Industrial equipment health state analysis method

Abstract

The invention discloses an industrial equipment health state analysis method, and relates to the technical field of industrial equipment health monitoring, and the method comprises the steps: obtaining vibration data, lubricating oil pollution image data and real-time motor current data of an automobile factory punching machine; generating a dynamic weight coefficient according to the ratio of the real-time motor current data to the rated current of the equipment; performing weighted summation to calculate a real-time health score; constructing a wear-fatigue coupling model to calculate and output an alarm prediction L value; and when the alarm prediction L value exceeds a set threshold value, generating an equipment maintenance instruction and adjusting equipment operation parameters. According to the method, through multi-physics field coupling modeling, dynamic weight distribution and closed-loop self-optimization mechanisms, the precision and real-time performance problems of equipment health state analysis in an automobile manufacturing scene are solved, the method is particularly suitable for high-load equipment such as a punching machine tool and a welding robot, and compared with a traditional method, the comprehensive operation and maintenance cost is reduced, and the non-planed downtime is shortened.

Description

Technical Field

[0001] The present invention relates to the technical field of industrial equipment health monitoring, and in particular to a method for analyzing the health status of industrial equipment. Background Art

[0002] The health management of core equipment such as stamping machines and welding robots in automobile manufacturing plants has become the key to ensuring production efficiency and product quality. The existing technologies mainly rely on single-type sensor data (such as vibration, current) or multi-source data fusion with fixed weights, and realize state evaluation through threshold alarm or life prediction models. However, in the automobile manufacturing scenario of high-speed continuous production, the health status of equipment is affected by the coupling of multiple factors such as mechanical wear, lubricating oil pollution, and electrical load. The traditional analysis methods have significant defects:

[0003] For example, in the high-load working condition of a stamping machine, the lubricating oil pollution particles and mechanical vibration are coupled with each other to accelerate the bearing wear. However, the existing technologies independently analyze the oil pollution image and vibration data, resulting in a deviation in health assessment. Moreover, the synergistic effect of the lubricating oil pollution coverage rate and vibration shock will exponentially accelerate the bearing fatigue. However, the traditional model still simply adds the "pollution rate + fatigue crack growth rate" to calculate the life, and the prediction error is large.

[0004] Therefore, there is an urgent need for a method for analyzing the health status of industrial equipment to solve the above technical problems. Summary of the Invention

[0005] The purpose of this part is to outline some aspects of the embodiments of the present invention and briefly introduce some preferred embodiments. Some simplifications or omissions may be made in this part, as well as in the abstract and title of the present application, to avoid obscuring the purpose of this part, the abstract, and the title. However, such simplifications or omissions cannot be used to limit the scope of the present invention.

[0006] In view of the problems existing in the above-mentioned prior art, the present invention is proposed.

[0007] To solve the above technical problems, the present invention provides the following technical solution: A method for analyzing the health status of industrial equipment, including:

[0008] First, obtain the vibration data, lubricating oil pollution image data, and real-time motor current data of the stamping machine in the automobile factory; perform harmonic decomposition on the vibration data to calculate the harmonic distortion rate, perform edge detection on the lubricating oil pollution image data to calculate the oil pollution coverage rate, and perform normalization processing on the real-time motor current data;

[0009] Then, generate a dynamic weight coefficient according to the ratio of the real-time motor current data to the rated current of the equipment;

[0010] Combined with the harmonic distortion rate, oil pollution coverage rate, and dynamic weight coefficient, calculate the real-time health score H1 through weighted summation;

[0011] Based on the real-time health score H1 and the time-domain integral of the oil pollution coverage rate, construct a wear-fatigue coupling model to calculate and output the alarm prediction L value. Finally, when the remaining life L value exceeds the set threshold, generate a device maintenance instruction and adjust the device operation parameters.

[0012] As a preferred solution of the industrial equipment health status analysis method described in the present invention, wherein: the vibration data, lubricating oil pollution image data, and real-time motor current data are synchronously processed through a spatio-temporal alignment model;

[0013] The spatio-temporal alignment model matches the data streams of different sensors in terms of time stamps and device space coordinates by identifying the mutation points in the current data.

