CN108062514A - A kind of ink roller of offset printing machine method for diagnosing faults based on three-dimensional spectrum analysis - Google Patents
A kind of ink roller of offset printing machine method for diagnosing faults based on three-dimensional spectrum analysis Download PDFInfo
- Publication number
- CN108062514A CN108062514A CN201711215783.5A CN201711215783A CN108062514A CN 108062514 A CN108062514 A CN 108062514A CN 201711215783 A CN201711215783 A CN 201711215783A CN 108062514 A CN108062514 A CN 108062514A
- Authority
- CN
- China
- Prior art keywords
- mrow
- signal
- msub
- printing machine
- feature
- 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.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/12—Classification; Matching
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01H—MEASUREMENT OF MECHANICAL VIBRATIONS OR ULTRASONIC, SONIC OR INFRASONIC WAVES
- G01H17/00—Measuring mechanical vibrations or ultrasonic, sonic or infrasonic waves, not provided for in the preceding groups
-
- 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
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
-
- 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
- G06N3/044—Recurrent networks, e.g. Hopfield networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/08—Feature extraction
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Signal Processing (AREA)
- Biophysics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Computer Hardware Design (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Biomedical Technology (AREA)
- Geometry (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Testing Of Devices, Machine Parts, Or Other Structures Thereof (AREA)
Abstract
The invention discloses a kind of ink roller of offset printing machine method for diagnosing faults based on three-dimensional spectrum analysis, specifically implement according to following steps:Step 1, the vibration signal under printing machine rubber roller bearings at both ends normal condition and malfunction is gathered, while is built and the corresponding emulation signal of fault-signal;Step 2, the vibration signal and corresponding emulation signal that collect are analyzed and handled, establish 3-D view;Step 3, feature extraction is carried out to the 3-D view of vibration signal using the method for gray level co-occurrence matrixes, and processing is weighted to the grain direction feature of gray level co-occurrence matrixes;Step 4, classification is identified to weighting processed feature using BP neural network respectively, the feature of different faults is recognized accurately.The method of the present invention effectively increases the correct probability of fault diagnosis, ensure that the work of printing machine normal table, has saved the plenty of time, production efficiency significantly improves.
Description
Technical field
The invention belongs to printing packaging equipment condition monitorings and intelligent fault diagnosis technical field, and in particular to one kind is based on
The ink roller of offset printing machine method for diagnosing faults of three-dimensional spectrum analysis.
Background technology
Printing industry is in occupation of critical role in the development of entire society, and printing is closely bound up with people’s lives, together
When press there is the double attribute of processing industry and cultural industry, be the important component of national economy.Modern society flies
Speed development, people are higher and higher for the requirement of various print qualities, and any machine in use all can be faulty
Occur, production efficiency will necessarily be reduced by breaking down, and the quality of product is caused to decline, causes serious economic loss.Work as failure
When serious, personal safety may be impacted.Therefore in actual production, failure is found in time, failure is solved, takes precautions against failure
Just seem particularly significant.
Complete ink system is to realize the important prerequisite of exquisite printing, and adjusting, maintenance, the maintenance of ink system are being printed
It is occupied an important position in machine operation.Offset press ink system is made of more than ten a ink rollers, including ink foundation roller, distributor rollers, rider roller,
Wavers, weight roller etc..One of the important spare part of rolling bearing as ink roller, the stability and reliability of operation are for offset printing
The working condition of machine plays decisive role.In the actual environment, ink roller bearing is not only subject to fountain solution, alcolhol burner organic solvent
Corrosion, while periodically movement roller is caused the abrasion of different faults occur.Ink roller bearings at both ends is in the operating condition
Forms of motion is synchronous, and bearings at both ends cycle, amplitude, movement locus are identical, has certain correlation;Multiple ink rollers are synchronous
Movement will be generated with frequency and frequency-doubled signal so that vibration signal is easily interfered with each other and coupled, and is increased fault signature and is carried
The difficulty taken;The relatively very noisy that multiple ink roller vibrations generate may flood bearing early-stage weak fault signal, only to single bearing
Analysis of vibration signal is easy to ignore fault signature;How to ink roller system bearing vibration signal rapid extraction and to have
The feature extraction of effect is significant for the status monitoring and fault diagnosis of printing machine inking system.
The major maintenance management method of equipment is exactly to be diagnosed according to vibration signal at present, for this kind of complexity of printing machine
Mechanical equipment for, analysis of vibration signal is undoubtedly maximally efficient means.Traditional vibration research spininess is to single zero
Time domain and the frequency domain character expansion of part vibration signal, comprehensively to analyze total, if can introduce 3-D view into
Row the Study on Fault is then more conducive to hold equipment state comprehensively, while has both the characteristics of directly perceived indirect.
The content of the invention
Object of the present invention is to provide a kind of ink roller of offset printing machine method for diagnosing faults based on three-dimensional spectrum analysis, are
The identification of ink system bearing abnormality provides accurate method directly perceived.
The technical solution adopted in the present invention is a kind of ink roller of offset printing machine fault diagnosis side based on three-dimensional spectrum analysis
Method is specifically implemented according to following steps:
Step 1, vibration signal of the printing machine rubber roller bearings at both ends under normal condition and malfunction is gathered, is built simultaneously
With the corresponding emulation signal of fault-signal;
Step 2, the vibration signal that collects and corresponding emulation signal are analyzed and handled, establish vibration signal and
Emulate the 3-D view of signal;
Step 3, feature extraction is carried out to the 3-D view of vibration signal using the method for gray level co-occurrence matrixes, and to gray scale
The grain direction feature of co-occurrence matrix is weighted processing;
Step 4, classification is identified to weighting processed feature using BP neural network respectively, difference is recognized accurately
The feature of failure, so as to complete the fault diagnosis of printing machine.
The features of the present invention also resides in:
Printing machine top roller bearing is rolling bearing in step 1;
Printing machine rubber roller bearings at both ends is gathered by high--speed multi--channel data acquisition system and acceleration transducer in step 1
Vibration signal under normal condition and malfunction;The emulation obtained by programming software under different types of faults state is believed
Number;
Malfunction in step 1 includes:In bearing inner race malfunction, outer ring malfunction, rolling element malfunction
One or more;
Step 2 specifically includes following steps:
Step 2.1, it is original to input vibration signal of the printing machine top roller bearing both ends of acquisition under normal and malfunction
Beginning signal x (t) finds the Local modulus maxima of original signal x (t) and the coenvelope with Cubic Spline Functions Fitting into former data
Line P, then find out all minimum points and it is used into cubic spline interpolation into lower envelope line Q;
Step 2.2, in calculating lower envelope average, be denoted as m (t):
Step 2.3, it is worth to one-component h (t) with what original signal x (t) subtracted envelope
H (t)=x (t)-m (t) (2)
Step 2.4, when component meet h (t) on time shaft Local Symmetric and | h (t) extreme point number-h (t) zero points
Number | during≤1 requirement, IMF components will be just saved as, are denoted as c (t);H (t) is expressed as inputting if being unsatisfactory for IMF definition
Signal repeats step 2.1~step 2.3, until meeting EMD requirements;
Step 2.5, c (t) is subtracted from original signal x (t), obtains residual signal r (t)
R (t)=x (t)-c (t) (3)
Step 2.6, step 2.1~step 2.5 is carried out using residual signal r (t) as original signal, until N ranks cannot
IMF extractions are carried out again, and most original signal is expressed as at last:
The Hilbert of x (t) converts y (t):
In formula (5), PV is Cauchy's principal value, and τ is time variable;
Analytic signal z (t) corresponding to x (t) is:
Z (t)=x (t)+iy (t)=a (t) eiθ(t) (6)
In formula (6), a (t) is known as the instantaneous amplitude of original signal x (t), and θ (t) is known as the instantaneous phase of original signal x (t)
Step 2.7, Hilbert conversion is carried out to every single order IMF of original signal x (t), obtains signal transient frequency:
Step 2.8, sampled by software programming, the vibration signal collected is converted respectively using HHT algorithms, is obtained
To corresponding amplitude over time-frequency 3-D view.
Step 3 specifically includes following steps:
Step 3.1, the three-dimensional time-frequency image of vibration signal is mapped on X-Y scheme, vibration is replaced with the light and shade of gray scale
Amplitude;
Step 3.2, the gray level co-occurrence matrixes of two-dimensional map are calculated, to four on 0 °, 45 °, 90 °, 135 ° of four directions
Gray feature extracts, and obtains the eigenmatrix of image;
Step 3.3, processing is weighted to the grain direction of eigenmatrix, to eliminate influence of the direction to diagnostic result,
The description characteristics of image of comprehensive and reasonable;
Step 3.3 concretely comprises the following steps,
Step 3.3.1 selects entropy to ask entropy flat on 0 °, 45 °, 90 °, 135 ° of four directions as fault signature
Average:
Step 3.3.2 calculates weight of the entropy on 0 °, 45 °, 90 °, 135 ° of four directions:
Step 3.3.3 combines the characteristic value of 0 °, 45 °, 90 °, 135 ° four direction according to weight factor, obtains at weighting
Feature vector after reason:
Step 3.3.4 passes through weighting treated feature vector drawing image characteristic statistics figure.
Step 4 specifically includes following steps:
Step 4.1, input layer unit number n is determined;
Step 4.2, output layer unit number m is determined;
Step 4.3, three layers of hidden layer BP neural network are chosen and carries out fault diagnosis.
The beneficial effects of the invention are as follows:
(1) present invention carries out EMD processing to bearing vibration signal using HHT methods and obtains IMF components, to IMF points
Amount carries out Hilbert conversion, obtains the Energy distribution graphics of vibration signal, compared with general two-dimensional time-domain frequency domain figure, shakes
Width-time-frequency three-dimensional spectrum more intuitively embodies signal frequency range and the distribution of energy, can observe amplitude simultaneously
At any time, the trend of frequency variation all has preferable temporal resolution simultaneously for low frequency, high-frequency signal, can be complete
The feature of original signal is shown, is laid the foundation for the feature extraction of 3-D view;
(2) the present invention is based on the gray scale depth is utilized to replace amplitude height that will vibrate graphics on the basis of gray level co-occurrence matrixes
As being mapped as two dimensional gray figure, and the vibrational image gray feature collected by different directions is weighted with weight factor
Processing, effectively eliminates influence of the direction to fault signature, so that fault signature cluster property and validity higher, but also
Fault signature extraction is more prone to;
(3) present invention innovatively proposes the method that ink roller bearings at both ends is carried out at the same time monitoring and analysis of vibration signal, by
The very noisy generated when printing machine works easily floods the vibration signal of rolling bearing, when an end bearing occurred early stage it is faint
It cannot still be found in time during fault signature;But it is compared by analyzing both ends bearing vibration signal, and it is three-dimensional using signal
Spectrogram finds exception, improves the precognition and recognition capability to early-stage weak fault, effectively accomplishes that failure early stage indicates;
(4) offset press based on three-dimensional spectrum analysis that the present invention is researched and proposed for the fault characteristic of top roller bearing system
Ink roller method for diagnosing faults is provided for new Research Thinking.Printing machine ink path system structure is complicated, multiple ink roller phase mutual connections
Transmission is touched, interacts and generates between each parts;The present invention is by putting forward ink roller bearings at both ends progress signal acquisition and feature
It takes, other interference signals is efficiently separated so as to obtain preferable fault signature, using neural network model to fault-signal feature
Quick and precisely classify, realize the efficient diagnosis of top roller bearing failure.
Description of the drawings
Fig. 1 is the printing machine fault diagnosis model of the method for the present invention.
Specific embodiment
The present invention is described in detail with reference to the accompanying drawings and detailed description.
The present invention is a kind of ink roller of offset printing machine method for diagnosing faults based on three-dimensional spectrum analysis, as shown in Figure 1, specifically pressing
Implement according to following steps:
Step 1, vibration signal of the printing machine rubber roller bearings at both ends under normal condition and malfunction is gathered, is built simultaneously
With the corresponding emulation signal of fault-signal;
Printing machine rubber roller bearings at both ends is gathered by high--speed multi--channel data acquisition system and acceleration transducer in step 1
Vibration signal under normal condition and malfunction;The emulation obtained by programming software under different types of faults state is believed
Number;Malfunction in step 1 includes:One kind in bearing inner race malfunction, outer ring malfunction, rolling element malfunction
It is or several;
Step 2, the vibration signal that collects and corresponding emulation signal are analyzed and handled, establish vibration signal and
Emulate the 3-D view of signal;
Step 2 specifically includes following steps:
Step 2.1, it is original to input vibration signal of the printing machine rubber roller bearings at both ends of acquisition under normal and malfunction
Beginning signal x (t) finds the Local modulus maxima of original signal x (t) and the coenvelope with Cubic Spline Functions Fitting into former data
Line P, then find out all minimum points and it is used into cubic spline interpolation into lower envelope line Q;
Step 2.2, in calculating lower envelope average, be denoted as m (t):
Step 2.3, it is worth to one-component h (t) with what original signal x (t) subtracted envelope
H (t)=x (t)-m (t) (2)
Step 2.4, when component meet h (t) on time shaft Local Symmetric and | h (t) extreme point number-h (t) zero points
Number | during≤1 requirement, IMF components will be just saved as, are denoted as c (t);H (t) is expressed as inputting if being unsatisfactory for IMF definition
Signal repeats step 2.1~step 2.3, until meeting EMD requirements;
Step 2.5, c (t) is subtracted from original signal x (t), obtains residual signal r (t)
R (t)=x (t)-c (t) (3)
Step 2.6, step 2.1~step 2.5 is carried out using residual signal r (t) as original signal, until n-th order cannot
IMF extractions are carried out again, and most original signal is expressed as at last:
The Hilbert of x (t) converts y (t):
In formula (5), PV is Cauchy's principal value, and τ is time variable;
Analytic signal z (t) corresponding to x (t) is:
Z (t)=x (t)+iy (t)=a (t) eiθ(t) (6)
In formula (6), a (t) is known as the instantaneous amplitude of original signal x (t), and θ (t) is known as the instantaneous phase of original signal x (t)
Step 2.7, Hilbert conversion is carried out to every single order IMF of original signal x (t), obtains signal transient frequency:
Step 2.8, the vibration signal and vibration signal that collect are converted respectively using HHT algorithms, is corresponded to
Amplitude over time-frequency 3-D view;
Step 3, feature extraction is carried out to the 3-D view of vibration signal using the method for gray level co-occurrence matrixes, and to gray scale
The grain direction feature of co-occurrence matrix is weighted processing;
Step 3 specifically includes following steps:
Step 3.1, the three-dimensional time-frequency image of vibration signal is mapped on X-Y scheme, vibration is replaced with the light and shade of gray scale
Amplitude;
Step 3.2, the gray level co-occurrence matrixes of two-dimensional map are calculated, four gray features on four direction are carried
It takes, obtains the eigenmatrix of image;
Step 3.3, processing is weighted to the grain direction of eigenmatrix, to eliminate influence of the direction to diagnostic result,
The description characteristics of image of comprehensive and reasonable;
Step 3.3 concretely comprises the following steps,
Step 3.3.1 selects entropy to ask entropy flat on 0 °, 45 °, 90 °, 135 ° of four directions as fault signature
Average:
Step 3.3.2 calculates weight of the entropy on 0 °, 45 °, 90 °, 135 ° of four directions:
Step 3.3.3 combines the characteristic value of 0 °, 45 °, 90 °, 135 ° four direction according to weight factor, obtains at weighting
Feature vector after reason:
Step 3.3.4 passes through weighting treated feature vector drawing image characteristic statistics figure;
Step 4, classification is identified to weighting processed feature using BP neural network respectively, difference is recognized accurately
The feature of failure, so as to complete the fault diagnosis of printing machine;
Step 4 specifically includes following steps:
Step 4.1, input layer unit number n is determined, since the present invention devises four kinds of states and obtains four gray features
Value is used as input, therefore n=4;
Step 4.2, determine output layer unit number m, ensure that each input sample has different vectors to correspond to, therefore m
=4;
Step 4.3, three layers of hidden layer BP neural network are chosen and carries out fault diagnosis.
Therefore, the present invention proposes a kind of ink roller of offset printing machine method for diagnosing faults based on three-dimensional spectrum analysis, by simultaneously
It gathers ink roller bearings at both ends vibration signal to carry out, when top roller bearing one end is broken down, three-dimensional characteristics of image will appear from different
Often, by comparing the three-dimensional spectral characteristic of bearings at both ends signal, the form of expression of abnormality is amplified, so as to beneficial to discovery early stage
Weak fault improves the timeliness and accuracy rate of fault diagnosis.
It is an advantage of the invention that:
(1) present invention carries out EMD processing to bearing vibration signal using HHT methods and obtains IMF components, to IMF points
Amount carries out Hilbert conversion, obtains the Energy distribution graphics of vibration signal, compared with general two-dimensional time-domain frequency domain figure, shakes
Width-time-frequency three-dimensional spectrum more intuitively embodies signal frequency range and the distribution of energy, can observe amplitude simultaneously
At any time, the trend of frequency variation all has preferable temporal resolution simultaneously for low frequency, high-frequency signal, can be complete
The feature of original signal is shown, is laid the foundation for the feature extraction of 3-D view;
(2) the present invention is based on the gray scale depth is utilized to replace amplitude height that will vibrate graphics on the basis of gray level co-occurrence matrixes
As being mapped as two dimensional gray figure, and the vibrational image gray feature collected by different directions is weighted with weight factor
Processing, effectively eliminates influence of the direction to fault signature, so that fault signature cluster property and validity higher, but also
Fault signature extraction is more prone to;
(3) present invention innovatively proposes the method that ink roller bearings at both ends is carried out at the same time monitoring and analysis of vibration signal, by
The very noisy generated when printing machine works easily floods the vibration signal of rolling bearing, when an end bearing occurred early stage it is faint
It cannot still be found in time during fault signature;But it is compared by analyzing both ends bearing vibration signal, and it is three-dimensional using signal
Spectrogram finds exception, improves the precognition and recognition capability to early-stage weak fault, effectively accomplishes that failure early stage indicates;
(4) offset press based on three-dimensional spectrum analysis that the present invention is researched and proposed for the fault characteristic of top roller bearing system
Ink roller method for diagnosing faults is provided for new Research Thinking.Printing machine ink path system structure is complicated, multiple ink roller phase mutual connections
Transmission is touched, interacts and generates between each parts;The present invention is by putting forward ink roller bearings at both ends progress signal acquisition and feature
It takes, other interference signals is efficiently separated so as to obtain preferable fault signature, using neural network model to fault-signal feature
Quick and precisely classify, realize the efficient diagnosis of top roller bearing failure.
Claims (6)
1. a kind of ink roller of offset printing machine method for diagnosing faults based on three-dimensional spectrum analysis, which is characterized in that specifically according to following step
It is rapid to implement:
Step 1, vibration signal of the printing machine rubber roller bearings at both ends under normal condition and malfunction is gathered, while is built and event
Hinder the corresponding emulation signal of signal;
Step 2, the vibration signal and corresponding emulation signal that collect are analyzed and handled, establish vibration signal and emulation
The 3-D view of signal;
Step 3, feature extraction is carried out to the 3-D view of vibration signal using the method for gray level co-occurrence matrixes, and to gray scale symbiosis
The grain direction feature of matrix is weighted processing;
Step 4, classification is identified to weighting processed feature using BP neural network respectively, different faults is recognized accurately
Feature, so as to complete the fault diagnosis of printing machine.
2. a kind of ink roller of offset printing machine method for diagnosing faults based on three-dimensional spectrum analysis according to claim 1, feature
It is, printing machine top roller bearing is rolling bearing in step 1;
Printing machine rubber roller bearings at both ends is gathered just by high--speed multi--channel data acquisition system and acceleration transducer in step 1
Vibration signal under normal state and malfunction;Emulation signal under different types of faults state is obtained by programming software;
Malfunction in step 1 includes:One in bearing inner race malfunction, outer ring malfunction, rolling element malfunction
Kind is several.
3. a kind of ink roller of offset printing machine method for diagnosing faults based on three-dimensional spectrum analysis according to claim 1, feature
It is, step 2 specifically includes following steps:
Step 2.1, it is original letter to input vibration signal of the printing machine top roller bearing both ends of acquisition under normal and malfunction
Number x (t), finds the Local modulus maxima of original signal x (t) and the coenvelope line P with Cubic Spline Functions Fitting into former data,
All minimum points are found out again and it is used into cubic spline interpolation into lower envelope line Q;
Step 2.2, in calculating lower envelope average, be denoted as m (t):
<mrow>
<mi>m</mi>
<mrow>
<mo>(</mo>
<mi>t</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<mfrac>
<mrow>
<mi>P</mi>
<mo>+</mo>
<mi>Q</mi>
</mrow>
<mn>2</mn>
</mfrac>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>1</mn>
<mo>)</mo>
</mrow>
</mrow>
Step 2.3, it is worth to one-component h (t) with what original signal x (t) subtracted envelope
H (t)=x (t)-m (t) (2)
Step 2.4, when component meet h (t) on time shaft Local Symmetric and | h (t) extreme point number-h (t) zero numbers |≤
During 1 requirement, IMF components will be just saved as, are denoted as c (t);H (t) is expressed as input signal if being unsatisfactory for IMF definition
Step 2.1~step 2.3 is repeated, until meeting EMD requirements;
Step 2.5, c (t) is subtracted from original signal x (t), obtains residual signal r (t)
R (t)=x (t)-c (t) (3)
Step 2.6, using residual signal r (t) as original signal carry out step 2.1~step 2.5, until N ranks cannot again into
Row IMF is extracted, and most original signal is expressed as at last:
<mrow>
<mi>x</mi>
<mrow>
<mo>(</mo>
<mi>t</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<munderover>
<mi>&Sigma;</mi>
<mrow>
<mi>j</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>n</mi>
</munderover>
<msub>
<mi>c</mi>
<mi>j</mi>
</msub>
<mrow>
<mo>(</mo>
<mi>t</mi>
<mo>)</mo>
</mrow>
<mo>+</mo>
<msub>
<mi>r</mi>
<mi>n</mi>
</msub>
<mrow>
<mo>(</mo>
<mi>t</mi>
<mo>)</mo>
</mrow>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>4</mn>
<mo>)</mo>
</mrow>
</mrow>
The Hilbert of x (t) converts y (t):
<mrow>
<mi>y</mi>
<mrow>
<mo>(</mo>
<mi>t</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<mfrac>
<mn>1</mn>
<mi>&pi;</mi>
</mfrac>
<mi>P</mi>
<mi>V</mi>
<mo>&Integral;</mo>
<mfrac>
<mrow>
<mi>x</mi>
<mrow>
<mo>(</mo>
<mi>&tau;</mi>
<mo>)</mo>
</mrow>
</mrow>
<mrow>
<mi>t</mi>
<mo>-</mo>
<mi>&tau;</mi>
</mrow>
</mfrac>
<mi>d</mi>
<mrow>
<mo>(</mo>
<mi>&tau;</mi>
<mo>)</mo>
</mrow>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>5</mn>
<mo>)</mo>
</mrow>
</mrow>
In formula (5), PV is Cauchy's principal value, and τ is time variable;
Analytic signal z (t) corresponding to x (t) is:
Z (t)=x (t)+iy (t)=a (t) eiθ(t) (6)
In formula (6), a (t) is known as the instantaneous amplitude of original signal x (t), and θ (t) is known as the instantaneous phase of original signal x (t)
<mrow>
<mi>a</mi>
<mrow>
<mo>(</mo>
<mi>t</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<msqrt>
<mrow>
<msup>
<mi>x</mi>
<mn>2</mn>
</msup>
<mrow>
<mo>(</mo>
<mi>t</mi>
<mo>)</mo>
</mrow>
<mo>+</mo>
<msup>
<mi>y</mi>
<mn>2</mn>
</msup>
<mrow>
<mo>(</mo>
<mi>t</mi>
<mo>)</mo>
</mrow>
</mrow>
</msqrt>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>7</mn>
<mo>)</mo>
</mrow>
</mrow>
<mrow>
<mi>&theta;</mi>
<mrow>
<mo>(</mo>
<mi>t</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<mi>arctan</mi>
<mo>&lsqb;</mo>
<mfrac>
<mrow>
<mi>y</mi>
<mrow>
<mo>(</mo>
<mi>t</mi>
<mo>)</mo>
</mrow>
</mrow>
<mrow>
<mi>x</mi>
<mrow>
<mo>(</mo>
<mi>t</mi>
<mo>)</mo>
</mrow>
</mrow>
</mfrac>
<mo>&rsqb;</mo>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>8</mn>
<mo>)</mo>
</mrow>
</mrow>
Step 2.7, Hilbert conversion is carried out to every single order IMF of original signal x (t), obtains signal transient frequency:
<mrow>
<mi>&omega;</mi>
<mrow>
<mo>(</mo>
<mi>t</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<mi>R</mi>
<mo>{</mo>
<munderover>
<mi>&Sigma;</mi>
<mrow>
<mi>i</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>n</mi>
</munderover>
<msub>
<mi>a</mi>
<mi>i</mi>
</msub>
<mrow>
<mo>(</mo>
<mi>t</mi>
<mo>)</mo>
</mrow>
<msup>
<mi>e</mi>
<mrow>
<mi>i</mi>
<mo>&Integral;</mo>
<msub>
<mi>&omega;</mi>
<mi>i</mi>
</msub>
<mrow>
<mo>(</mo>
<mi>t</mi>
<mo>)</mo>
</mrow>
<mi>d</mi>
<mi>t</mi>
</mrow>
</msup>
<mo>}</mo>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>9</mn>
<mo>)</mo>
</mrow>
</mrow>
Step 2.8, sampled by software programming, the vibration signal collected is converted respectively using HHT algorithms, obtained pair
The amplitude over time answered-frequency 3-D view.
4. a kind of ink roller of offset printing machine method for diagnosing faults based on three-dimensional spectrum analysis according to claim 1, feature
It is, step 3 specifically includes following steps:
Step 3.1, the three-dimensional time-frequency image of vibration signal is mapped on X-Y scheme, the width of vibration is replaced with the light and shade of gray scale
Value;
Step 3.2, the gray level co-occurrence matrixes of two-dimensional map are calculated, to four gray scales on 0 °, 45 °, 90 °, 135 ° of four directions
Feature extracts, and obtains the eigenmatrix of image;
Step 3.3, processing is weighted to the grain direction of eigenmatrix, to eliminate influence of the direction to diagnostic result, comprehensively
Rational description characteristics of image.
5. a kind of ink roller of offset printing machine method for diagnosing faults based on three-dimensional spectrum analysis according to claim 4, feature
It is, step 3.3 concretely comprises the following steps,
Step 3.3.1 selects entropy to seek average value of the entropy on 0 °, 45 °, 90 °, 135 ° of four directions as fault signature:
Step 3.3.2 calculates weight of the entropy on 0 °, 45 °, 90 °, 135 ° of four directions:
<mrow>
<msub>
<mi>kk</mi>
<mi>i</mi>
</msub>
<mo>=</mo>
<mfrac>
<msub>
<mi>M</mi>
<msub>
<mi>E</mi>
<mi>i</mi>
</msub>
</msub>
<mrow>
<munder>
<mo>&Sigma;</mo>
<mi>i</mi>
</munder>
<msub>
<mi>M</mi>
<msub>
<mi>E</mi>
<mi>i</mi>
</msub>
</msub>
</mrow>
</mfrac>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>10</mn>
<mo>)</mo>
</mrow>
</mrow>
Step 3.3.3 combines the characteristic value of 0 °, 45 °, 90 °, 135 ° four direction according to weight factor, after obtaining weighting processing
Feature vector:
<mrow>
<msup>
<mi>E</mi>
<mo>&prime;</mo>
</msup>
<mo>=</mo>
<munder>
<mo>&Sigma;</mo>
<mi>i</mi>
</munder>
<msub>
<mi>kk</mi>
<mi>i</mi>
</msub>
<mo>&times;</mo>
<msub>
<mi>E</mi>
<mi>i</mi>
</msub>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>11</mn>
<mo>)</mo>
</mrow>
</mrow>
Step 3.3.4 passes through weighting treated feature vector drawing image characteristic statistics figure.
6. a kind of ink roller of offset printing machine method for diagnosing faults based on three-dimensional spectrum analysis according to claim 1, feature
It is, step 4 specifically includes following steps:
Step 4.1, input layer unit number n is determined;
Step 4.2, output layer unit number m is determined;
Step 4.3, three layers of hidden layer BP neural network are chosen and carries out fault diagnosis.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201711215783.5A CN108062514A (en) | 2017-11-28 | 2017-11-28 | A kind of ink roller of offset printing machine method for diagnosing faults based on three-dimensional spectrum analysis |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201711215783.5A CN108062514A (en) | 2017-11-28 | 2017-11-28 | A kind of ink roller of offset printing machine method for diagnosing faults based on three-dimensional spectrum analysis |
Publications (1)
Publication Number | Publication Date |
---|---|
CN108062514A true CN108062514A (en) | 2018-05-22 |
Family
ID=62135785
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201711215783.5A Pending CN108062514A (en) | 2017-11-28 | 2017-11-28 | A kind of ink roller of offset printing machine method for diagnosing faults based on three-dimensional spectrum analysis |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108062514A (en) |
Cited By (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109272017A (en) * | 2018-08-08 | 2019-01-25 | 太原理工大学 | The vibration signal mode identification method and system of distributed fiberoptic sensor |
CN109632081A (en) * | 2018-11-23 | 2019-04-16 | 积成电子股份有限公司 | Vibration of wind generating set feature extraction and otherness sentence method for distinguishing |
CN110161330A (en) * | 2019-05-10 | 2019-08-23 | 广东石油化工学院 | The vibration sound detection method and device of running state of transformer based on grey topology degree |
CN110516305A (en) * | 2019-07-26 | 2019-11-29 | 西安交通大学 | Intelligent fault diagnosis method under small sample based on attention mechanism meta-learning model |
CN110597212A (en) * | 2019-09-27 | 2019-12-20 | 红云红河烟草(集团)有限责任公司 | Method for fault early warning of wrapping equipment based on equipment vibration imaging diagnosis |
CN111401136A (en) * | 2020-02-22 | 2020-07-10 | 上海交通大学 | Plunger pump cavitation degree detection method and device and terminal |
CN112414715A (en) * | 2020-11-05 | 2021-02-26 | 西安工程大学 | Bearing fault diagnosis method based on mixed feature and improved gray level co-occurrence algorithm |
CN112629647A (en) * | 2020-11-24 | 2021-04-09 | 同济大学 | Real-time identification, monitoring and early warning method for vortex vibration event of large-span suspension bridge |
CN113156832A (en) * | 2020-10-23 | 2021-07-23 | 齐鲁工业大学 | Method and system for rapidly analyzing performance parameters of ink path system of offset press |
WO2023119632A1 (en) * | 2021-12-24 | 2023-06-29 | 富士通フロンテック株式会社 | Deterioration-determining system and deterioration-determining method |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103234748A (en) * | 2013-04-02 | 2013-08-07 | 北京工业大学 | Klingelnberg bevel gear fault diagnosis method based on sensitive IMF (instinct mode function) components |
CN106053074A (en) * | 2016-08-02 | 2016-10-26 | 北京航空航天大学 | Rolling bearing sound signal fault feature extraction method based on STFT and rotation inertia entropy |
CN106441888A (en) * | 2016-09-07 | 2017-02-22 | 广西大学 | High-speed train rolling bearing fault diagnosis method |
-
2017
- 2017-11-28 CN CN201711215783.5A patent/CN108062514A/en active Pending
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103234748A (en) * | 2013-04-02 | 2013-08-07 | 北京工业大学 | Klingelnberg bevel gear fault diagnosis method based on sensitive IMF (instinct mode function) components |
CN106053074A (en) * | 2016-08-02 | 2016-10-26 | 北京航空航天大学 | Rolling bearing sound signal fault feature extraction method based on STFT and rotation inertia entropy |
CN106441888A (en) * | 2016-09-07 | 2017-02-22 | 广西大学 | High-speed train rolling bearing fault diagnosis method |
Non-Patent Citations (1)
Title |
---|
丁继业: "基于三维谱图分析的印刷机胶辊轴承系统故障诊断研究", 《万方学位论文》 * |
Cited By (14)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109272017B (en) * | 2018-08-08 | 2022-07-12 | 太原理工大学 | Vibration signal mode identification method and system of distributed optical fiber sensor |
CN109272017A (en) * | 2018-08-08 | 2019-01-25 | 太原理工大学 | The vibration signal mode identification method and system of distributed fiberoptic sensor |
CN109632081A (en) * | 2018-11-23 | 2019-04-16 | 积成电子股份有限公司 | Vibration of wind generating set feature extraction and otherness sentence method for distinguishing |
CN110161330A (en) * | 2019-05-10 | 2019-08-23 | 广东石油化工学院 | The vibration sound detection method and device of running state of transformer based on grey topology degree |
CN110516305A (en) * | 2019-07-26 | 2019-11-29 | 西安交通大学 | Intelligent fault diagnosis method under small sample based on attention mechanism meta-learning model |
CN110597212A (en) * | 2019-09-27 | 2019-12-20 | 红云红河烟草(集团)有限责任公司 | Method for fault early warning of wrapping equipment based on equipment vibration imaging diagnosis |
CN111401136A (en) * | 2020-02-22 | 2020-07-10 | 上海交通大学 | Plunger pump cavitation degree detection method and device and terminal |
CN113156832B (en) * | 2020-10-23 | 2022-09-23 | 齐鲁工业大学 | Method and system for rapidly analyzing performance parameters of ink path system of offset press |
CN113156832A (en) * | 2020-10-23 | 2021-07-23 | 齐鲁工业大学 | Method and system for rapidly analyzing performance parameters of ink path system of offset press |
CN112414715A (en) * | 2020-11-05 | 2021-02-26 | 西安工程大学 | Bearing fault diagnosis method based on mixed feature and improved gray level co-occurrence algorithm |
CN112414715B (en) * | 2020-11-05 | 2022-09-27 | 西安工程大学 | Bearing fault diagnosis method based on mixed feature and improved gray level symbiosis algorithm |
CN112629647B (en) * | 2020-11-24 | 2022-04-08 | 同济大学 | Real-time identification, monitoring and early warning method for vortex vibration event of large-span suspension bridge |
CN112629647A (en) * | 2020-11-24 | 2021-04-09 | 同济大学 | Real-time identification, monitoring and early warning method for vortex vibration event of large-span suspension bridge |
WO2023119632A1 (en) * | 2021-12-24 | 2023-06-29 | 富士通フロンテック株式会社 | Deterioration-determining system and deterioration-determining method |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108062514A (en) | A kind of ink roller of offset printing machine method for diagnosing faults based on three-dimensional spectrum analysis | |
CN100485342C (en) | Integrated supporting vector machine mixed intelligent diagnosing method for mechanical fault | |
Hasan et al. | Acoustic spectral imaging and transfer learning for reliable bearing fault diagnosis under variable speed conditions | |
CN108181107B (en) | The Wind turbines bearing mechanical method for diagnosing faults of meter and more class objects | |
Cui et al. | Quantitative trend fault diagnosis of a rolling bearing based on Sparsogram and Lempel-Ziv | |
CN104729853B (en) | A kind of rolling bearing performance degradation assessment device and method | |
CN110110768B (en) | Rolling bearing fault diagnosis method based on parallel feature learning and multiple classifiers | |
CN104155108B (en) | A kind of Fault Diagnosis of Roller Bearings based on vibration time frequency analysis | |
CN108896296A (en) | A kind of wind turbine gearbox method for diagnosing faults based on convolutional neural networks | |
CN104697767B (en) | Rotor system fault diagnosis method and device based on vibration analysis | |
CN109782603A (en) | The detection method and monitoring system of rotating machinery coupling fault | |
CN109934206A (en) | A kind of rotary machinery fault diagnosis method under non-stationary operating condition | |
CN108171263A (en) | Based on the Fault Diagnosis of Roller Bearings for improving variation mode decomposition and extreme learning machine | |
CN110068462A (en) | A kind of motor bearings method for diagnosing faults and device | |
CN106092565A (en) | A kind of Vibration Analysis method and system thereof | |
CN108827605A (en) | A kind of mechanical breakdown characteristic automatic extraction method based on improvement sparseness filtering | |
CN113128338B (en) | Intelligent diagnosis method for printer roller faults under small sample | |
CN102797671A (en) | Fault detection method and device of reciprocating compressor | |
CN102054179A (en) | Online state monitoring and fault diagnosis device and method for rotary machine | |
CN106596149A (en) | Method for monitoring and diagnosing flat wheel state of rail vehicle based on VMD | |
CN103033359A (en) | Wind turbine generator main gear fault diagnosis method of multi-feature multilevel comprehensive judgment | |
CN103018043A (en) | Fault diagnosis method of variable-speed bearing | |
CN107908812A (en) | A kind of valve fault diagnosis method based on HHT and neutral net | |
CN103471849A (en) | Bearing fault diagnosis system of multi-layer relevance vector machine on basis of dual combination | |
CN108520125A (en) | A kind of method and system of prediction cutting-tool wear state |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20180522 |