CN110631849A - Online identification method and system for wear state of friction system - Google Patents
Online identification method and system for wear state of friction system Download PDFInfo
- Publication number
- CN110631849A CN110631849A CN201910733557.9A CN201910733557A CN110631849A CN 110631849 A CN110631849 A CN 110631849A CN 201910733557 A CN201910733557 A CN 201910733557A CN 110631849 A CN110631849 A CN 110631849A
- Authority
- CN
- China
- Prior art keywords
- friction
- wear
- friction coefficient
- correlation dimension
- friction system
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
- G01M99/00—Subject matter not provided for in other groups of this subclass
- G01M99/007—Subject matter not provided for in other groups of this subclass by applying a load, e.g. for resistance or wear testing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/02—Preprocessing
- G06F2218/04—Denoising
- G06F2218/06—Denoising by applying a scale-space analysis, e.g. using wavelet analysis
Landscapes
- Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Biomedical Technology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Testing Of Devices, Machine Parts, Or Other Structures Thereof (AREA)
- Length Measuring Devices With Unspecified Measuring Means (AREA)
Abstract
A friction system wear state on-line identification method comprises the steps of firstly collecting friction coefficient signals through a sensor, secondly, carrying out noise reduction processing on the friction coefficient signals, and calculating a characterization parameter-correlation dimension value of the friction coefficient signals; and then comparing the data with the correlation dimension prediction value of the friction system database to judge the wear state of the friction system. The invention provides a friction system wear state online identification method and system, which can monitor and identify the friction system wear state online in real time, effectively judge the dynamic characteristics of the friction system and prevent sudden failure of a mechanical system; the method has the advantages of high identification accuracy and high reliability, and avoids loss of a lot of system characteristic information caused by simple analysis of the time evolution of the univariate friction coefficient signal.
Description
Technical Field
The invention relates to the field of friction wear and wear state identification, in particular to a friction system wear state online identification method and system.
Background
In engineering application, the running-in time of a mechanical friction pair is always expected to be shortened, the time of a stable abrasion state is prolonged, an equipment failure node is obtained in advance, and parts or maintenance equipment are replaced in time, so that the equipment loss is reduced, and the economic benefit is improved. Therefore, it is important to monitor and control the wear status of the mechanical equipment.
At present, engineers mainly rely on long-term working experience to judge and monitor the operating state of mechanical equipment, which has a very high requirement on the work literacy of workers. The wear state is usually judged through a vibration signal or a noise signal, and the alertness of workers to the running state of the equipment can be caused only when a friction system suddenly vibrates violently or generates harsh squeaking sound. Or the current wear state of the friction system is judged by intermittently selecting oil samples to perform abrasive particle analysis. However, these methods cannot monitor the wear state of the friction system in real time, and it is more difficult to predict the wear state in advance and prevent the mechanical parts from being seriously damaged.
Disclosure of Invention
In order to overcome the defects in the prior art, the invention provides an online identification method and system for the wear state of a friction system, which solve the problem that the wear state of the friction system cannot be identified and predicted online in real time, can monitor and identify the wear state of the friction system online in real time, effectively judge the dynamic characteristics of the friction system and prevent sudden failure of a mechanical system; the method has the advantages of high identification accuracy and high reliability, and avoids loss of a lot of system characteristic information caused by simple analysis of the time evolution of the univariate friction coefficient signal.
The technical scheme adopted by the invention for solving the technical problems is as follows:
an on-line identification method for a wear state of a friction system comprises the following steps:
(1) acquiring friction coefficient signals generated in a friction wear test of a friction system under different system parameters;
(2) denoising the friction coefficient signal by using a wavelet method, and calculating the correlation dimension value of the friction coefficient signal in a stable wear state by using a chaos theory;
(3) calling an association dimension prediction value corresponding to a test system parameter from a friction system database, and comparing the association dimension prediction value with an association dimension value obtained through continuous calculation in real time;
(4) and (4) judging the current wear state according to the comparison result, and continuing the step (2) and the step (3) after the friction system enters the stable wear state.
Further, in the step (1), the load and the speed of a friction system and the initial surface roughness of the upper and lower samples are respectively changed, a plurality of groups of friction and wear tests are carried out, and friction coefficient signals are collected in real time.