[0014] As a preferred solution of the industrial equipment health status analysis method described in the present invention, wherein: the harmonic distortion rate is the amplitude ratio of the fundamental wave to the third harmonic of the vibration signal, and the oil pollution coverage rate is the percentage of the pixel area of the particulate matter region in the oil to the total pixel area of the device monitoring region.

[0015] As a preferred solution of the industrial equipment health status analysis method described in the present invention, wherein: the calculation formula of the dynamic weight coefficient is:

[0016] α = current current / device design limit current;

[0017] Among them, the value range of the dynamic weight coefficient is [0 - 1]. When the real-time current exceeds the rated current, α is forcibly set to 1.

[0018] As a preferred solution of the industrial equipment health status analysis method described in the present invention, wherein: the calculation formula of the wear-fatigue coupling model is:

[0019] L = ∫H1(t)·C(t)·Ddt / δ;

[0020] Among them, H1(t) represents the health score at time t, and D represents the material damage rate coefficient;

[0021] H1(t) = α·(1 - THD) + (1 - α)·(1 - C(t));

[0022] THD represents the harmonic distortion rate, C(t) represents the oil pollution coverage rate at time t; δ represents the material fatigue limit threshold. When L > 1, an alarm is triggered. And when the real-time current exceeds 90% of the rated current, an acceleration aging coefficient β is introduced in the integral calculation, and β > 1.

[0023] As a preferred solution of the industrial equipment health status analysis method described in the present invention, wherein: adjusting the equipment operating parameters includes: dynamically reducing the stamping frequency of the stamping machine to 70%-90% of the rated value according to the mapping relationship between the L value and the remaining life of the equipment; specifically including:

[0024] The first threshold ≤ L < the second threshold; the stamping frequency of the equipment is reduced to 90% of the rated value;

[0025] The second threshold ≤ L < the third threshold; the stamping frequency of the equipment is reduced to 80% of the rated value;

[0026] L ≥ the third threshold; the stamping frequency of the equipment is reduced to 70% of the rated value;

[0027] And associate the maintenance instruction with the real-time health score H1 and store it in the maintenance log database.

[0028] As a preferred solution of the industrial equipment health status analysis method described in the present invention, wherein: after each equipment maintenance, compare the actual remaining life data with the predicted life corresponding to the predicted L value, and update the integral coefficient of the wear-fatigue coupling model through a Bayesian network; specifically, update the material fatigue limit threshold δ;

[0029] That is, when the actual life is greater than the predicted life, then increase the threshold δ.

[0030] As a preferred solution of the industrial equipment health status analysis method described in the present invention, wherein: the vibration data is collected by a high-frequency vibration sensor, the sampling frequency is 10 kHz and the frequency band is limited to 0.1 - 500 Hz;

[0031] The lubricating oil contamination image data is taken by an industrial microscope camera, and the shooting interval is 5 minutes;

[0032] The real-time motor current data is collected by a high-precision current sensor, the range is 0 - 500 A and the accuracy is ±0.5%.

[0033] The present invention also discloses a computer device, including a memory and a processor, the memory stores a computer program, and when the processor executes the computer program, it implements the steps of the above-mentioned industrial equipment health status analysis method.

[0034] The present invention also discloses a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, it implements the steps of the above-mentioned industrial equipment health status analysis method.

[0035] The beneficial effects of the present invention:

[0036] 1. The present invention adjusts the weights of vibration and oil contamination data in real time through motor current data, solves the evaluation deviation caused by traditional fixed weights, and at the same time constructs an integral model to quantify the synergistic effects of health scores, contamination coverage rates, and damage rates, greatly reducing the life prediction error. An acceleration aging coefficient triggered by current is introduced to improve the prediction accuracy under high-load conditions, and the fatigue limit threshold of materials is dynamically updated based on a Bayesian network to achieve adaptive calibration of model parameters.

