Bearing fault diagnosis method based on signal analysis

Abstract

The invention belongs to the field of bearing fault diagnosis, relates to a data analysis technology, and is used for solving the problem that a mode of performing fault diagnosis and analysis by adopting a fixed algorithm in the prior art is not suitable for all bearing types and application scenes, in particular to a bearing fault diagnosis method based on signal analysis. Comprising the following steps: carrying out fault diagnosis test on a bearing: marking a to-be-tested bearing as a test object, allocating an application scene to the test object, obtaining the type of the test object, obtaining the life duration of the test object according to the bearing type and the application scene, and carrying out fault diagnosis test on the bearing; generating a test period of which the duration is equal to the service life duration of the test object, dividing the test period into a plurality of test time periods, and putting the test object into an application scene for running test; according to the invention, the diagnosis model of the test object is dynamically switched according to the diagnosis validity of the test time period, so that sufficient diagnosis validity data is provided for bearings of different types and different application scenes, and the result accuracy of mode analysis is improved.

Description

Technical Field

[0001] The present invention belongs to the field of bearing fault diagnosis, involves data analysis technology, and specifically is a bearing fault diagnosis method based on signal analysis. Background Art

[0002] Bearing fault diagnosis is to identify and judge various abnormal conditions that occur during the operation of bearings through various detection means and technologies, so as to take corresponding maintenance and repair measures to ensure the normal operation of the equipment. Accurate fault diagnosis can help detect potential problems of bearings in time, such as cracks and wear, thus avoiding more serious faults and ensuring the safe and stable operation of the equipment.

[0003] The invention patent with the publication number CN117890111A discloses a training method of a bearing fault diagnosis model, a bearing fault diagnosis method and a device. This fault diagnosis method trains a pre-established bearing fault diagnosis model based on a second fusion feature to obtain a trained bearing fault diagnosis model; wherein, the input data of the pre-established bearing fault diagnosis model is the second fusion feature, and the training label includes the fault type of the first bearing; it can effectively improve the accuracy of bearing fault diagnosis; however, this fault diagnosis method can only use a single fault diagnosis model to perform bearing fault diagnosis processing, and the method of using a fixed algorithm for fault diagnosis analysis cannot be applied to all bearing types and application scenarios, resulting in the inability to improve the overall fault diagnosis accuracy.

[0004] In view of the above technical problems, this application proposes a solution. Summary of the Invention

[0005] The purpose of the present invention is to provide a bearing fault diagnosis method based on signal analysis to solve the problem that the method of using a fixed algorithm for fault diagnosis analysis in the prior art cannot be applied to all bearing types and application scenarios;

[0006] The technical problem that the present invention needs to solve is: how to provide a bearing fault diagnosis method based on signal analysis that can match the diagnosis model according to the bearing type and application scenario.

[0007] The purpose of the present invention can be achieved by the following technical solutions:

[0008] The bearing fault diagnosis method based on signal analysis includes the following steps:

[0009] Step 1: Conduct a fault diagnosis test on the bearing: Mark the bearing to be tested as the test object, assign an application scenario to the test object, obtain the type of the test object, obtain the service life duration of the test object based on the bearing type and application scenario, generate a test cycle with a duration equal to the service life duration of the test object and divide the test cycle into several test periods, and put the test object into the application scenario for operation testing;

[0010] Step 2: Assign a diagnosis model at the start time of the test period;

[0011] Step 3: Conduct an analysis of the effectiveness of the fault diagnosis at the end time of the test period;

[0012] Step 4: Conduct a diagnosis mode analysis at the end time of the test cycle: Mark all the test periods corresponding to the test objects of the same type and application scenario as analysis periods, calculate the variance of the diagnosis accuracy rates of all the analysis periods to obtain a stability coefficient, and mark the diagnosis mode of the test object as static diagnosis or dynamic diagnosis through the stability coefficient;

[0013] Step 5: Conduct a fault diagnosis analysis on the bearing: Mark the bearing to be fault diagnosed as the diagnosis object, obtain the type and application scenario of the diagnosis object, obtain the diagnosis mode of the diagnosis object based on the type and application scenario of the diagnosis object, and conduct a fault diagnosis on the diagnosis object according to the obtained diagnosis mode.

