CN110375974B - Method for monitoring state of rotating mechanical equipment based on data boundary form after planarization - Google Patents
Method for monitoring state of rotating mechanical equipment based on data boundary form after planarization Download PDFInfo
- Publication number
- CN110375974B CN110375974B CN201910669452.1A CN201910669452A CN110375974B CN 110375974 B CN110375974 B CN 110375974B CN 201910669452 A CN201910669452 A CN 201910669452A CN 110375974 B CN110375974 B CN 110375974B
- Authority
- CN
- China
- Prior art keywords
- monitoring
- boundary
- points
- characteristic parameters
- point
- 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.)
- Active
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
- G01M13/00—Testing of machine parts
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
- G01M13/00—Testing of machine parts
- G01M13/02—Gearings; Transmission mechanisms
- G01M13/021—Gearings
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
- G01M13/00—Testing of machine parts
- G01M13/02—Gearings; Transmission mechanisms
- G01M13/022—Power-transmitting couplings or clutches
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
- G01M13/00—Testing of machine parts
- G01M13/02—Gearings; Transmission mechanisms
- G01M13/028—Acoustic or vibration analysis
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
- G01M13/00—Testing of machine parts
- G01M13/04—Bearings
- G01M13/045—Acoustic or vibration analysis
Abstract
A method for monitoring the state of rotary mechanical equipment based on the boundary form of data after planarization comprises the steps of extracting the characteristic parameters of sample points of the rotary mechanical equipment, utilizing a constellation diagram technology to map the characteristic parameters of the sample points onto a plane in a dimensionality reduction mode, eliminating outliers in a distance-based mode, calculating the boundary form of a plane point set, obtaining a boundary point set, and constructing a visual monitoring model; then extracting characteristic parameters of monitoring points of the rotary mechanical equipment, inputting the characteristic parameters of the monitoring points into a visual monitoring model, and judging whether the characteristic parameters are abnormal or not; if the monitoring point is normal, updating the monitoring model by using a window sliding method; if the monitoring point is abnormal, alarming by using the monitoring model after the boundary form expansion; the invention can real-time depict the operation state of the rotating mechanical equipment, thereby establishing the direct correlation between the state and the graph and providing new technical support for state monitoring.
Description
Technical Field
The invention relates to the technical field of rotating machinery equipment state detection, in particular to a rotating machinery equipment state monitoring method based on planarized data boundary morphology.
Background
With the continuous large-scale, complicated and intelligent transformation of mechanical equipment and the wide application of a state monitoring system, the data volume of the operation state of the equipment acquired by people is larger and larger. Under the background of mass data, how to mine hidden knowledge in equipment operation data and enhance the intuitiveness of a decision process of a state monitoring and fault diagnosis technology are the hot spots of current research.
The traditional state monitoring technology is characterized by a characteristic quantity mode, but the characteristic quantity contains less hidden knowledge of equipment, and the change of a single characteristic is not enough to judge the change of the equipment state; and the decision making process is invisible, similar to black box operation, people are difficult to find the hidden knowledge in the data.
The prior art has the defects of single characterization characteristic quantity and invisible decision process.
Disclosure of Invention
In order to overcome the disadvantages of the prior art, an object of the present invention is to provide a method for monitoring the state of a rotating mechanical device based on a planarized data boundary shape, which can depict the operation status of the device in real time, thereby establishing a direct relationship between the state and the pattern, and providing a new technical support for state monitoring.
In order to achieve the purpose, the technical scheme adopted by the invention is as follows:
a method for monitoring the state of rotating mechanical equipment based on the shape of a data boundary after planarization comprises the following steps:
step 1: extracting characteristic parameters of sample points of the rotary mechanical equipment;
step 2: using a constellation diagram technology to map the characteristic parameters of the sample points onto a plane in a dimensionality reduction way;
and step 3: removing outliers in a distance-based mode;
and 4, step 4: calculating the boundary form of the plane point set, obtaining a boundary point set, and constructing a visual monitoring model;
and 5: extracting characteristic parameters of monitoring points of the rotary mechanical equipment;
step 6: inputting the characteristic parameters of the monitoring points into a visual monitoring model, and judging whether the characteristic parameters are abnormal or not; if the monitoring point is normal, updating the monitoring model by using a window sliding method; and if the monitoring point is abnormal, alarming by using the monitoring model after the boundary form is expanded.