Still further, the step (2) further comprises the following steps:
calculating time delay tau and an optimal embedding dimension m of a friction coefficient signal time sequence;
secondly, reconstructing a phase space of the friction coefficient signal according to the following formula;
calculating the correlation dimension value of the reconstructed friction coefficient signal by using a chaos theory, wherein the calculation method comprises the following steps:
wherein n is a data point xiThe number of (2); n is a vector XiThe number of (2); cm(r) is the correlation integral; h (·) is a step function; r is a small scalar; d is the correlation dimension.
Still further, the establishing of the friction system database in the step (3) comprises the following steps:
selecting load, speed and initial surface roughness of an upper sample and a lower sample as main system parameters of a friction system, and carrying out a large number of friction and wear tests by changing different levels of the four parameters;
performing wavelet denoising processing on the friction coefficient signal, and calculating the correlation dimension value of the friction coefficient signal;
establishing a neural network model, selecting a training sample and a testing sample, and training and testing a neural network; when the neural network meets the prediction precision requirement, determining a prediction neural network model;
predicting the running-in time, the friction coefficient value and the correlation dimension value of the four system parameters under the corresponding stable wear states under different level combinations based on the prediction neural network model;
finally, establishing a friction system database by using an MATLAB program according to the running-in time, the friction coefficient value and the correlation dimension prediction value corresponding to a large number of different system parameter combinations.
Furthermore, in step (4), when the values of the three consecutive correlation dimensions all satisfy D ≧ D1Under the condition D1The correlation dimension prediction value indicates that the friction system is in a stable abrasion state at the moment; otherwise, the friction system is in a running-in wear state, and the friction wear test is continued;
when three consecutive correlation dimension values all satisfy D<D1When the condition is met, the friction system is indicated to start to enter a severe abrasion state, and the machine is stopped in time; otherwise, the friction system is in a stable abrasion state, and the friction abrasion test is continued.
A system based on a friction system wear state online identification method comprises a signal acquisition module and a signal processing module;
the signal acquisition module comprises a torque sensor and a data acquisition card, acquires a friction coefficient signal voltage value in the abrasion process of the friction system through the torque sensor, and transmits data to a PC client through the data acquisition card so as to acquire friction coefficient signals under different system parameters and different abrasion states; the system parameters comprise load, speed, initial surface roughness of an upper sample and initial surface roughness of a lower sample; the wear state comprises three states of running-in wear, stable wear and severe wear;
the signal processing module comprises a friction system database, an A \ D converter and a signal amplifier, wherein the friction system database is used for corresponding the correlation dimension prediction values under different system parameters to corresponding running-in wear, stable wear and severe wear states; converting the voltage signal into a friction coefficient signal time sequence through an A \ D converter, acquiring the friction coefficient signal in real time through a signal amplifier, and calling an MATLAB correlation dimension calculation program through an LabVIEW program to calculate a correlation dimension value; an MATLAB program is called through a LabVIEW program block diagram to compare and analyze the correlation dimension value obtained in real time with the correlation dimension prediction value in the friction system database, and the abrasion state of the current friction system is judged; and displaying the change of the friction coefficient signal generated by the friction system in real time along with the time and the corresponding correlation dimension value through a LabVIEW user interface.
The invention has the following beneficial effects: the online wear state identification system can acquire a friction coefficient signal in real time, continuously calculate the correlation dimension value of the friction coefficient signal, compare the correlation dimension value with the corresponding correlation dimension value in the database and judge the current wear state of the friction system in real time; the system realizes online real-time identification of the wear state, has the advantages of strong reliability and high accuracy, provides a powerful basis for prediction of the wear state, reduces the loss of mechanical equipment, and improves the economic benefit of production.
Drawings
Fig. 1 is a flow chart of online identification of a wear state of a friction system.
FIG. 2 is a flow chart of the friction system database creation of the present invention.
Fig. 3 is a structural diagram of a wear state identification system of the friction system of the present invention.
Detailed Description
The invention is further described below with reference to the accompanying drawings.
As shown in fig. 1, in the method for online identifying the wear state of a friction system, firstly, a friction coefficient signal is acquired through a sensor, secondly, noise reduction processing is performed on the friction coefficient signal, and a characterization parameter-correlation dimension value of the friction coefficient signal is calculated; and then comparing the data with the correlation dimension prediction value of the friction system database to judge the wear state of the friction system. The method specifically comprises the following steps:
(1) and changing the load and speed of a friction system and the initial surface roughness of the upper and lower samples, performing a plurality of groups of friction and wear tests, and acquiring friction coefficient signals in real time, wherein the corresponding sampling frequency is 300 Hz.