[0014] Further, in Step 2, randomly select a diagnosis model at the start time of the first test period and mark it as the assigned model for the test object, and conduct a fault diagnosis on the test object using the assigned model during the test period. The diagnosis models include the empirical mode decomposition model, the hidden Markov model, wavelet analysis, and the long short-term memory network model; Conduct an effectiveness determination at the start time of any test period after the first test period: If the fault diagnosis effectiveness in the previous test period meets the requirements, continue to use the assigned model for fault diagnosis in the current test period; If the fault diagnosis effectiveness in the previous test period does not meet the requirements, randomly select a diagnosis model other than the assigned model as the new assigned model, and conduct a fault diagnosis on the test object using the new assigned model during the test period.

[0015] Further, in Step 3, obtain the diagnosis accuracy rate of the assigned model during the test period at the end time of the test period, and compare the diagnosis accuracy rate with a preset accuracy threshold: If the diagnosis accuracy rate is less than the accuracy threshold, it is determined that the fault diagnosis effectiveness during the test period does not meet the requirements; If the diagnosis accuracy rate is greater than or equal to the accuracy threshold, it is determined that the fault diagnosis effectiveness during the test period meets the requirements.

[0016] Furthermore, the specific process of marking the diagnostic mode of the test object as static diagnosis or dynamic diagnosis includes: comparing the smoothness coefficient with the preset smoothness threshold: if the smoothness coefficient is less than the smoothness threshold, the diagnostic mode of the corresponding test object is marked as static diagnosis, and the static applicable model is marked; if the smoothness coefficient is greater than or equal to the smoothness threshold, the diagnostic mode of the corresponding test object is marked as dynamic diagnosis, and the dynamic applicable model is marked.

[0017] Furthermore, the specific process of marking the static applicable model includes: summing up and averaging the diagnostic accuracy of the analysis period when a single diagnostic model is used as the allocation model for fault diagnosis to obtain the accurate performance value of the corresponding diagnostic model, and marking the diagnostic model with the largest accurate performance value as the static applicable model of the corresponding type and application scenario.

[0018] Furthermore, the specific process of marking the dynamically applicable model includes: arranging the analysis time periods in chronological order according to the test cycles to which they belong to obtain an analysis sequence, forming an analysis subset from the diagnostic accuracy of the analysis time periods with the same serial number in the analysis sequence, summing and averaging the corresponding elements in the analysis subset of the analysis time periods in which a single diagnostic model is used as the allocation model for fault diagnosis to obtain the diagnostic mean ZJ of the diagnostic model, performing variance calculation on the corresponding elements in the analysis subset of the analysis time periods in which a single diagnostic model is used as the allocation model for fault diagnosis to obtain the diagnostic discrete value ZL of the diagnostic model; obtaining the diagnostic priority coefficient ZY of the diagnostic model relative to the analysis subset by numerically calculating the diagnostic mean ZJ and the diagnostic discrete value ZL; marking the diagnostic model with the largest diagnostic priority coefficient ZY as the dynamically applicable model with the corresponding serial number in the analysis subset.

[0019] Furthermore, the specific process of using static diagnosis to perform fault diagnosis on the diagnosis object includes: obtaining a static applicable model of the diagnosis object, and using the static applicable model to perform full-process fault diagnosis on the diagnosis object.

[0020] Furthermore, the specific process of using dynamic diagnosis to perform fault diagnosis on the diagnosis object includes: generating a diagnosis cycle with the same duration as the test cycle for the diagnosis object, dividing the diagnosis cycle into a number of diagnosis time periods, the number of diagnosis time periods is equal to the number of test time periods, calling the dynamic applicable model corresponding to the diagnosis time period sequence number at the beginning of the diagnosis time period, and using the dynamic applicable model to perform fault diagnosis on the diagnosis object within the diagnosis time period.