The rotating mechanical equipment in the step 1 comprises a fan, a motor, a gear box, rotor equipment of a coupler and a bearing; the characteristic parameters of the sample points are derived from vibration signals of the rotary mechanical equipment, and the vibration signals comprise speed signals, acceleration signals and impact signals; and extracting speed signals or acceleration signals for rotor equipment of a fan, a coupling, a gear box and a motor, and extracting impact signals for a bearing.
In the step 1, the characteristic parameters of the sample points comprise dimensional parameter indexes and dimensionless parameter indexes, the dimensional parameter indexes comprise standard deviation, variance, peak-to-peak value, square root amplitude, average amplitude and root-mean-square value, and the dimensionless parameter indexes comprise kurtosis indexes, peak value indexes, waveform indexes, pulse indexes and margin indexes.
The basic principle of the constellation diagram technology in the step 2 is that N-dimensional features of data are fused into a point to be displayed in a semicircle, sample points closer to each other in the constellation diagram are similar, sample points farther from each other are different, and the positions of different types of data in the constellation diagram are different.
And 3, eliminating outliers based on a distance mode, measuring the distance between the objects by defining a distance function, and determining the outliers in the data set according to the obvious deviation of the distances of most objects in the data.
The boundary point set in step 4 is based on the principle that the distribution form of the boundary points of the plane point set replaces the distribution area of the state data.
The characteristic parameters of the monitoring points in the step 5 comprise dimensional parameter indexes and dimensionless parameter indexes, the dimensional parameter indexes comprise standard deviation, variance, peak-to-peak value, square root amplitude, average amplitude and root-mean-square value, and the dimensionless parameter indexes comprise kurtosis indexes, peak value indexes, waveform indexes, pulse indexes and margin indexes.
And 6, judging whether the monitoring points are abnormal or not according to the relative position relationship between the monitoring points and the boundary form.
The window sliding method in the step 6 updates the model by using the boundary points of the original model and the two points, so as to achieve the purpose of improving the algorithm execution efficiency.
In the step 6, the boundary form expansion is to obtain the distance between the flattened points of the adjacent data in the normal state of the equipment, expand the boundary by using the maximum likelihood estimation of the parameters, and multiply the horizontal and vertical coordinates of the primary boundary point set by the expansion proportion coefficient respectively to obtain a new boundary point set.
In the step 6, in order to reduce false alarm when alarming, two characteristics of the boundary form and the expansion ratio coefficient after data planarization are integrated to serve as the alarming condition of the monitoring model, and the point is considered as an abnormal point only when the alarming condition is exceeded; in order to make the monitoring model have better visualization effect, the arrangement is made that the normal point is displayed as green, and the abnormal point is displayed as red.
Compared with the prior art, the method has the following advantages:
1. the method provides a rotating mechanical equipment state monitoring method based on the planarized data boundary form, and realizes real-time tracking of the rotating mechanical equipment state.
2. The method disclosed by the invention utilizes a constellation diagram technology to map characteristic parameters of sample points onto a plane in a dimensionality reduction manner, eliminates outliers in a distance-based manner, finally calculates the boundary form of a plane point set, obtains a boundary point set, constructs a visual monitoring model, and provides a more visual display for the monitoring process.
3. The method updates the monitoring model in real time by using a window sliding method, and alarms by using the monitoring model after the boundary form is expanded, thereby increasing the fault tolerance of an alarm mechanism, reducing false alarm and providing more visual evidence for the alarm process.
Drawings
FIG. 1 is a flow chart of the method of the present invention.
FIG. 2 is an external view of the machine tool according to the embodiment.
FIG. 3 is a diagram of the piston parts of the embodiment.
FIG. 4 is a top view of an example tool.
Fig. 5 is a monitoring diagram of the 201 st sampling of the embodiment.
Fig. 6 is a 301 th sampling monitoring chart of the embodiment.
Fig. 7 is a 401 th sampling monitoring chart of the embodiment.
Fig. 8 is a monitoring diagram of the 406 th sampling of the embodiment.
FIG. 9 is a graph showing the variation of the standard deviation of the examples.
