Mechanical fault diagnosis method based on digital-analog linkage

Abstract

The invention provides a mechanical fault diagnosis method based on digital-analog linkage. The method comprises the following steps: S1, establishing a model; s2, adding model constraints; s3, acquiring operation data; s4, simulating mechanical operation; s5, carrying out mechanical calibration; s6, performing actual operation of the machine; s7, recording operation data; s8, mechanical maintenance; according to the method, the digital diagnosis model and the three-dimensional mechanical model are constructed to be synchronized with the actual machine, the machine operation simulation is performed before diagnosis, and the machine and the constructed model are synchronized in the operation process, so that the troubleshooting of fault reasons is facilitated, corresponding constraints are added to the three-dimensional mechanical model, and corresponding effects are added to the three-dimensional mechanical model; the mechanical operation abnormal parts can be quickly analyzed through the three-dimensional mechanical model; according to the method, the digital diagnosis model and the three-dimensional mechanical model at different stages are subjected to file recording, so that a corresponding data model is conveniently and quickly called for troubleshooting and analysis when a machine generates related abnormity while deep learning of a feeding intelligent algorithm is facilitated.

Description

Technical Field

[0001] The invention relates to a fault diagnosis method, in particular to a mechanical fault diagnosis method based on digital-analog linkage, and belongs to the technical field of fault diagnosis. Background Art

[0002] Mechanical failure is defined as the failure of a sub-system at a given level that "loses the specified function", or the output of a sub-system at a given level does not match the expected output. In addition, from the perspective of equipment maintenance, failure is defined as: malfunction of the equipment operation, or system or partial failure of the equipment. From the perspective of the diagnostic object, failure can be considered as a contradiction between the observed value of the system and the predicted value obtained from the system's behavioral model. From the perspective of state recognition, the failure of the equipment is defined as its abnormal state. Some experts believe that equipment failure is a phenomenon in which the equipment becomes abnormal during operation and cannot meet the predetermined performance requirements, or the parameters characterizing its working performance exceed a certain specified limit, which may cause the equipment to lose part or all of its functions. A Chinese patent application (CN202410939862.4) discloses a mechanical fault diagnosis method and device based on digital-analog linkage. The method classifies the multi-dimensional features of each sample noise-reduced vibration data based on the classification module of the fault diagnosis model to obtain the fault diagnosis results of the corresponding sample noise-reduced vibration data, and updates the parameters of the fault diagnosis model based on the fault diagnosis results and fault labels of each sample noise-reduced vibration data to obtain a trained fault diagnosis model, so that the vibration signal of the equipment to be diagnosed can be diagnosed based on the trained fault diagnosis model. However, when diagnosing mechanical faults, the existing diagnostic methods are not convenient for diagnosing the wear and aging of mechanical transmission parts, and the service life of internal components cannot be determined by manual judgment alone; In order to solve the above technical problems, a mechanical fault diagnosis method based on digital-analog linkage is proposed. Summary of the invention

[0003] In view of this, the present invention provides a mechanical fault diagnosis method based on digital-analog linkage to solve or alleviate the technical problems existing in the prior art and at least provide a beneficial option.

[0004] The technical solution of the present invention is implemented as follows: a mechanical fault diagnosis method based on digital-analog linkage comprises the following steps: S1. Build the model; S2, add model constraints; S3, obtaining operation data; S4, simulate mechanical operation; S5, Mechanical calibration; S6. Actual operation of the machinery; S7, operation data recording; S8. Mechanical maintenance.

[0005] Further preferably, in S1, when establishing a model for mechanical fault diagnosis, a digital diagnostic model and a three-dimensional mechanical model are established respectively, the digital diagnostic model completes data analysis and collation, and the three-dimensional mechanical model completes mechanical operation simulation, and the size of the three-dimensional mechanical model is the same as the actual mechanical size or is proportionally enlarged or reduced.

