to the left and right of the meshing frequency. The specific expression for the peak outlier index is as follows: ； in, This represents the peak outlier index at the i-th fault characteristic frequency. Let represent the spectral peak values at the i fault characteristic frequencies, and N represent the number of spectral lines in the sideband range of the third-order turning frequencies to the left and right of the meshing frequency. Indicates the amplitude of the spectral line; The significance index for spectral peak shape is constructed as follows: the ratio of the peak value at the fault characteristic frequency to the peak values at two discrete frequency points to the left and right of the peak value is used to evaluate the significance of the spectral peak shape. The specific expression for the significance index of spectral peak shape is as follows: ； in, This represents the significance index of the spectral peak shape at the i-th fault characteristic frequency. This represents the spectral peak at the i characteristic frequencies of the fault. This indicates the amplitude of the frequency point to the left of the meshing frequency. This indicates the amplitude of the frequency point to the right of the meshing frequency.
2. The automatic gearbox fault diagnosis method according to claim 1, characterized in that, By fusing the peak outlier index and the spectral peak shape significance index, a fault feature frequency identification method is obtained, expressed as follows: ； When fusion index Greater than If so, it is determined that a fault characteristic component exists at the fault characteristic frequency; , and These are constant coefficients.
3. The automatic gearbox fault diagnosis method according to claim 1, characterized in that, In step 4, when it is determined that a fault feature component exists at a certain fault feature frequency, the identification result is recorded as 1; when it is determined that a fault feature component does not exist at a certain fault feature frequency, the identification result is recorded as 0. When the identification result is 1, the fault feature component count is accumulated once; Set the warning threshold for the measuring points, accumulate the counts of four fault characteristic components at the driving end and non-driving end respectively, and output the diagnostic results based on the fault characteristic component counts, specifically: If the cumulative counts of the four fault feature components on the driving end and the non-driving end are less than the warning threshold, then the "no fault" diagnosis result is output. If the count of fault characteristic components exceeds the warning threshold at either the driving end or the non-driving end, a "fault reminder" diagnostic result will be output. If the count of fault characteristic components exceeds the warning threshold at both the drive end and the non-drive end, the faulty gear is located based on the characteristic frequency value and the fault mechanism, and the "gear fault location" diagnostic result is output.
4. The automatic gearbox fault diagnosis method according to claim 1, characterized in that, In step 1, the meshing frequency of each gearbox system is calculated using the following formula: fm=Z1·fn1=Z2·fn2 ； In the formula, f m The meshing frequency, f n1 , f n2 These are the rotational frequencies of the driving gear and driven gear of the gearbox, respectively. Z 1. Z 2 represents the number of teeth on the driving gear and driven gear of the gearbox, respectively; The formula for calculating the rotational frequency of the gear shaft is as follows: f r =n r / 60 ； In the formula, n r This represents the rotational speed of the gear shaft.
5. The automatic gearbox fault diagnosis method according to claim 1, characterized in that, In step 2, the vibration signals of various gearbox systems are subjected to fast Fourier transform to obtain the spectrum of the vibration signals.
6. The automatic gearbox fault diagnosis method according to claim 1, characterized in that, In step 2, the meshing frequency is f m The rotational frequency of the gear shaft is denoted as f r The sideband of the first-order turn frequency to the left of the meshing frequency is... f m -f r ； The sideband of the first-order turn frequency to the right of the meshing frequency is... f m +f r ； The sideband of the second-order turn frequency to the left of the meshing frequency is... f m -2f r ； The sideband of the second-order turn-frequency to the right of the meshing frequency is... f m + 2f r 。