Detailed Description
The present invention will be further explained with reference to the following examples, but the scope of the present invention is not limited thereto, and the present example uses data of piston machining process of Shandong double Port piston Co., Ltd, and the example device machining process has five processes of rough turning skirt part, semi-finish turning skirt part, rough turning seam allowance, semi-finish turning seam allowance and finish turning seam allowance. Because the change that utilizes current monitoring with low costs, simple to operate and can better reflect the cutter, consequently gather the current signal in the cutter course of working in the experiment, the process selects half finish turning skirt section stage, experimental facilities and collection parameter specifically as follows:
the model of the machine tool: TNC-300 numerically controlled lathe;
the acquisition equipment: an OR35 signal acquisition instrument, a FLUKE i200s current clamp, and a ThinkPad T400 notebook computer;
the type of the workpiece is as follows: a GG-2813 piston;
the type of the cutter is as follows: MITSUISHI-WNMG 080404-FY;
processing parameters are as follows: the rotating speed of the main shaft is 450rpm, the feeding speed is 108mm/min, and the cutting depth is 1.7 mm;
sampling frequency: 6.4 KHz;
sampling length: 9212;
the machine tool is shown in fig. 2, the workpiece is shown in fig. 3, and the tool is shown in fig. 4.
Referring to fig. 1, a method for monitoring the condition of a rotating machine based on the shape of a planarized data boundary, comprising the steps of:
step 1: extracting characteristic parameters of sample points of the rotary mechanical equipment;
in the experiment, 17 workpieces are machined, and 27 groups of data are collected for each workpiece; when the 16 th workpiece is machined, the cutter is seriously damaged, and the surface of the workpiece is obviously scratched; selecting 12 time domain characteristic indexes of the acceleration signal, and establishing a monitoring model; the time domain characteristics of the tool data comprise dimensional parameter indexes and dimensionless parameter indexes, the dimensional parameter indexes comprise standard deviation, variance, peak-to-peak value, square root amplitude, average amplitude and root mean square value, and the dimensionless parameter indexes comprise kurtosis indexes, peak value indexes, waveform indexes, pulse indexes and margin indexes;
step 2: using a constellation diagram technology to map the characteristic parameters of the sample points onto a plane in a dimensionality reduction way;
and step 3: removing outliers in a distance-based mode;
and 4, step 4: calculating the boundary form of the plane point set, obtaining a boundary point set, and constructing a visual monitoring model;
and 5: extracting characteristic parameters of monitoring points of the rotary mechanical equipment, wherein the characteristic parameters of the monitoring points comprise dimensional parameter indexes and dimensionless parameter indexes, the dimensional parameter indexes comprise standard deviation, variance, peak-to-peak value, square root amplitude value, average amplitude value and root-mean-square value, and the dimensionless parameter indexes comprise kurtosis indexes, peak value indexes, waveform indexes, pulse indexes and margin indexes;
step 6: inputting the characteristic parameters of the monitoring points into a visual monitoring model, and judging whether the characteristic parameters are abnormal or not; if the monitoring point is normal, updating the monitoring model by using a window sliding method; and if the monitoring point is abnormal, alarming by using the monitoring model after the boundary form is expanded.
And the boundary form expansion is used for solving the distance between the flattened points of the adjacent data in the normal state of the equipment, and the maximum likelihood estimation of the parameters is used for expanding the boundary. Multiplying the horizontal and vertical coordinates of the primary boundary point set by the expansion ratio coefficient respectively to obtain a new boundary point set;
in the embodiment, two characteristics of the boundary form and the expansion scale coefficient after data planarization are integrated to serve as the alarm condition of the monitoring model, and the point is considered as an abnormal point only when the alarm condition is exceeded simultaneously; if the monitoring point is normal, updating the monitoring model by using a window sliding method, namely using the boundary point of the original model and the two points; if the monitoring point is abnormal, alarming by using the monitoring model after the boundary form expansion; the monitoring model has better visualization effect, the normal points are displayed as green, the abnormal points are displayed as red, and the dynamic change process of the monitoring model is as shown in figures 5, 6, 7 and 8, because the monitoring process is a dynamic graph change process, the intermediate results are very many, and a part of the intermediate process and the monitoring process with alarm are mainly displayed; the model starts to alarm at the 406 th sampling, and the first sampling of the 16 th workpiece is consistent with the actual situation. The variation trend of the standard deviation of the tool current signal is shown in fig. 9, and it can be obtained that the tool signal fluctuates sharply at the 406 th sampling.
For the cutter used in the embodiment, the life cycle is short, and only about 15 workpieces can be processed in the whole life cycle; the performance degradation of the cutter is rapid according to the change trend of the standard deviation of the current signal of the cutter, when the data of the cutter is processed, the expansion coefficient of the boundary form is relatively large because the data changes violently, the model does not give an alarm when the change degree of the data is unchanged, and the alarm line set based on the expanded boundary form has certain self-adaptability.