[0006] Further preferably, in S2, when adding constraints to the three-dimensional mechanical model after the model is established, it is necessary to add corresponding material feature constraints for each simulated component, including physical properties, chemical properties, mechanical properties and other properties such as flexibility, hardness, processing performance, electrical insulation, heat resistance and fatigue resistance.

[0007] Further preferably, in S3, the operation data obtained are the required parameters of the actual operation of the machine, numbers are assigned to the obtained operation data and recorded in groups, and multiple groups of data are obtained to meet multiple simulation requirements.

[0008] Further preferably, in said S4, when simulating the operation of the machinery, the remaining life of each mechanical component and the operation abnormality prediction are performed through the digital diagnosis model, and the three-dimensional mechanical model performs actual operation simulation based on the prediction results of the digital diagnosis model.

[0009] Further preferably, in S5, the machine is calibrated to a desired state, and the machine is associated with the three-dimensional machine model to synchronize the data of the two.

[0010] Further preferably, in S6, the three-dimensional mechanical model operates synchronously during the operation of the machine, and the machine and the three-dimensional mechanical model are simultaneously monitored in real time during the operation of the machine and the operation data is recorded.

[0011] Further preferably, in S7, when recording the operation data after operation, the data of the digital diagnostic model and the three-dimensional mechanical model are archived and recorded together, and the digital diagnostic model is fed into the intelligent algorithm to realize mechanical deep learning.

[0012] Further preferably, in said S8, after the machine maintenance, the data of the associated digital diagnosis model and the three-dimensional machine model are updated to keep the digital diagnosis model and the three-dimensional machine model consistent with the actual machine data.

[0013] The embodiment of the present invention has the following advantages due to the adoption of the above technical solution: First, the present invention constructs a digital diagnosis model and a three-dimensional mechanical model that is synchronized with the actual machine, performs mechanical operation simulation before diagnosis, and synchronizes the machine with the constructed model during operation, so as to facilitate troubleshooting of the cause of the fault, and the troubleshooting is more comprehensive. The three-dimensional mechanical model is added with corresponding constraints and effects, so as to facilitate rapid analysis of abnormal parts of the mechanical operation through the three-dimensional mechanical model; 2. The present invention records digital diagnostic models and three-dimensional mechanical models at different stages, which facilitates the feeding of intelligent algorithms for deep learning and the rapid calling of corresponding data models for troubleshooting and analysis when relevant abnormalities occur in the machinery.

[0014] The above summary is for illustrative purposes only and is not intended to be limiting in any way. In addition to the illustrative aspects, embodiments and features described above, further aspects, embodiments and features of the present invention will be readily apparent by reference to the accompanying drawings and the following detailed description. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0016] Figure 1 The present invention is a flow chart of the method. DETAILED DESCRIPTION

[0017] In the following, only some exemplary embodiments are briefly described. As those skilled in the art will appreciate, the described embodiments may be modified in various ways without departing from the spirit or scope of the present invention. Therefore, the drawings and descriptions are considered to be exemplary and non-restrictive in nature.

[0018] The embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0019] like Figure 1 As shown, an embodiment of the present invention provides a mechanical fault diagnosis method based on digital-analog linkage, comprising the following steps: S1. Build the model; S2, add model constraints; S3, obtaining operation data; S4, simulate mechanical operation; S5, Mechanical calibration; S6. Actual operation of the machinery; S7, operation data recording; S8. Mechanical maintenance.