[0021] The present invention has the following beneficial effects:

[0022] 1. Through the fault diagnosis test of the bearing, the diagnostic model of the test object is dynamically switched according to the diagnostic effectiveness of the test period, thereby providing sufficient diagnostic effectiveness data for bearings of different types and different application scenarios, and improving the accuracy of the results of pattern analysis;

[0023] 2. By performing a fault diagnosis effectiveness analysis at the end of the test period and marking the diagnostic mode of the test object according to the effectiveness analysis result, static diagnosis has low requirements for computing power but is only applicable to application scenarios where the impact of switching the diagnostic model on diagnostic effectiveness is small. Dynamic diagnosis can perform dynamic switching of the diagnostic model according to the phased effectiveness analysis result, but it has high requirements for computing power. Therefore, allocating a diagnostic mode for the test object can balance the accuracy of the diagnostic result and the rationality of computing power resource utilization from the application perspective;

[0024] 3. By performing a fault diagnosis analysis on the bearing, after allocating a diagnostic mode for the diagnostic object, the corresponding diagnostic mode can be used for fault diagnosis analysis. Bearings of different types and application scenarios can all perform high-accuracy fault diagnosis processing by matching the corresponding diagnostic algorithm with the static applicable model or the dynamic applicable model. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0026] Figure 1 It is the system block diagram of Embodiment 1 of the present invention;

[0027] Figure 2 It is the method flowchart of Embodiment 2 of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0028] The following will clearly and completely describe the technical solutions of the present invention in conjunction with the embodiments. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention.

[0029] Embodiment 1: As Figure 1 shown, a bearing fault diagnosis method based on signal analysis includes the following steps:

[0030] Step 1: Conduct a fault diagnosis test on the bearing: Mark the bearing to be tested as the test object, assign an application scenario to the test object, obtain the type of the test object, obtain the service life duration of the test object according to the bearing type and application scenario, generate a test cycle with a duration equal to the service life duration of the test object and divide the test cycle into several test periods, and put the test object into the application scenario for running tests; Dynamically switch the diagnostic model of the test object according to the diagnostic effectiveness of the test period, so as to provide sufficient diagnostic effectiveness data for bearings of different types and different application scenarios and improve the accuracy of the pattern analysis results;

[0031] Step 2: Allocate the diagnostic model at the start time of the test period: Randomly select a diagnostic model at the start time of the first test period and mark it as the allocated model of the test object, and use the allocated model to conduct a fault diagnosis on the test object during the test period. The diagnostic models include the empirical mode decomposition model, the hidden Markov model, wavelet analysis, and the long short-term memory network model;

[0032] The empirical mode decomposition (EMD) model is suitable for the analysis of non-stationary signals. The EMD method can decompose the vibration signal into multiple intrinsic mode functions (IMFs), extract the energy of important IMF components as feature quantities, and then use the BP neural network for pattern classification to achieve the diagnosis of bearing faults; The hidden Markov model (HMM) extracts the Mel-frequency cepstral coefficient (MFCC) features of the bearing audio signal, and uses two methods, the discrete HMM and the continuous Gaussian mixture density HMM, for modeling and diagnostic research. The HMM model has good anti-noise ability, is suitable for the analysis of audio signals, has a fast operation speed, and a high diagnostic accuracy; Wavelet analysis can effectively extract the characteristic components of bearing faults by selecting appropriate basis functions to decompose the signal. For example, the spectral kurtosis method based on wavelet packet transform separates the effective components of bearing damage characteristics by finding the optimal wavelet packet components; The long short-term memory network (LSTM) is a deep learning model suitable for processing time series data. By collecting vibration signal data related to rolling bearings, performing data preprocessing (such as removing noise, filtering, downsampling, etc.), and then using the LSTM model for training and testing, the diagnosis of bearing faults is achieved.

[0033] Perform effectiveness determination at the start time of any test period after the first test period: If the fault diagnosis effectiveness in the previous test period meets the requirements, continue to use the allocated model for fault diagnosis in the current test period; If the fault diagnosis effectiveness in the previous test period does not meet the requirements, randomly select a diagnostic model other than the allocated model as the new allocated model, and use the new allocated model to conduct a fault diagnosis on the test object during the test period;

[0034] Step 3: Conduct fault diagnosis effectiveness analysis at the end of the test period: Obtain the diagnostic accuracy rate of the allocation model during the test period at the end of the test period, and compare the diagnostic accuracy rate with a preset accuracy threshold. If the diagnostic accuracy rate is less than the accuracy threshold, it is determined that the fault diagnosis effectiveness during the test period does not meet the requirements. If the diagnostic accuracy rate is greater than or equal to the accuracy threshold, it is determined that the fault diagnosis effectiveness during the test period meets the requirements. Mark the diagnostic mode of the test object according to the effectiveness analysis result. Static diagnosis has low requirements for computing power, but is only applicable to application scenarios where the impact of switching the diagnostic model on diagnostic effectiveness is small. Dynamic diagnosis can perform dynamic switching of the diagnostic model according to the phased effectiveness analysis result, but it has high requirements for computing power. Therefore, allocating a diagnostic mode for the test object can balance the accuracy of the diagnostic result and the rationality of computing power resource utilization from an application perspective;

[0035] Step 4: Conduct diagnostic mode analysis at the end of the test cycle: Mark all test periods of test objects of the same type and application scenario as analysis periods. Calculate the variance of the diagnostic accuracy rates of all analysis periods to obtain a stability coefficient, and compare the stability coefficient with a preset stability threshold. If the stability coefficient is less than the stability threshold, mark the diagnostic mode of the corresponding test object as static diagnosis. Sum and average the diagnostic accuracy rates of the analysis periods using a single diagnostic model as the allocation model for fault diagnosis to obtain the accurate performance value of the corresponding diagnostic model. Mark the diagnostic model with the largest accurate performance value as the static applicable model for the corresponding type and application scenario;

[0036] If the stability coefficient is greater than or equal to the stability threshold, mark the diagnostic mode of the corresponding test object as dynamic diagnosis. Arrange the analysis periods in the time order of the test cycles to which they belong to obtain an analysis sequence. The diagnostic accuracy rates of the analysis periods with the same serial number in the analysis sequence form an analysis subset. Sum and average the corresponding elements of the analysis periods using a single diagnostic model as the allocation model for fault diagnosis in the analysis subset to obtain the diagnostic mean value ZJ of the diagnostic model. Calculate the variance of the corresponding elements of the analysis periods using a single diagnostic model as the allocation model for fault diagnosis in the analysis subset to obtain the diagnostic discrete value ZL of the diagnostic model. Obtain the diagnostic priority coefficient ZY of the diagnostic model relative to the analysis subset through the formula ZY = k1×ZJ - k2×ZL, where k1 and k2 are both proportionality coefficients, and k1 > k2 > 1. Mark the diagnostic model with the largest diagnostic priority coefficient ZY as the dynamic applicable model for the corresponding serial number of the analysis subset. After allocating the diagnostic mode for the diagnostic object, the corresponding diagnostic mode can be used for fault diagnosis analysis. Bearings of different types and application scenarios can perform high-accuracy fault diagnosis processing according to the static applicable model or dynamic applicable model in combination with the corresponding diagnostic algorithm;

[0037] Step 5: Conduct fault diagnosis and analysis on the bearing: Mark the bearing to be fault-diagnosed as the diagnosis object, obtain the type and application scenario of the diagnosis object, obtain the diagnosis mode of the diagnosis object according to the type and application scenario of the diagnosis object. The specific process of using static diagnosis to conduct fault diagnosis on the diagnosis object includes: obtaining the static applicable model of the diagnosis object, and using the static applicable model to conduct full-process fault diagnosis on the diagnosis object; The specific process of using dynamic diagnosis to conduct fault diagnosis on the diagnosis object includes: generating a diagnosis cycle with the same duration as the test cycle for the diagnosis object, dividing the diagnosis cycle into several diagnosis time periods, the number of diagnosis time periods is equal to the number of test time periods, calling the dynamic applicable model corresponding to the diagnosis time period serial number at the start moment of the diagnosis time period, and using the dynamic applicable model to conduct fault diagnosis on the diagnosis object within the diagnosis time period.

[0038] Example 2: As Figure 2 shown, the bearing fault diagnosis system based on signal analysis includes a diagnosis test module, a model allocation module, an effective analysis module, a mode analysis module, and a diagnosis processing module. The diagnosis test module, the model allocation module, the effective analysis module, the mode analysis module, and the diagnosis processing module are sequentially connected for communication.

[0039] The diagnosis test module is used to conduct fault diagnosis tests on the bearing and generate a test cycle and test time periods for it;

[0040] The model allocation module is used to conduct diagnosis model allocation at the start moment of the test time period;

[0041] The effective analysis module is used to conduct fault diagnosis effectiveness analysis at the end moment of the test time period;

[0042] The mode analysis module is used to conduct diagnosis mode analysis at the end moment of the test cycle;

[0043] The diagnosis processing module is used to conduct fault diagnosis and analysis on the bearing using static diagnosis or dynamic diagnosis.