In summary, the method for monitoring the state of the rotating mechanical equipment based on the planarized data boundary morphology can effectively monitor the abnormal state of the mechanical equipment, and the method can visualize the monitoring process and reflect the running condition of the rotating mechanical equipment in real time.
Claims (9)
1. A method for monitoring the condition of rotating mechanical equipment based on the form of a planarized data boundary is characterized by comprising the following steps:
step 1: extracting characteristic parameters of sample points of the rotary mechanical equipment;
step 2: using a constellation diagram technology to map the characteristic parameters of the sample points onto a plane in a dimensionality reduction way;
and step 3: removing outliers in a distance-based mode;
and 4, step 4: calculating the boundary form of the plane point set, obtaining a boundary point set, and constructing a visual monitoring model;
and 5: extracting characteristic parameters of monitoring points of the rotary mechanical equipment;
step 6: inputting the characteristic parameters of the monitoring points into a visual monitoring model, and judging whether the characteristic parameters are abnormal or not; if the monitoring point is normal, updating the monitoring model by using a window sliding method; if the monitoring point is abnormal, alarming by using the monitoring model after the boundary form expansion;
in the step 6, the boundary form expansion is to obtain the distance between the flattened points of the adjacent data in the normal state of the equipment, expand the boundary by using the maximum likelihood estimation of the parameters, and multiply the horizontal and vertical coordinates of the primary boundary point set by the expansion proportion coefficient respectively to obtain a new boundary point set.
2. The method of claim 1, wherein the method further comprises the step of: the rotating mechanical equipment in the step 1 comprises a fan, a motor, a gear box, rotor equipment of a coupler and a bearing; the characteristic parameters of the sample points are derived from vibration signals of the rotary mechanical equipment, and the vibration signals comprise speed signals, acceleration signals and impact signals; and extracting speed signals or acceleration signals for rotor equipment of a fan, a coupling, a gear box and a motor, and extracting impact signals for a bearing.
3. The method of claim 1, wherein the method further comprises the step of: the sample point characteristic parameters in the step 1 and the monitoring point characteristic parameters in the step 5 comprise dimensional parameter indexes and dimensionless parameter indexes, the dimensional parameter indexes comprise standard deviation, variance, peak-peak value, square root amplitude value, average amplitude value and root-mean-square value, and the dimensionless parameter indexes comprise kurtosis indexes, peak value indexes, waveform indexes, pulse indexes and margin indexes.
4. The method of claim 1, wherein the method further comprises the step of: the basic principle of the constellation diagram technology in the step 2 is that N-dimensional features of data are fused into a point to be displayed in a semicircle, sample points closer to each other in the constellation diagram are similar, sample points farther from each other are different, and the positions of different types of data in the constellation diagram are different.
5. The method of claim 1, wherein the method further comprises the step of: and 3, eliminating outliers based on a distance mode, measuring the distance between the objects by defining a distance function, and determining the outliers in the data set according to the obvious deviation of the distances of most objects in the data.
6. The method of claim 1, wherein the method further comprises the step of: the boundary point set in step 4 is based on the principle that the distribution form of the boundary points of the plane point set replaces the distribution area of the state data.
7. The method of claim 1, wherein the method further comprises the step of: and 6, judging whether the monitoring points are abnormal or not according to the relative position relationship between the monitoring points and the boundary form.
8. The method of claim 1, wherein the method further comprises the step of: and 6, updating the model by using the boundary points of the original model and the model input data updating points by using the window sliding method so as to achieve the aim of improving the algorithm execution efficiency.