(2) Every 10000 data points in the continuously collected friction coefficient signal are taken as a stage, the noise reduction processing is carried out on the friction coefficient signal by using a wavelet method, and then the time delay tau and the optimal embedding dimension m are calculated; performing phase space reconstruction on the friction coefficient signal according to the following formula;
finally, the correlation dimension value of the reconstructed friction coefficient signal is calculated using the following method.
Wherein n is a data point xiThe number of (2); n is a vector XiThe number of (2); cm(r) is the correlation integral; h (·) is a step function; r is a small scalar; d is the correlation dimension.
(3) Calling correlation dimension predicted value D corresponding to test system parameters from friction system database1And comparing the correlation dimension value D with the correlation dimension value D obtained by continuous calculation in real time;
(4) judging the current wear state according to the comparison result, and when the three continuous correlation dimension values all meet the condition that D is more than or equal to D1When the condition is met, the friction system is in a stable abrasion state; otherwise, the friction system is in a running-in wear state, and the friction wear test is continued;
(5) when the friction system enters a stable abrasion state, continuing the step (2) and the step (3);
(6) judging the current wear state according to the comparison result, and when the three continuous correlation dimension values all meet D<D1And when the condition is met, the friction system is indicated to be in a severe abrasion state, and the machine is stopped in time. Otherwise, the friction system is in a stable abrasion state, and the friction abrasion test is continued.
The friction system database in step (3) is established as shown in fig. 2, and the specific process is as follows:
selecting load, speed, initial surface roughness of an upper sample and initial surface roughness of a lower sample as main system parameters of a friction system, carrying out a large number of friction wear tests by changing different levels of the four parameters, and repeating the tests for at least three times on each system parameter in order to ensure the reliability of test data;
after all tests are finished, performing wavelet denoising processing on the friction coefficient signals, and calculating the correlation dimension values of the friction coefficient signals;
establishing a neural network model, taking the load, the speed, the initial surface roughness of the upper sample and the initial surface roughness of the lower sample as input parameters, taking the running-in time, the friction coefficient value and the correlation dimension value corresponding to the stable abrasion state as output parameters, respectively selecting a training sample and a test sample, training and testing the neural network, and finally determining a prediction neural network model;
predicting the running-in time, the friction coefficient value and the correlation dimension value of the four system parameters under the corresponding stable wear states under different level combinations based on the prediction neural network model;
finally, establishing a friction system database by using an MATLAB program according to the running-in time, the friction coefficient value and the correlation dimension prediction value corresponding to a large number of different system parameter combinations.
As shown in fig. 3, an online identification system for wear status of friction system comprises a signal acquisition module and a signal processing module, wherein,
the signal acquisition module is used for acquiring a friction coefficient signal voltage value in the abrasion process of the friction system through the torque sensor and then transmitting data to the PC client through the data acquisition card so as to acquire friction coefficient signals under different abrasion states under different system parameters; the system parameters comprise load, speed, initial surface roughness of an upper sample and initial surface roughness of a lower sample; the wear state comprises three states of running-in wear, stable wear and severe wear.
The signal processing module includes:
the friction system database corresponds the correlation dimension prediction values corresponding to different system parameters to corresponding running-in wear, stable wear and severe wear states;
converting the voltage signal into a friction coefficient signal time sequence through an A \ D converter, acquiring the friction coefficient signal in real time through a signal amplifier, and calling an MATLAB correlation dimension calculation program through an LabVIEW program to calculate a correlation dimension value;
calling an MATLAB program through a LabVIEW program block diagram to compare and analyze the correlation dimension value obtained in real time with the correlation dimension prediction value in the friction system database, and judging the wear state of the current friction system; and displaying the change of the friction coefficient signal generated by the friction system in real time along with the time and the corresponding correlation dimension value through a LabVIEW user interface.
The invention discloses a real-time online wear state identification method based on a prediction neural network, which comprises the steps of developing an online identification system through a signal acquisition module and a signal processing module, acquiring a friction coefficient signal online, calculating a characteristic parameter value of the friction coefficient signal, comparing the characteristic parameter value with a corresponding parameter value in a database, and displaying the current wear state of the friction system in real time. The system can effectively, accurately and real-timely identify the current running-in state and provide a powerful basis for predicting the wear state in advance. The system realizes the intellectualization of the friction system wear state identification and avoids unnecessary economic loss.
The embodiments are only for illustrating the technical idea of the present invention, and the technical idea of the present invention is not limited thereto, and any modifications made on the basis of the technical scheme according to the technical idea of the present invention fall within the scope of the present invention.