[0044] Bearing fault diagnosis method based on signal analysis. During operation, the bearing to be tested is marked as the test object, an application scenario is assigned to the test object, a test cycle with a duration equal to the service life of the test object is generated and the test cycle is divided into several test periods. The test object is put into the application scenario for operation test. At the start moment of the test period, a diagnosis model is assigned. At the end moment of the test period, the diagnosis accuracy rate of the assigned model within the test period is obtained. Whether the fault diagnosis effectiveness within the test period meets the requirements is determined according to the diagnosis accuracy rate. At the end moment of the test cycle, a diagnosis mode analysis is carried out and the diagnosis mode of the test object is marked as static diagnosis or dynamic diagnosis. The bearing to be subjected to fault diagnosis is marked as the diagnosis object, the type and application scenario of the diagnosis object are obtained, the diagnosis mode of the diagnosis object is obtained according to the type and application scenario of the diagnosis object, and fault diagnosis analysis is carried out on the diagnosis object according to the obtained diagnosis mode.

[0045] The above content is only an example and illustration of the structure of the present invention. Those skilled in the art of this technology can make various modifications or supplements to the described specific embodiments or use similar methods to replace them. As long as they do not deviate from the structure of the invention or exceed the scope defined by this claim book, they should all fall within the protection scope of the present invention.

[0046] The above formulas are all obtained by collecting a large amount of data for software simulation and selecting a formula close to the true value. The coefficients in the formula are set by those skilled in the art according to the actual situation. For example: the formula ZY = k1×ZJ - k2×ZL; those skilled in the art collect multiple groups of sample data and set corresponding diagnosis priority coefficients for each group of sample data; substitute the set diagnosis priority coefficients and the collected sample data into the formula, and any two formulas form a binary linear equation system. Screen the calculated coefficients and take the average value to obtain the values of k1 and k2 as 2.83 and 2.05 respectively;

[0047] The magnitude of the coefficient is a specific value obtained by quantifying each parameter, which is convenient for subsequent comparison. Regarding the magnitude of the coefficient, it depends on the amount of sample data and the diagnosis priority coefficients initially set by those skilled in the art for each group of sample data; as long as it does not affect the proportional relationship between the parameter and the quantified value, for example, the diagnosis priority coefficient is proportional to the numerical value of the diagnosis mean.

[0048] In the description of this specification, the descriptions referring to the terms "one embodiment", "example", "specific example", etc. mean that the specific features, structures, materials or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner.

[0049] The preferred embodiments of the present invention disclosed above are only used to help explain the present invention. The preferred embodiments do not describe all the details in detail, nor do they limit the present invention to only the specific implementation manners. Obviously, according to the content of this specification, many modifications and variations can be made. This specification selects and specifically describes these embodiments in order to better explain the principle and practical application of the present invention, so that those skilled in the art can well understand and utilize the present invention. The present invention is only limited by the claims and their full scope and equivalents.

Claims

1. A bearing fault diagnosis method based on signal analysis, characterized in that It includes the following steps: Step 1: Conduct a fault diagnosis test on the bearing: Mark the bearing to be tested as the test object, assign an application scenario to the test object, obtain the type of the test object, obtain the service life duration of the test object according to the bearing type and application scenario, generate a test cycle with a duration equal to the service life duration of the test object and divide the test cycle into several test periods, and put the test object into the application scenario for running tests; Step 2: Assign a diagnostic model at the start time of the test period; Step 3: Conduct an analysis of the effectiveness of fault diagnosis at the end time of the test period; Step 4: Conduct a diagnostic mode analysis at the end time of the test cycle: Mark all test periods corresponding to test objects of the same type and application scenario as analysis periods, calculate the variance of the diagnostic accuracy rates of all analysis periods to obtain a stability coefficient, and mark the diagnostic mode of the test object as static diagnosis or dynamic diagnosis through the stability coefficient; Step 5: Conduct a fault diagnosis analysis on the bearing: Mark the bearing to be fault diagnosed as the diagnosis object, obtain the type and application scenario of the diagnosis object, obtain the diagnostic mode of the diagnosis object according to the type and application scenario of the diagnosis object, and conduct a fault diagnosis on the diagnosis object according to the obtained diagnostic mode; Effectiveness determination is carried out at the start time of any test period after the first test period: If the fault diagnosis effectiveness in the previous test period meets the requirements, the current test period continues to use the assigned model for fault diagnosis; If the fault diagnosis effectiveness in the previous test period does not meet the requirements, randomly select a diagnostic model outside the assigned model as the new assigned model, and use the new assigned model to conduct a fault diagnosis on the test object within the test period; In Step 3, obtain the diagnostic accuracy rate of the assigned model within the test period at the end time of the test period, and compare the diagnostic accuracy rate with a preset accuracy threshold: If the diagnostic accuracy rate is less than the accuracy threshold, it is determined that the fault diagnosis effectiveness within the test period does not meet the requirements; If the diagnostic accuracy rate is greater than or equal to the accuracy threshold, it is determined that the fault diagnosis effectiveness within the test period meets the requirements.