9. The method of claim 1, wherein the method further comprises the step of: in the step 6, in order to reduce false alarm when alarming, two characteristics of the boundary form and the expansion ratio coefficient after data planarization are integrated to serve as the alarming condition of the monitoring model, and the point is considered as an abnormal point only when the alarming condition is exceeded; in order to make the monitoring model have better visualization effect, the arrangement is made that the normal point is displayed as green, and the abnormal point is displayed as red.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910669452.1A CN110375974B (en) | 2019-07-24 | 2019-07-24 | Method for monitoring state of rotating mechanical equipment based on data boundary form after planarization |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910669452.1A CN110375974B (en) | 2019-07-24 | 2019-07-24 | Method for monitoring state of rotating mechanical equipment based on data boundary form after planarization |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110375974A CN110375974A (en) | 2019-10-25 |
| CN110375974B true CN110375974B (en) | 2020-06-02 |
Family
ID=68255295
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910669452.1A Active CN110375974B (en) | 2019-07-24 | 2019-07-24 | Method for monitoring state of rotating mechanical equipment based on data boundary form after planarization |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110375974B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113029567A (en) * | 2021-02-26 | 2021-06-25 | 上海电机学院 | Fault signal diagnosis method based on feature fusion equipotential planet map |
Family Cites Families (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101699359B (en) * | 2009-10-28 | 2012-01-04 | 上海理工大学 | Method for visualizing fault state monitoring |
| EP2835707B1 (en) * | 2013-08-05 | 2016-07-27 | ABB Technology AG | A method for condition monitoring of distributed drive-trains |
| CN103487275B (en) * | 2013-10-08 | 2016-04-06 | 中国矿业大学(北京) | Based on coal mine equipment state recognition and the method for early warning of two-dimensional projections |
| CN106323635B (en) * | 2016-08-04 | 2018-10-19 | 中国航空综合技术研究所 | A kind of rolling bearing fault on-line checking and state evaluating method |
| CN108182498A (en) * | 2018-01-15 | 2018-06-19 | 国网黑龙江省电力有限公司电力科学研究院 | The restorative reconstructing method of distribution network failure |
| CN109872511B (en) * | 2019-02-26 | 2021-07-06 | 西安交通大学 | Self-adaptive two-stage alarm method for monitoring axial displacement sudden change |
-
2019
- 2019-07-24 CN CN201910669452.1A patent/CN110375974B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN110375974A (en) | 2019-10-25 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110587377B (en) | Method for monitoring defects of milling cutter on line | |
| Karpuschewski et al. | Grinding monitoring system based on power and acoustic emission sensors | |
| CN104634571B (en) | A kind of Fault Diagnosis of Roller Bearings based on LCD MF | |
| CN108490880A (en) | A kind of numerical control machine tool cutting cutting-tool wear state method of real-time | |
| CN105867305A (en) | Method for real-time monitoring of numerical control machining states of complicated structural components based on machining features | |
| CN101879690B (en) | Batch drilling process quality monitoring method based on multiple sensor signals | |
| CN106514434A (en) | Milling tool abrasion monitoring method based on data | |
| Gao et al. | Chatter detection and stability region acquisition in thin-walled workpiece milling based on CMWT | |
| CN109500657B (en) | Vision-based broken cutter detection method and system | |
| US20190160619A1 (en) | Tool State Detection Method of Machine Tool and System Thereof | |
| CN110375974B (en) | Method for monitoring state of rotating mechanical equipment based on data boundary form after planarization | |
| Liu et al. | Real-time cutting tool state recognition approach based on machining features in NC machining process of complex structural parts | |
| CN103786093B (en) | What a kind of grinding status monitored in real time graphically detects display packing | |
| CN114473870B (en) | Grinding machine monitoring system and monitoring method | |
| CN110434676A (en) | A kind of boring monitoring chatter method of multisensor time-frequency characteristics fusion | |
| CN114896753A (en) | Method for identifying instantaneous friction energy consumption distribution characteristics of rear cutter face of high-energy-efficiency milling cutter | |
| CN103941645A (en) | Thin-wall part complex working condition machining state monitoring method | |
| Yao et al. | Real-time chatter detection and automatic suppression for intelligent spindles based on wavelet packet energy entropy and local outlier factor algorithm | |
| CN110346130A (en) | A kind of boring flutter detection method based on empirical mode decomposition and time-frequency multiple features | |
| Wolszczak et al. | Monitoring of cutting conditions with the empirical mode decomposition | |
| TWI670138B (en) | Method for predicting tool wear in an automatic processing machine | |
| JP7134062B2 (en) | Machine tool abnormality diagnosis system, abnormality diagnosis method, abnormality diagnosis program | |
| Zhong et al. | Correlation analysis of cutting force and acoustic emission signals for tool condition monitoring | |
| CN114508490A (en) | Intelligent split centrifugal pump, network fault point positioning method and readable storage module applied to centrifugal pump | |
| CN112380646B (en) | Method for researching turning temperature change and turning vibration coupling characteristics of different-abrasion cutters |
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 |