[0037] 2. The present invention solves the problems of accuracy and real-time performance in analyzing the health status of equipment in the automotive manufacturing scenario through multi-physics field coupling modeling, dynamic weight allocation, and a closed-loop self-optimization mechanism. It is particularly applicable to high-load equipment such as stamping machines and welding robots, reducing the comprehensive operation and maintenance costs and the unplanned downtime compared with traditional methods. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] To more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts. Among them:

[0039] Figure 1 is a schematic diagram of the overall process of a method for analyzing the health status of industrial equipment proposed by the present invention;

[0040] Figure 2 is a schematic diagram of the principle framework of a method for analyzing the health status of industrial equipment proposed by the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0041] To make the above objects, features, and advantages of the present invention more obvious and understandable, the following will provide a detailed description of the specific embodiments of the present invention with reference to the accompanying drawings of the specification.

[0042] Many specific details are set forth in the following description to facilitate a thorough understanding of the present invention. However, the present invention can also be implemented in other ways different from those described herein. Those skilled in the art can make similar extensions without departing from the connotation of the present invention. Therefore, the present invention is not limited by the specific embodiments disclosed below.

[0043] Secondly, the so-called "one embodiment" or "embodiment" herein refers to a specific feature, structure, or characteristic that can be included in at least one implementation of the present invention. The appearances of "in one embodiment" in different places in this specification do not all refer to the same embodiment, nor are they separate or selectively exclusive embodiments from other embodiments.

[0044] Embodiment 1

[0045] Reference Figure 1-2 , for an embodiment of the present invention, a method for analyzing the health status of industrial equipment is provided. This method includes the following steps:

[0046] Step 1: Obtain the vibration data, lubricating oil pollution image data, and real-time motor current data of the stamping machine tool in the automobile factory; perform harmonic decomposition on the vibration data to calculate the harmonic distortion rate, perform edge detection on the lubricating oil pollution image data to calculate the oil pollution coverage rate, and perform normalization processing on the real-time motor current data.

[0047] By collecting multi-source data from the stamping machine tool, it provides the basic input for subsequent health status analysis. The vibration data reflects the mechanical structure state, the lubricating oil pollution image characterizes the concentration of wear particles, and the motor current data monitors the equipment load. Through the above data, mechanical wear and electrical load factors are comprehensively covered.

[0048] The vibration data, lubricating oil pollution image data, and real-time motor current data are synchronously processed through a spatio-temporal alignment model; the spatio-temporal alignment model matches the data streams of different sensors in terms of time stamps and equipment spatial coordinates by identifying the mutation points in the current data. The spatio-temporal alignment model synchronizes the data based on current mutations (such as motor startup / overload) to solve the data delay caused by changes in the production rhythm. For example, when the current suddenly increases from 100A to 450A, the system aligns the time stamps of the vibration and oil data. Thus, the accuracy of multi-modal data fusion is improved.

[0049] The vibration data is collected by a high-frequency vibration sensor with a sampling frequency of 10kHz and a frequency band limited to 0.1 - 500Hz; the lubricating oil pollution image data is taken by an industrial microscope camera with a shooting interval of 5 minutes; the real-time motor current data is collected by a high-precision current sensor with a range of 0 - 500A and an accuracy of ±0.5%. The sensor parameters are adapted to the automobile factory environment. For example, the vibration sensor filters low-frequency noise (ground vibration <0.1Hz), and the current sensor range covers the peak load of the stamping machine tool, thus avoiding data overload and distortion.

[0050] Furthermore, the harmonic distortion rate calculation formula is: THD = (A3 / A1) × 100%.

[0051] A1 represents the fundamental wave of the vibration signal; A3 represents the amplitude of the third harmonic. The mechanical shock anomaly is quantified through its calculation result. The oil pollution coverage rate is: C = (S C / S z ) × 100%.