[0020] In one embodiment, in S1, when establishing a model for mechanical fault diagnosis, a digital diagnosis model and a three-dimensional mechanical model are established respectively, the digital diagnosis model completes data analysis and collation, and the three-dimensional mechanical model completes mechanical operation simulation, the size of the three-dimensional mechanical model is the same as the actual mechanical size or is proportionally enlarged and reduced, the digital diagnosis model calculates the probability contribution value of each data point in the operation data through the kernel density function, and determines the redundant residual threshold interval according to the probability contribution value, and finally realizes accurate diagnosis of the fault; Taking the degradation characteristic with a growth trend as an example, the linear Wiener process modeling of the evolution of the composite health index over time is considered. The model is described as

[0021] in, For the Devices in The initial degradation at time , is the drift coefficient, reflecting the The degradation rate of a device, For the The diffusion coefficient of each device characterizes the random uncertainty of the degradation process. The standard Brownian motion reflects the time-varying randomness of the degradation process; Based on the degradation process established above, through the concept of first arrival time, the life of the equipment can be defined as

[0022] in, is the failure threshold corresponding to the composite health indicator of the fused multi-source sensors; For the random degradation process (2) and the definition of the first arrival time lifetime (3), the lifetime It obeys the inverse Gaussian distribution, and its probability density function, mathematical expectation and variance are

[0023]

[0024]

[0025] To predict the equipment life based on the above results, we first need to determine the model parameters. And the failure threshold vv, for this reason, the composite health indicator dataset constructed based on multi-source sensor data is recorded The incremental data set is ,in, ; Based on the standard Brownian motion, the increments are independent, Gaussian, and Markov-like, and the composite health index increment data set is obtained The likelihood function is

[0026] in, is the unknown parameter set of the established degradation model; According to the maximum likelihood estimation method, Respectively After the derivative is equal to zero, we can get The maximum likelihood estimate of is

[0027]

[0028]

[0029] in, , , ,

[0030] According to the above parameter estimation results, The predicted value (point estimate) of the lifetime of a randomly degraded device can be expressed as

[0031]

[0032] , Among them, the failure threshold and fusion coefficient Still to be determined.

[0033] In one embodiment, in S2, when adding constraints to the three-dimensional mechanical model after the model is established, it is necessary to add corresponding material feature constraints for each simulated component, including physical properties, chemical properties, mechanical properties and other properties such as flexibility, hardness, processing performance, electrical insulation, heat resistance and fatigue resistance.

[0034] In one embodiment, in S3, the operation data obtained are the required parameters when the machine is actually running, numbers are assigned to the obtained operation data and recorded in groups, and multiple groups of data are obtained to meet multiple simulation requirements.

[0035] In one embodiment, in S4, when simulating the operation of the machine, the remaining life of each machine component and the operation abnormality prediction are predicted through the digital diagnosis model, and the three-dimensional machine model performs actual operation simulation based on the prediction results of the digital diagnosis model.

[0036] In one embodiment, in S5, the machine is calibrated to an expected state, and the machine is associated with a three-dimensional machine model to synchronize data of the two. After calibration, a multi-source sensor is configured for the machine, and the remaining life of the randomly degraded equipment is predicted under the monitoring of the multi-source sensor. make For the Randomly degraded devices The sensors in Performance degradation monitoring data collected at all times, is the number of random degradation devices that need to be monitored. Assume that there are a total of According to the multi-source sensor data, the composite health index after fusing the multi-sensor monitoring data is expressed as

[0037] in, Indicates The fusion coefficient of a sensor measures the proportion of the sensor in the data fusion process. Assuming that The corresponding monitoring time of a randomly degraded device is , then the observation data corresponding to the composite health index is ,in, For the The number of monitoring data of randomly degraded devices is , without loss of generality, only the case where the monitoring time interval of randomly degraded devices is equal interval is considered, that is, .

[0038] In one embodiment, in S6, the three-dimensional mechanical model is synchronously operated during the operation of the machine, and the machine and the three-dimensional mechanical model are simultaneously monitored in real time during the operation of the machine and the operation data is recorded.

[0039] In one embodiment, in S7, when recording the operation data after operation, the data of the digital diagnostic model and the three-dimensional mechanical model are archived and recorded together, and the digital diagnostic model is fed into the intelligent algorithm to realize mechanical deep learning.

[0040] In one embodiment, in S8, after the machine maintenance, the associated digital diagnosis model and the three-dimensional machine model are updated to keep the digital diagnosis model and the three-dimensional machine model consistent with the actual machine data.

[0041] When the present invention is working: a separate digital diagnosis model and a three-dimensional mechanical model are established for the machine, corresponding constraints are added to the three-dimensional mechanical model, the actual machine operation is simulated through the three-dimensional mechanical model, and data is fed to the digital diagnosis model for deep learning, the digital diagnosis model and the three-dimensional mechanical model are fed with actual operation parameters before the machine runs, data prediction is performed through the digital diagnosis model and operation simulation is performed through the three-dimensional mechanical model, the actual operation is calibrated according to the model prediction results, and the simulated operation anomalies are analyzed and the abnormal parts of the machine are inferred, the machine is synchronized with the model when the machine is actually operated, the machine and the model are monitored in real time during the operation, the operation anomaly information is obtained as soon as possible, the operation data is archived and recorded after the operation, and the model data is updated after the machine maintenance.

[0042] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of various changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A mechanical fault diagnosis method based on digital-analog linkage, characterized in that: The following steps are involved: S1. Build the model; S2, add model constraints; S3, obtaining operation data; S4, simulate mechanical operation; S5, Mechanical calibration; S6. Actual operation of the machinery; S7, operation data recording; S8. Mechanical maintenance.

2. A mechanical fault diagnosis method based on digital-analog linkage according to claim 1, characterized in that: In S1, when establishing a model for mechanical fault diagnosis, a digital diagnosis model and a three-dimensional mechanical model are established respectively, the digital diagnosis model completes data analysis and collation, and the three-dimensional mechanical model completes mechanical operation simulation, and the size of the three-dimensional mechanical model is the same as the actual mechanical size or is proportionally enlarged or reduced.

3. A mechanical fault diagnosis method based on digital-analog linkage according to claim 1, characterized in that: In S2, when adding constraints to the three-dimensional mechanical model after the model is established, it is necessary to add corresponding material feature constraints for each simulated component, including physical properties, chemical properties, mechanical properties and other properties such as flexibility, hardness, processing performance, electrical insulation, heat resistance and fatigue resistance.

4. A mechanical fault diagnosis method based on digital-analog linkage according to claim 1, characterized in that: In S3, the operation data obtained are the required parameters of the actual operation of the machine. Numbers are assigned to the obtained operation data and recorded in groups, and multiple groups of data are obtained to meet multiple simulation requirements.

5. The mechanical fault diagnosis method based on digital-analog linkage according to claim 1 is characterized in that: In S4, when simulating the operation of the machine, the remaining life of each machine component and the operation abnormality prediction are predicted through the digital diagnosis model, and the three-dimensional machine model performs actual operation simulation based on the prediction results of the digital diagnosis model.

6. The mechanical fault diagnosis method based on digital-analog linkage according to claim 1 is characterized in that: In S5 , the machine is calibrated to the expected state, and the machine is associated with the three-dimensional machine model to synchronize the data of the two.

7. The mechanical fault diagnosis method based on digital-analog linkage according to claim 1 is characterized in that: In S6, the three-dimensional mechanical model operates synchronously during the operation of the machine, and the machine and the three-dimensional mechanical model are simultaneously monitored in real time during the operation of the machine, and the operation data are recorded.

8. The mechanical fault diagnosis method based on digital-analog linkage according to claim 1 is characterized in that: In the S7, when recording the operation data after operation, the data of the digital diagnosis model and the three-dimensional mechanical model are archived and recorded together, and the digital diagnosis model is fed into the intelligent algorithm to realize mechanical deep learning.

9. The mechanical fault diagnosis method based on digital-analog linkage according to claim 1 is characterized in that: In S8, after the machine maintenance, the data of the associated digital diagnosis model and the three-dimensional machine model are updated to keep the digital diagnosis model and the three-dimensional machine model consistent with the actual machine data.

Images