Claims (6)
1. An on-line identification method for a friction system wear state is characterized in that: the method comprises the following steps:
(1) acquiring friction coefficient signals generated in a friction wear test of a friction system under different system parameters;
(2) denoising the friction coefficient signal by using a wavelet method, and calculating the correlation dimension value of the friction coefficient signal in a stable wear state by using a chaos theory;
(3) calling an association dimension prediction value corresponding to a test system parameter from a friction system database, and comparing the association dimension prediction value with an association dimension value obtained through continuous calculation in real time;
(4) and (4) judging the current wear state according to the comparison result, and continuing the step (2) and the step (3) after the friction system enters the stable wear state.
2. A friction system wear status on-line identification method as claimed in claim 1, characterized in that: in the step (1), the load and the speed of a friction system and the initial surface roughness of an upper sample and a lower sample are respectively changed, a plurality of groups of friction and wear tests are carried out, and friction coefficient signals are collected in real time.
3. A friction system wear status on-line recognition method as claimed in claim 2, characterized in that: the step (2) further comprises the following steps:
calculating time delay tau and an optimal embedding dimension m of a friction coefficient signal time sequence;
secondly, reconstructing a phase space of the friction coefficient signal according to the following formula;
calculating the correlation dimension value of the reconstructed friction coefficient signal by using a chaos theory, wherein the calculation method comprises the following steps:
wherein n is a data point xiThe number of (2); n is a vector XiThe number of (2); cm(r) is the correlation integral; h (·) is a step function; r is a small scalar; d is the correlation dimension.
4. A friction system wear status on-line recognition method as claimed in claim 3, characterized in that: the establishment of the friction system database in the step (3) comprises the following steps:
selecting load, speed and initial surface roughness of an upper sample and a lower sample as main system parameters of a friction system, and carrying out a large number of friction and wear tests by changing different levels of the four parameters;
performing wavelet denoising processing on the friction coefficient signal, and calculating the correlation dimension value of the friction coefficient signal;
establishing a neural network model, selecting a training sample and a testing sample, and training and testing a neural network; when the neural network meets the prediction precision requirement, determining a prediction neural network model;
predicting the running-in time, the friction coefficient value and the correlation dimension value of the four system parameters under the corresponding stable wear states under different level combinations based on the prediction neural network model;
finally, establishing a friction system database by using an MATLAB program according to the running-in time, the friction coefficient value and the correlation dimension prediction value corresponding to a large number of different system parameter combinations.
5. A friction system wear state on-line identification method as claimed in claim 3 or 4 wherein: in the step (4), when the values of the three continuous correlation dimensions all meet that D is larger than or equal to D1Under the condition D1The correlation dimension prediction value indicates that the friction system is in a stable abrasion state at the moment; otherwise, the friction system is in a running-in wear state, and the friction wear test is continued;
when three consecutive correlation dimension values all satisfy D<D1When the condition is met, the friction system is indicated to start to enter a severe abrasion state, and the machine is stopped in time; otherwise, the friction system is in a stable abrasion state, and the friction abrasion test is continued.
6. A system based on the friction system wear state online identification method of claim 1, characterized in that: the device comprises a signal acquisition module and a signal processing module;
the signal acquisition module comprises a torque sensor and a data acquisition card, acquires a friction coefficient signal voltage value in the abrasion process of the friction system through the torque sensor, and transmits data to a PC client through the data acquisition card so as to acquire friction coefficient signals under different system parameters and different abrasion states; the system parameters comprise load, speed, initial surface roughness of an upper sample and initial surface roughness of a lower sample; the wear state comprises three states of running-in wear, stable wear and severe wear;