2. The bearing fault diagnosis method based on signal analysis according to claim 1, wherein In Step 2, randomly select a diagnostic model and mark it as the assigned model of the test object at the start time of the first test period, and use the assigned model to conduct a fault diagnosis on the test object within the test period. The diagnostic models include empirical mode decomposition model, hidden Markov model, wavelet analysis, and long short-term memory network model; Effectiveness determination is carried out at the start time of any test period after the first test period: If the fault diagnosis effectiveness in the previous test period meets the requirements, the current test period continues to use the assigned model for fault diagnosis; If the fault diagnosis effectiveness in the previous test period does not meet the requirements, randomly select a diagnostic model outside the assigned model as the new assigned model, and use the new assigned model to conduct a fault diagnosis on the test object within the test period.

3. The bearing fault diagnosis method based on signal analysis according to claim 2, wherein The specific process of marking the diagnostic mode of the test object as static diagnosis or dynamic diagnosis includes: comparing the smoothness coefficient with the preset smoothness threshold: if the smoothness coefficient is less than the smoothness threshold, the diagnostic mode of the corresponding test object is marked as static diagnosis, and the static applicable model is marked; if the smoothness coefficient is greater than or equal to the smoothness threshold, the diagnostic mode of the corresponding test object is marked as dynamic diagnosis, and the dynamic applicable model is marked.

4. The bearing fault diagnosis method based on signal analysis according to claim 3, wherein The specific process of marking the static applicable model includes: summing up and averaging the diagnostic accuracy of the single diagnostic model as the allocation model for fault diagnosis during the analysis period to obtain the accurate performance value of the corresponding diagnostic model, and marking the diagnostic model with the largest accurate performance value as the static applicable model of the corresponding type and application scenario.

5. The bearing fault diagnosis method based on signal analysis according to claim 3, wherein The specific process of marking the dynamically applicable model includes: arranging the analysis time periods in the time sequence of the test cycles to which they belong to obtain an analysis sequence, forming an analysis subset by the diagnostic accuracy of the analysis time periods with the same serial number in the analysis sequence, summing and averaging the corresponding elements in the analysis subset for the analysis time periods in which a single diagnostic model is used as an allocation model for fault diagnosis to obtain the diagnostic mean ZJ of the diagnostic model, performing variance calculation on the corresponding elements in the analysis subset for the analysis time periods in which a single diagnostic model is used as an allocation model for fault diagnosis to obtain the diagnostic discrete value ZL of the diagnostic model; obtaining the diagnostic priority coefficient ZY of the diagnostic model relative to the analysis subset by numerically calculating the diagnostic mean ZJ and the diagnostic discrete value ZL; marking the diagnostic model with the largest diagnostic priority coefficient ZY as the dynamically applicable model with the corresponding serial number in the analysis subset.

6. The bearing fault diagnosis method based on signal analysis according to claim 4, wherein The specific process of using static diagnosis to perform fault diagnosis on the diagnosis object includes: obtaining a static applicable model of the diagnosis object, and using the static applicable model to perform full-process fault diagnosis on the diagnosis object.

7. The bearing fault diagnosis method based on signal analysis according to claim 5, characterized in that The specific process of using dynamic diagnosis to perform fault diagnosis on the diagnosis object includes: generating a diagnosis cycle with the same duration as the test cycle for the diagnosis object, dividing the diagnosis cycle into a number of diagnosis time periods, the number of diagnosis time periods is equal to the number of test time periods, calling the dynamic applicable model corresponding to the diagnosis time period sequence number at the beginning of the diagnosis time period, and using the dynamic applicable model to perform fault diagnosis on the diagnosis object within the diagnosis time period.

Images