[0052] S C represents the pixel area of the particulate matter region in the oil; S zThe total pixel area of the device monitoring region characterizes key wear and electrical characteristics through its calculation results. Real-time electromechanical current normalization maps the original value to the 0-1 interval.

[0053] Step 2: Generate a dynamic weight coefficient according to the ratio of the real-time motor current data to the rated current of the device; specifically: the calculation formula for the dynamic weight coefficient is:

[0054] α = A d / A j ;

[0055] where, A d represents the current current, A j represents the rated current designed for the device. When the load is high, vibration data is emphasized (due to increased mechanical stress), and when the load is low, oil contamination data is emphasized.

[0056] The value range of the dynamic weight coefficient is limited to [0-1] to avoid distortion of weight calculation under over-limit working conditions. When overloaded, α is forcibly set to 1, and at this time, the health score completely depends on the vibration data. Thus, the mechanical failure risk under high load is preferentially warned.

[0057] Step 3: Combine the harmonic distortion rate, oil contamination coverage rate, and dynamic weight coefficient, and calculate the real-time health score H1 through weighted summation.

[0058] H1(t) represents the health score at time t;

[0059] H1(t) = α·(1 - THD) + (1 - α)·(1 - C(t)).

[0060] where, THD represents the harmonic distortion rate, and C(t) represents the oil contamination coverage rate at time t. In the formula, both THD and C(t) are normalized to 0-1 values, and their calculation results comprehensively reflect the dynamic integration of the mechanical and contamination states.

[0061] Step 4: Based on the real-time health score H1 and the time-domain integral of the oil contamination coverage rate, construct a wear-fatigue coupling model to calculate and output the alarm prediction L value. Finally, when the alarm prediction L value exceeds the set threshold, an alarm is issued, and a device maintenance instruction is generated and the device operation parameters are adjusted.

[0062] Specifically: the calculation formula for the wear-fatigue coupling model is:

[0063] L = ∫H1(t)·C(t)·Ddt / δ;

[0064] Among them, δ represents the fatigue limit threshold of the material, D represents the material damage rate coefficient, with the unit of cycles per hour. The result calculated by the above integral model reflects the quantification of the synergistic effect of pollution and fatigue, dynamically correlates the health score with material damage, replaces the traditional linear superposition model, triggers an early warning when L>1, and when the real-time current exceeds 90% of the rated current, the integral calculation introduces an accelerated aging coefficient β, and β>1.

[0065] The above-mentioned accelerated aging coefficient is used to cope with the non-linear damage in the high-current environment. That is to say, assuming that the current exceeds 90% of the rated current design value of the equipment, the calculation formula of the wear-fatigue coupling model becomes: L = β×∫H1(t)·C(t)·Ddt / δ. Thereby improving the prediction accuracy under high-current conditions and ensuring that the early warning can be triggered in advance.

[0066] The maintenance instructions include: recommended maintenance time, list of replacement parts.

[0067] Adjusting the equipment operation parameters includes: dynamically reducing the stamping frequency of the stamping machine to 70%-90% of the rated value according to the mapping relationship between the L value and the remaining life of the equipment; specifically including:

[0068] The first threshold ≤ L < the second threshold; the stamping frequency of the equipment is reduced to 90% of the rated value;

[0069] The second threshold ≤ L < the third threshold; the stamping frequency of the equipment is reduced to 80% of the rated value;

[0070] L ≥ the third threshold; the stamping frequency of the equipment is reduced to 70% of the rated value;

[0071] And associate the maintenance instructions with the real-time health score H1 and store them in the maintenance log database.

[0072] Example 2

[0073] The difference from Example 1 is that: after each equipment maintenance, compare the actual remaining life data with the predicted life mapped by the predicted L value (the time required to reach L = 1 under a constant damage rate), and update the integral coefficient of the wear-fatigue coupling model through the Bayesian network; specifically, update the material fatigue limit threshold δ.

[0074] In the following scenario: the fatigue limit threshold of the bearing material of a certain automobile stamping machine is initially set to δ = 1.2×10 6 cycles.

[0075] In 3 consecutive maintenance cycles, the predicted remaining life of the model is 15% smaller than the actual life (that is, the predicted life is 85,000 stamping times, and the actual life is 100,000 stamping times). It shows that the current δ value setting (1.2×10 6)Smaller → The model believes that the material is more prone to fatigue → Earlier warning (predicted short lifespan).

[0076] Bayesian network parameter update:

[0077] S1: Prior probability setting: Assume that the possible range of the δ value is 1.0×10 6 ~1.5×10 6 cycles, and the initial prior probability distribution is a uniform distribution.

[0078] S2: Likelihood function construction:

[0079] According to the formula L = ∫H1(t)·C(t)·Ddt / δ,

[0080] D is the damage rate (unit: cycles per hour), calculated based on the actual stamping frequency of historical data;

[0081] When the actual lifespan is 100,000 stamping cycles, the predicted lifespan should be 100,000 cycles, but the original model calculates that the lifespan corresponding to L = 1.0 is 85,000 cycles.

[0082] It can be seen that when the δ value is corrected to δ' = δ×(actual lifespan / predicted lifespan), that is

[0083] Posterior probability calculation: The Bayesian network updates the probability distribution of the δ' value through error data, and the range of the corrected δ' value converges to 1.35×10 6 ~1.45×10 6 and finally selects δ' = 1.41×10 6 .

[0084] S3: Finally, the updated δ' is re-substituted into the calculation formula of the model for verification.

[0085] The advantage of the Bayesian network lies in small-sample learning: Only 3 - 5 maintenance data are required to adjust the parameters, which is suitable for the characteristics of small data volume in industrial scenarios.

[0086] In summary, the present invention adjusts the weights of vibration and oil contamination data in real time through motor current data, solves the evaluation deviation caused by traditional fixed weights, and at the same time constructs an integral model to quantify the synergistic effect of health score, contamination coverage rate, and damage rate, greatly reducing the life prediction error. An acceleration aging coefficient triggered by current is introduced to improve the prediction accuracy under high-load conditions. The fatigue limit threshold of the material is dynamically updated based on the Bayesian network to achieve adaptive calibration of model parameters. Through multi-physical field coupling modeling, dynamic weight allocation, and closed-loop self-optimization mechanism, the accuracy and real-time performance of equipment health status analysis in the automotive manufacturing scenario are solved, especially applicable to high-load equipment such as stamping machines and welding robots. Compared with traditional methods, the comprehensive operation and maintenance cost is reduced, and the unplanned downtime is reduced.

[0087] This embodiment further provides a computer device applicable to a method for analyzing the health status of industrial equipment, including: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement a method for analyzing the health status of industrial equipment as proposed in the above embodiment.

[0088] The computer device may be a terminal, and the computer device includes a processor, a memory, a communication interface, a display screen, and an input device connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be achieved through WIFI, a carrier network, NFC (Near Field Communication), or other technologies. The display screen of the computer device may be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device may be a touch layer covered on the display screen, or a button, a trackball, or a touchpad provided on the housing of the computer device, or an external keyboard, a touchpad, or a mouse, etc.

[0089] This embodiment also provides a storage medium, on which a computer program is stored. When the program is executed by a processor, it implements the method for analyzing the health status of industrial equipment proposed in the above embodiment; the storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM for short), electrically erasable programmable read-only memory (EEPROM for short), erasable programmable read-only memory (EPROM for short), programmable read-only memory (PROM for short), read-only memory (ROM for short), magnetic memory, flash memory, magnetic disk or optical disc.

[0090] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention, and they should all be covered by the scope of the claims of the present invention.

Claims

1. A method for analyzing the health status of industrial equipment, characterized in that, Including: Firstly, obtain the vibration data, lubricating oil pollution image data and real-time motor current data of the stamping machine tool in the automobile factory; perform harmonic decomposition on the vibration data to calculate the harmonic distortion rate, perform edge detection on the lubricating oil pollution image data to calculate the oil pollution coverage rate, and perform normalization processing on the real-time motor current data; Then, generate a dynamic weight coefficient according to the ratio of the real-time motor current data to the rated current of the equipment; Next, combine the harmonic distortion rate, oil pollution coverage rate and dynamic weight coefficient, and calculate the real-time health score H1 through weighted summation; Based on the real-time health score H1 and the time-domain integral of the oil pollution coverage rate, construct a wear-fatigue coupling model to calculate and output the warning prediction L value. Finally, when the warning prediction L value exceeds the set threshold, generate an equipment maintenance instruction and adjust the equipment operation parameters.

2. The method for analyzing the health status of an industrial device according to claim 1, wherein: The vibration data, lubricating oil pollution image data and real-time motor current data are synchronously processed through a spatio-temporal alignment model; The spatio-temporal alignment model matches the data streams of different sensors in terms of timestamp and equipment spatial coordinates by identifying the mutation points in the current data.

3. The method for analyzing the health status of an industrial device according to claim 2, wherein: The harmonic distortion rate is the amplitude ratio of the fundamental wave to the third harmonic of the vibration signal, and the oil pollution coverage rate is the percentage of the pixel area of the particulate matter region in the oil fluid to the total pixel area of the equipment monitoring region.

4. The method for analyzing the health status of an industrial device according to claim 3, wherein: The calculation formula of the dynamic weight coefficient is: α = current current / equipment design limit current; Among them, the value range of the dynamic weight coefficient is [0 - 1]. When the real-time current exceeds the rated current, α is forcibly set to 1.

5. The method for analyzing the health state of an industrial device according to claim 4, wherein: The calculation formula of the wear-fatigue coupling model is: L = ∫H1(t)·C(t)·Ddt / δ; Among them, H1(t) represents the health score at time t, and D represents the material damage rate coefficient; H1(t) = α·(1 - THD) + (1 - α)·(1 - C(t)); THD represents the harmonic distortion rate, C(t) represents the oil pollution coverage rate at time t; δ represents the material fatigue limit threshold. When L > 1, a warning is triggered. And when the real-time current exceeds 90% of the rated current, an acceleration aging coefficient β is introduced in the integral calculation, and β > 1.

6. The method for analyzing the health state of an industrial device according to claim 5, wherein: Adjusting the equipment operation parameters includes: dynamically reducing the stamping frequency of the stamping machine tool to 70% - 90% of the rated value according to the mapping relationship between the L value and the remaining life of the equipment; specifically including: The first threshold ≤ L < the second threshold; the stamping frequency of the equipment is reduced to 90% of the rated value; The second threshold ≤ L < the third threshold; the stamping frequency of the equipment is reduced to 80% of the rated value; L ≥ the third threshold; the stamping frequency of the equipment is reduced to 70% of the rated value; And associate the maintenance instruction with the real-time health score H1 and store it in the maintenance log database.

7. The method for analyzing the health status of an industrial device according to claim 6, characterized in that: After each equipment maintenance, compare the actual remaining life data with the predicted life corresponding to the predicted L value, and update the integral coefficient of the wear-fatigue coupling model through a Bayesian network; specifically, update the material fatigue limit threshold δ; That is, when the actual life is greater than the predicted life, the material fatigue limit threshold δ is increased.

8. The method for analyzing the health status of industrial equipment according to claim 7, wherein: The vibration data is collected by a high-frequency vibration sensor, the sampling frequency is 10kHz and the frequency band is limited to 0.1 - 500Hz; The lubricating oil pollution image data is captured by an industrial microscope camera with a shooting interval of 5 minutes; The real-time motor current data is collected by a high-precision current sensor with a measuring range of 0 - 500 A and an accuracy of ±0.5%.

9. A computer device, comprising a memory and a processor, the memory storing a computer program, characterized in that: When the processor executes the computer program, it implements the steps of an industrial equipment health status analysis method according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by the processor, it implements the steps of an industrial equipment health status analysis method according to any one of claims 1 to 7.

Images