the signal processing module comprises a friction system database, an A \ D converter and a signal amplifier, wherein the friction system database is used for corresponding the correlation dimension prediction values under different system parameters to corresponding running-in wear, stable wear and severe wear states; converting the voltage signal into a friction coefficient signal time sequence through an A \ D converter, acquiring the friction coefficient signal in real time through a signal amplifier, and calling an MATLAB correlation dimension calculation program through an LabVIEW program to calculate a correlation dimension value; an MATLAB program is called through a LabVIEW program block diagram to compare and analyze the correlation dimension value obtained in real time with the correlation dimension prediction value in the friction system database, and the abrasion state of the current friction system is judged; and displaying the change of the friction coefficient signal generated by the friction system in real time along with the time and the corresponding correlation dimension value through a LabVIEW user interface.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910733557.9A CN110631849B (en) | 2019-08-09 | 2019-08-09 | Online identification method and system for wear state of friction system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910733557.9A CN110631849B (en) | 2019-08-09 | 2019-08-09 | Online identification method and system for wear state of friction system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110631849A true CN110631849A (en) | 2019-12-31 |
| CN110631849B CN110631849B (en) | 2021-04-06 |
Family
ID=68970041
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910733557.9A Active CN110631849B (en) | 2019-08-09 | 2019-08-09 | Online identification method and system for wear state of friction system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110631849B (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111323366A (en) * | 2020-03-09 | 2020-06-23 | 上海中科深江电动车辆有限公司 | Synchronizer friction coefficient-based detection method and corresponding wear test bed |
| CN112525749A (en) * | 2020-11-19 | 2021-03-19 | 扬州大学 | Tribology state online identification method based on friction signal recursion characteristic |
| CN115166031A (en) * | 2022-05-10 | 2022-10-11 | 清华大学 | Method for determining friction performance and friction test equipment |
| CN117074291A (en) * | 2023-10-17 | 2023-11-17 | 西南交通大学 | Non-contact texture friction prediction method |
-
2019
- 2019-08-09 CN CN201910733557.9A patent/CN110631849B/en active Active
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111323366A (en) * | 2020-03-09 | 2020-06-23 | 上海中科深江电动车辆有限公司 | Synchronizer friction coefficient-based detection method and corresponding wear test bed |
| CN112525749A (en) * | 2020-11-19 | 2021-03-19 | 扬州大学 | Tribology state online identification method based on friction signal recursion characteristic |
| CN115166031A (en) * | 2022-05-10 | 2022-10-11 | 清华大学 | Method for determining friction performance and friction test equipment |
| CN115166031B (en) * | 2022-05-10 | 2024-04-19 | 清华大学 | Friction performance determining method and friction test equipment |
| CN117074291A (en) * | 2023-10-17 | 2023-11-17 | 西南交通大学 | Non-contact texture friction prediction method |
| CN117074291B (en) * | 2023-10-17 | 2024-01-02 | 西南交通大学 | Non-contact texture friction prediction method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110631849B (en) | 2021-04-06 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110631849B (en) | Online identification method and system for wear state of friction system | |
| CN111037365B (en) | Cutter state monitoring data set enhancing method based on generative countermeasure network | |
| CN113305645B (en) | Numerical control machine tool cutter residual life prediction method based on hybrid neural model | |
| Boutros et al. | Detection and diagnosis of bearing and cutting tool faults using hidden Markov models | |
| Camci et al. | Feature evaluation for effective bearing prognostics | |
| CN103034170B (en) | Numerical control machine tool machining performance prediction method based on intervals | |
| CN109212966B (en) | Multi-working-condition dynamic benchmarking mechanical equipment residual life prediction method | |
| CN101520652A (en) | Method for evaluating service reliability of numerical control equipment | |
| CN111458149B (en) | Method and system for predicting performance and service life of rolling bearing | |
| CN108544303A (en) | A kind of main shaft of numerical control machine tool fault diagnosis method and system | |
| CN112525749B (en) | Tribology state online identification method based on friction signal recursion characteristic | |
| CN103217248A (en) | Method for detecting tensile force of bridge steel cable | |
| Medjaher et al. | Feature extraction and evaluation for health assessment and failure prognostics | |
| CN115034137A (en) | RVM and degradation model-based two-stage hybrid prediction method for residual life of bearing | |
| CN109710955A (en) | Based on LCD-recurrence quantification analysis rolling bearing fault diagnosis and health evaluating method | |
| CN113742993A (en) | Method, device, equipment and storage medium for predicting life loss of dry-type transformer | |
| CN109408998B (en) | Fatigue life evaluation method by rapidly acquiring stress spectrum based on sample continuous increment | |
| CN103203670A (en) | Grinding chatter prediction method based on maximum information entropy and directional divergence | |
| CN118060972A (en) | Cutter wear monitoring method based on discrete wavelet decomposition | |
| CN113469408A (en) | Running state trend prediction method and system for phase modulator | |
| CN110633314A (en) | Internet of vehicles data processing method and device | |
| CN105675321A (en) | Determination method for equipment performance degradation radar map | |
| CN109165396A (en) | A kind of equipment remaining life prediction technique of failure evolution trend | |
| CN115719008A (en) | Industrial robot joint fault prediction method, equipment and storage medium | |
| CN108021725B (en) | Method for determining reversing point position of oil pumping unit in electric diagram data |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |