CN115184016A - Elevator bearing fault detection method - Google Patents

Elevator bearing fault detection method Download PDF

Info

Publication number
CN115184016A
CN115184016A CN202211081144.5A CN202211081144A CN115184016A CN 115184016 A CN115184016 A CN 115184016A CN 202211081144 A CN202211081144 A CN 202211081144A CN 115184016 A CN115184016 A CN 115184016A
Authority
CN
China
Prior art keywords
vibration signal
sub
points
peak
segment
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
Application number
CN202211081144.5A
Other languages
Chinese (zh)
Inventor
江冰倩
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Jiangsu Dongkong Automation Technology Co ltd
Original Assignee
Jiangsu Dongkong Automation Technology Co ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Jiangsu Dongkong Automation Technology Co ltd filed Critical Jiangsu Dongkong Automation Technology Co ltd
Priority to CN202211081144.5A priority Critical patent/CN115184016A/en
Publication of CN115184016A publication Critical patent/CN115184016A/en
Pending legal-status Critical Current

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
    • G01M13/04—Bearings
    • G01M13/045—Acoustic or vibration analysis
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
    • G06F17/10—Complex mathematical operations

Abstract

The invention relates to the field of data compression, in particular to a method for detecting a fault of a bearing of an elevator, which comprises the following steps: acquiring the period of a vibration signal, wherein the period divides the vibration signal into a plurality of sub-vibration signal segments; obtaining an effective peak point and an effective valley point of each sub-vibration signal segment; respectively obtaining the smear degree of each sub-vibration signal section according to the distance between the effective peak/valley points of each sub-vibration signal section, and obtaining the importance of each sub-vibration signal section according to the amplitude of the effective peak/valley points; obtaining a self-adaptive bit block of each sub-vibration signal segment according to the importance and the smear degree, carrying out DACs (digital addressable coding) coding on each sub-vibration signal segment to obtain a coded vibration signal, and transmitting the coded vibration signal to an analysis server to decode the coded vibration signal to obtain a decoded vibration signal; and outputting the fault state of the bearing by taking the decoded vibration signal as the input of the neural network. The invention realizes the real-time and rapid detection of the bearing fault of the elevator.

Description

Elevator bearing fault detection method
Technical Field
The invention relates to the field of data compression, in particular to a method for detecting a fault of a bearing of a lifter.
Background
With the development of science and technology, the application of high-precision industrial production equipment is more and more common, and the frequency of faults and the maintenance cost of the equipment are higher and higher. An elevator is an important mechanical device in the industrial neighborhood, and the bearing failure of the elevator is a main failure in the operation process of the elevator. Therefore, it is important to detect a bearing failure of the elevator.
The elevator bearing fault diagnosis is based on bearing vibration frequency, extracts abnormal signal characteristics in the bearing operation process by accurately acquiring vibration signals of the bearing and detecting the frequency reflecting the bearing state in the vibration signals, and transmits the abnormal signal characteristics to a terminal for analysis, thereby judging the fault occurrence reason. The traditional method for transmitting the vibration signals adopts a DACs algorithm for compression transmission, however, the traditional DACs algorithm needs to manually set the size of a bit block for compression, and does not consider the actual vibration signals, if the size of the bit block is set to be too large, the reading speed can be ensured to be high, but the excessive bit-filling data can cause data redundancy, and the compression ratio is low; if the bit block sets up the undersize, can guarantee that the compression ratio is great, but can lead to reading speed slower, consequently comparatively important to the setting of bit block size, the bit block sets up too big or undersize, can lead to can't guarantee the speed of reading trouble data when guaranteeing the compression ratio, finally can't be to the real-time short-term test of trouble, consequently, await urgent need a lift bearing fault detection method to solve the current problem that can't be to the real-time short-term test of trouble.
Disclosure of Invention
The invention provides a method for detecting faults of a bearing of an elevator, which aims to solve the existing problems.
The invention discloses a method for detecting faults of a bearing of a lifter, which adopts the following technical scheme: the method comprises the following steps:
s1, collecting vibration signals of a bearing in a working state of an elevator, setting different time intervals for the vibration signals, acquiring the similarity of the amplitude values of corresponding signal points at the interval points of each time interval, obtaining the period of the vibration signals by using the time interval corresponding to the maximum similarity, and dividing the vibration signals into a plurality of sub-vibration signal sections according to the period of the vibration signals;
s2, obtaining amplitudes of all signal points in the sub-vibration signal section, obtaining peak values and valley values of the sub-vibration signal section according to the amplitudes of all the signal points in the sub-vibration signal section, and obtaining effective peak value points and effective valley value points of the sub-vibration signal section by using the obtained peak values and valley values and corresponding threshold values;
s3, obtaining the smear degree of the corresponding sub-vibration signal section according to the distance between the effective peak points and the distance between the effective valley points of each sub-vibration signal section, and obtaining the importance degree of the corresponding sub-vibration signal section according to the amplitude values of the effective peak points and the effective valley points of the sub-vibration signal section;
and S4, obtaining the size of a bit block of the corresponding sub-vibration signal section according to the importance and the smear degree of the sub-vibration signal section, carrying out DACs (digital audio coding) on the corresponding sub-vibration signal section according to the size of the bit block to obtain a coded vibration signal, transmitting the coded vibration signal to an analysis server, decoding the coded vibration signal by using the analysis server to obtain a decoded vibration signal, and obtaining the fault state of the bearing according to the decoded vibration signal.
Further, the period of the vibration signal is determined as follows:
acquiring all time intervals corresponding to the maximum similarity;
acquiring maximum value points corresponding to all time intervals;
and acquiring the mean value of the abscissa distances of the adjacent maximum value points, and taking the obtained mean value as the period of the vibration signal.
Further, the expression of the similarity of the signal amplitude corresponding to each time interval point is as follows:
Figure DEST_PATH_IMAGE002
in the formula:
Figure DEST_PATH_IMAGE004
representing the similarity of the amplitudes of the corresponding signal points at the interval points of each time interval,
Figure DEST_PATH_IMAGE006
representing time interval points
Figure DEST_PATH_IMAGE008
Of the amplitude of the vibration signal of (a),
Figure DEST_PATH_IMAGE010
representing time interval points
Figure DEST_PATH_IMAGE012
Of the amplitude of the vibration signal of (a),
Figure DEST_PATH_IMAGE014
the time interval is represented by the time interval,
Figure DEST_PATH_IMAGE016
representing the average of the amplitudes of all signal points of the vibration signal,
Figure DEST_PATH_IMAGE018
representing the length of the time segment of the vibration signal.
Further, the method for obtaining the peak value and the valley value of the sub-vibration signal segment is as follows:
if the amplitudes of two adjacent points of the signal points in the sub-vibration signal section are smaller than the amplitude of the signal point, the signal point is a peak point;
if the amplitudes of two adjacent points of the signal points in the sub-vibration signal section are larger than the amplitude of the signal point, the signal point is a wave valley point;
and taking the amplitudes corresponding to the wave peak point and the wave valley point as the peak value and the valley value of the sub-vibration signal segment.
Further, the specific step of obtaining the effective peak point and the effective valley point of the sub-vibration signal segment by using the obtained peak value and the valley value and the corresponding threshold value includes:
acquiring a first peak point of the sub-vibration signal segment;
acquiring an abscissa of a position corresponding to a first peak point of the sub-vibration signal segment in the sub-vibration signal segment of the next period adjacent to the sub-vibration signal segment;
acquiring the range of the abscissa after the addition and subtraction of the abscissa of the corresponding position by half a period;
obtaining the amplitudes of all signal points in the range of the abscissa, and calculating to obtain the average value of the amplitudes of all the signal points;
and if the difference value is smaller than the threshold value, the first peak point of the sub-vibration signal segment is the effective peak point of the sub-vibration signal segment, and in the same way, the effective valley point of the sub-vibration signal segment is obtained.
Further, the specific expression of the smear degree of each sub-vibration signal segment is as follows:
Figure DEST_PATH_IMAGE020
in the formula:
Figure DEST_PATH_IMAGE022
denotes the first
Figure DEST_PATH_IMAGE024
The extent of smearing of the segment vibration signal segments,
Figure DEST_PATH_IMAGE026
is shown as
Figure 982266DEST_PATH_IMAGE024
The number of the effective peak points of the segment vibration signal segment,
Figure DEST_PATH_IMAGE028
denotes the first
Figure 227302DEST_PATH_IMAGE024
Sub-vibration signal section one
Figure DEST_PATH_IMAGE030
Effective peak/valley point and
Figure DEST_PATH_IMAGE032
the distance of the abscissa between the effective peak points/effective valley points,
Figure DEST_PATH_IMAGE034
is shown as
Figure 760701DEST_PATH_IMAGE024
In the sub-vibration signal section
Figure 483807DEST_PATH_IMAGE032
The abscissa of the point of the effective peak,
Figure DEST_PATH_IMAGE036
is shown as
Figure 924278DEST_PATH_IMAGE024
In the sub-vibration signal section
Figure 809057DEST_PATH_IMAGE030
The abscissa of each valid peak point.
Further, the expression of the importance of each sub-vibration signal segment is:
Figure DEST_PATH_IMAGE038
in the formula:
Figure 89472DEST_PATH_IMAGE022
is shown as
Figure 350689DEST_PATH_IMAGE024
The extent of smearing of the segment vibration signal segments,
Figure 848929DEST_PATH_IMAGE026
is shown as
Figure 904610DEST_PATH_IMAGE024
The number of effective peak points/effective valley points of the segment vibration signal segment,
Figure 721256DEST_PATH_IMAGE028
is shown as
Figure 723847DEST_PATH_IMAGE024
In the sub-vibration signal section
Figure 802225DEST_PATH_IMAGE030
Effective peak/valley point and
Figure 294386DEST_PATH_IMAGE032
a significant peakThe distance on the abscissa between the value point/the effective valley point,
Figure 598329DEST_PATH_IMAGE034
is shown as
Figure 201348DEST_PATH_IMAGE024
In the sub-vibration signal section
Figure 408601DEST_PATH_IMAGE032
The abscissa of the individual effective peak/effective valley points,
Figure 806084DEST_PATH_IMAGE036
is shown as
Figure 597323DEST_PATH_IMAGE024
In the sub-vibration signal section
Figure 738454DEST_PATH_IMAGE030
The abscissa of each effective peak/effective valley point.
Further, a specific expression of the adaptive bit block of each sub-vibration signal segment is as follows:
Figure DEST_PATH_IMAGE040
in the formula:
Figure DEST_PATH_IMAGE042
is as follows
Figure 813595DEST_PATH_IMAGE024
The initial block of bits of the sub-vibration signal segment,
Figure 319663DEST_PATH_IMAGE022
is shown as
Figure 660514DEST_PATH_IMAGE024
The extent of smearing of the segment vibration signal segments,
Figure DEST_PATH_IMAGE044
is shown as
Figure 339757DEST_PATH_IMAGE024
The importance of a sub-vibration signal segment,
Figure DEST_PATH_IMAGE046
is as follows
Figure 649166DEST_PATH_IMAGE024
The over-parameters of the sub-vibration signal segments,
Figure DEST_PATH_IMAGE048
is shown as
Figure 217813DEST_PATH_IMAGE024
A block of bits of a sub-vibration signal segment.
Further, the method for obtaining the fault state of the bearing according to the decoded vibration signal comprises the following steps:
acquiring a kurtosis factor, a peak factor and a pulse factor of a vibration signal;
using kurtosis factors, peak factors and pulse factors of vibration signals as input of a neural network, using fault states of a bearing as output of the neural network, and training the neural network to obtain a trained neural network;
acquiring a kurtosis factor, a peak factor and a pulse factor of the decoded vibration signal;
and taking the kurtosis factor, the peak factor and the pulse factor of the decoded vibration signal as the input of the trained neural network, and outputting the fault state of the bearing.
The invention has the beneficial effects that: the invention relates to a method for detecting faults of an elevator bearing, which comprises the steps of dividing a vibration signal into a plurality of sub vibration signal segments by obtaining the period of the vibration signal, calculating the smear degree and the importance degree of each sub vibration signal segment, and obtaining a self-adaptive bit block of DACs codes of each sub vibration signal segment according to the smear degree and the importance degree of each sub vibration signal segment, thereby achieving the purpose of ensuring the speed of reading fault data while ensuring the compression ratio, and finally utilizing a neural network to achieve the real-time and rapid detection of faults.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the drawings without creative efforts.
FIG. 1 is a flow chart of an embodiment of a method of elevator bearing fault detection of the present invention;
FIG. 2 is a graph of vibration signals for a method of elevator bearing fault detection in accordance with the present invention;
FIG. 3 is an exemplary diagram of the detailed process of DACs encoding in an elevator bearing fault detection method of the present invention;
fig. 4 is a diagram illustrating a specific process of DACs decoding in an elevator bearing fault detection method according to the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
An embodiment of a method for detecting a fault of a bearing of an elevator according to the present invention, as shown in fig. 1, includes:
s1, collecting vibration signals of a bearing in the working state of an elevator, setting different time intervals for the vibration signals, obtaining the similarity of the amplitude values of corresponding signal points at the interval points of each time interval, obtaining the period of the vibration signals by using the time interval corresponding to the maximum similarity, and dividing the vibration signals into a plurality of sub vibration signal sections according to the period of the vibration signals.
The elevator bearing vibration data acquisition platform is arranged and used for acquiring bearing vibration data conditions in the working state of an elevator, adopts a piezoelectric acceleration sensor with a magnetic base as a device for acquiring vibration signals, and is arranged in the 6 o' clock direction of an outer ring of a bearing.
The acquisition platform arranged in the invention comprises a driving part, namely a servo motor, a speed reducer and a coupling, a supporting part, namely a complete rolling bearing and a bearing seat, a bearing part, namely a bearing for acquiring data, an acquisition sensor and a data transmission system part.
The acquired vibration signal data are transmitted to a processor to analyze the acquired vibration signal, and the signal of a rolling bearing of the elevator is in periodic change in the running process, as shown in fig. 2, if an outer ring, an inner ring or a rolling body, namely a ball, of the rolling bearing of the elevator fails, abnormality also occurs in the periodic vibration signal, so that the period of the vibration signal needs to be acquired first to extract the fault characteristics, and then the vibration signal in each period is analyzed, wherein the step of acquiring the period of the vibration signal is as follows:
carrying out smooth denoising on the collected vibration signals to obtain denoised vibration signals, setting time intervals, setting initial time interval points, calculating the similarity of signal point amplitudes corresponding to each time interval point according to the signal point amplitudes corresponding to the time interval points, wherein the specific expression is as follows:
Figure DEST_PATH_IMAGE002A
in the formula:
Figure 108277DEST_PATH_IMAGE004
representing the similarity of the signal amplitude corresponding to each time interval point,
Figure 89746DEST_PATH_IMAGE006
indicating the point of the initial time interval
Figure 624633DEST_PATH_IMAGE008
Of the amplitude of the vibration signal of (a),
Figure 800399DEST_PATH_IMAGE010
representing time interval points
Figure 991209DEST_PATH_IMAGE012
Of the amplitude of the vibration signal of (a),
Figure 779299DEST_PATH_IMAGE014
the time interval is represented by a time interval,
Figure 168692DEST_PATH_IMAGE016
representing the average of the amplitudes of all signal points of the acquired vibration signal,
Figure 515360DEST_PATH_IMAGE018
representing the length of the time period of the acquired vibration signal.
Calculating the similarity of the signal amplitude corresponding to each time interval point to obtain all the time intervals corresponding to the maximum similarity, and when the similarity is maximum, indicating that the corresponding time interval point is a similarity function
Figure 990203DEST_PATH_IMAGE004
The maximum value point, and the distance between the adjacent maximum value points represents the period of the vibration signal, therefore, the mean value of the abscissa distance of the adjacent maximum value points is obtained, the mean value is the period of the vibration signal, and the specific expression is as follows:
Figure DEST_PATH_IMAGE050
in the formula:
Figure DEST_PATH_IMAGE052
representing similarity functions
Figure 612944DEST_PATH_IMAGE004
To (1) a
Figure DEST_PATH_IMAGE054
The maximum value point of the number of pixels,
Figure DEST_PATH_IMAGE056
representing similarity functions
Figure 247056DEST_PATH_IMAGE004
To (1) a
Figure DEST_PATH_IMAGE058
The maximum value point of the number of the pixels,
Figure DEST_PATH_IMAGE060
representing similarity functions
Figure 325477DEST_PATH_IMAGE004
The number of the maximum value points of (a),
Figure DEST_PATH_IMAGE062
representing the period of the vibration signal.
Thus, the period of the collected vibration signal is obtained
Figure 851399DEST_PATH_IMAGE062
And according to the period
Figure 479826DEST_PATH_IMAGE062
And dividing the vibration signal into a plurality of sub vibration signal segments.
S2, obtaining amplitudes of all signal points in the sub-vibration signal section, obtaining peak values and valley values of the sub-vibration signal section according to the amplitudes of all the signal points in the sub-vibration signal section, and obtaining effective peak value points and effective valley value points of the sub-vibration signal section by utilizing the obtained peak values and valley values and corresponding threshold values.
And S3, obtaining the smear degree of the corresponding sub-vibration signal section according to the distance between the effective peak points and the distance between the effective valley points of each sub-vibration signal section, and obtaining the importance degree of the corresponding sub-vibration signal section according to the amplitude values of the effective peak points and the effective valley points of the sub-vibration signal section.
Different fault causes correspond to different waveforms. For the inner ring fault, the vibration amplitude of the inner ring fault is obviously increased, and an amplitude modulation phenomenon can be generated; for the outer ring fault, the vibration amplitude of the outer ring fault is obviously increased, and obvious pulse impact occurs; for the rolling element fault, the vibration amplitude of the rolling element is increased less obviously, and an amplitude modulation phenomenon is generated. And acquiring the smear degree and the importance degree in each sub-vibration signal segment, and determining the size of the self-adaptive bit block according to the smear degree and the importance degree, so as to perform self-adaptive DACs coding on the vibration signal.
Determining the second vibration signal by analyzing the waveform characteristics of the sub-vibration signal segments of the vibration signal
Figure 843812DEST_PATH_IMAGE024
Effective peak point set of sub-vibration signal segments
Figure DEST_PATH_IMAGE064
And effective valley point set
Figure DEST_PATH_IMAGE066
Go on to
Figure 161310DEST_PATH_IMAGE024
And calculating the smear degree and the importance of the sub-vibration signal segments.
Wherein, the first
Figure 938642DEST_PATH_IMAGE024
The calculation process of the effective peak point and the effective valley point of the sub-vibration signal segment is as follows: firstly, to the first
Figure 872226DEST_PATH_IMAGE024
Comparing the signal values of all the points of the sub-vibration signal section, and if the amplitudes of two adjacent points of the signal points in the sub-vibration signal section are smaller than the amplitude of the signal point according to the characteristics of the peak value and the valley value, the signal point is a peak point; and if the amplitudes of two adjacent points of the signal point in the sub-vibration signal segment are larger than the amplitude of the signal point, the signal point is a wave valley point. To the first
Figure 825138DEST_PATH_IMAGE024
Analyzing the sub-vibration signal segment to obtain the second vibration signal segment
Figure 684510DEST_PATH_IMAGE024
Peak point set of sub-vibration signal segments
Figure DEST_PATH_IMAGE068
And valley point set
Figure DEST_PATH_IMAGE070
Wherein
Figure DEST_PATH_IMAGE072
。
Resulting set of peak points
Figure DEST_PATH_IMAGE074
Sum valley point
Figure DEST_PATH_IMAGE076
False peaks and false valleys may occur due to the influence of noise, requiring further analysis. According to the periodic transformation of the vibration signal segment, if the peak point is
Figure 136088DEST_PATH_IMAGE074
Sum valley point
Figure 371898DEST_PATH_IMAGE076
If the partial points in (1) are repeated in different periods or have greater similarity, and the positions of the partial points in different periods are similar, the partial points are effective peak points or effective valley points. Because the number of the initial peak points and the initial valley points obtained due to the influence of the noise in different periods is not in one-to-one correspondence, if the amplitude variation of the initial peak points and the initial valley points in different periods is to be calculated, the influence of the noise needs to be considered.
In the invention, the effective peak point is calculated by taking the current sub-vibration signal segment required to be calculated as a target period, and the first peak point is used
Figure DEST_PATH_IMAGE078
Analyzing for starting point, and calculating the next period
Figure DEST_PATH_IMAGE080
Axis coordinate point
Figure DEST_PATH_IMAGE082
And the range of the abscissa of the point
Figure DEST_PATH_IMAGE084
Calculating the mean value of the amplitudes of all the points in the range and the points
Figure DEST_PATH_IMAGE086
Amplitude of
Figure DEST_PATH_IMAGE088
A difference value of (1), setting a threshold value
Figure DEST_PATH_IMAGE090
If at all
Figure 468730DEST_PATH_IMAGE086
Amplitude of
Figure 499003DEST_PATH_IMAGE088
The difference value with the amplitude mean value is less than the threshold value
Figure 155987DEST_PATH_IMAGE090
Then point of
Figure 929908DEST_PATH_IMAGE086
For the effective peak, similarly all the peak and valley points are calculated to obtain the first
Figure 122992DEST_PATH_IMAGE024
A set of valid peak points and a set of valid valley points for the sub-vibration signal segments. Obtaining an effective peak point set according to the steps
Figure DEST_PATH_IMAGE092
And effective valley point set
Figure DEST_PATH_IMAGE094
。
The invention obtains the bit block size in the current sub-vibration signal segment for compression coding by calculating the smear degree and the importance degree of the signal in the current sub-vibration signal segment. By analyzing the sub-vibration signal segments, the more abnormal the sub-vibration signal segments are, the greater the smear degree of the vibration signal is. The smear degree is obtained by calculation according to the distance between the effective peak points and the distance between the effective valley points in the sub-vibration signal section, wherein the closer the distance between the adjacent effective peak points is, the denser the vibration signal is, and the greater the smear degree of the vibration signal is; the closer the distance between adjacent effective valley points is, the denser the vibration signal is, and the greater the smear degree of the vibration signal is. Wherein the first step
Figure 950265DEST_PATH_IMAGE024
The smear degree calculation expression of the segment vibration signal segment is as follows:
Figure DEST_PATH_IMAGE020A
in the formula:
Figure 693880DEST_PATH_IMAGE022
is shown as
Figure 5912DEST_PATH_IMAGE024
The degree of smear of the segment vibration signal segment,
Figure 53503DEST_PATH_IMAGE026
is shown as
Figure 425578DEST_PATH_IMAGE024
The number of effective peak points/effective valley points of the segment vibration signal segment,
Figure 387980DEST_PATH_IMAGE028
is shown as
Figure 503704DEST_PATH_IMAGE024
In the sub-vibration signal section
Figure 405801DEST_PATH_IMAGE030
Effective peak/valley point and
Figure 447313DEST_PATH_IMAGE032
the distance of the abscissa between the effective peak points/effective valley points,
Figure 67650DEST_PATH_IMAGE034
is shown as
Figure 987064DEST_PATH_IMAGE024
In the sub-vibration signal section
Figure 743668DEST_PATH_IMAGE032
The abscissa of the individual effective peak/effective valley points,
Figure 959011DEST_PATH_IMAGE036
denotes the first
Figure 66644DEST_PATH_IMAGE024
In the sub-vibration signal section
Figure 727432DEST_PATH_IMAGE030
The abscissa of each effective peak/effective valley point.
According to the prior art, the vibration amplitude of the failed bearing vibration signal is obviously different from that of the normal bearing vibration signal, so that the amplitude of the effective peak point and the amplitude of the effective valley point are calculated, and the amplitude threshold value of the data statistics of the normal bearing vibration signal in the prior art is compared with that of the normal bearing vibration signal
Figure DEST_PATH_IMAGE096
And
Figure DEST_PATH_IMAGE098
the importance of the current sub-vibration signal segment is characterized by the difference, wherein the larger the amplitude of the effective peak point is, the larger the amplitude of the effective valley point is, and the larger the importance of the current sub-vibration signal segment is. Then it is first
Figure 57369DEST_PATH_IMAGE024
The calculation expression of the importance of the sub-vibration signal segment is as follows:
Figure DEST_PATH_IMAGE038A
in the formula:
Figure 161722DEST_PATH_IMAGE044
is shown as
Figure 491072DEST_PATH_IMAGE024
The importance of a sub-vibration signal segment,
Figure DEST_PATH_IMAGE100
representing a set of valid peak points
Figure DEST_PATH_IMAGE102
The maximum value of the amplitude of the medium effective peak point,
Figure DEST_PATH_IMAGE104
representing a set of valid valley points
Figure DEST_PATH_IMAGE106
The maximum value of the amplitude of the middle effective valley point,
Figure 234513DEST_PATH_IMAGE096
an amplitude threshold for a valid peak point for normal bearing vibration signal data statistics,
Figure 201594DEST_PATH_IMAGE098
and the amplitude threshold value of the effective valley point is counted by the normal bearing vibration signal data.
In this way,to obtain the first
Figure 522854DEST_PATH_IMAGE024
The smear degree and the importance degree of the sub-vibration signal segment can be obtained in the same way.
And S4, obtaining the size of a bit block of the corresponding sub-vibration signal section according to the importance and the smear degree of the sub-vibration signal section, carrying out DACs (digital audio coding) on the corresponding sub-vibration signal section according to the size of the bit block to obtain a coded vibration signal, transmitting the coded vibration signal to an analysis server, decoding the coded vibration signal by using the analysis server to obtain a decoded vibration signal, and obtaining the fault state of the bearing according to the decoded vibration signal.
And calculating the size of a bit block when each sub-vibration signal is compressed according to the smear degree and the importance degree in different sub-vibration signal segments, namely obtaining a self-adaptive bit block, and performing DACs (digital audio coding) compression coding. The greater the smear degree in the sub-vibration signal segment is, the greater the importance degree is, the more abnormal the vibration signal of the segment is shown, the more important the vibration signal of the segment is, the larger the bit number needs to be set, and the data reading speed is ensured. Get the first
Figure 339500DEST_PATH_IMAGE024
Initial bit block of the sub-vibration signal segment, then
Figure 138829DEST_PATH_IMAGE024
Bit block size of sub-vibration signal segments
Figure 556822DEST_PATH_IMAGE048
The calculation process of (2) is as follows:
Figure DEST_PATH_IMAGE040A
in the formula:
Figure 675082DEST_PATH_IMAGE042
is as follows
Figure 713445DEST_PATH_IMAGE024
The initial block of bits of the sub-vibration signal segment,
Figure 316464DEST_PATH_IMAGE022
is shown as
Figure 287831DEST_PATH_IMAGE024
The extent of smearing of the segment vibration signal segments,
Figure 183850DEST_PATH_IMAGE044
is shown as
Figure 178351DEST_PATH_IMAGE024
The importance of a sub-vibration signal segment,
Figure 319482DEST_PATH_IMAGE046
is as follows
Figure 145356DEST_PATH_IMAGE024
The over-parameters of the sub-vibration signal segments,
Figure 215205DEST_PATH_IMAGE048
is a first
Figure 493740DEST_PATH_IMAGE024
Bit block size of a sub-vibration signal segment.
Wherein: the invention gives an empirical reference value
Figure DEST_PATH_IMAGE108
，
Figure 321451DEST_PATH_IMAGE046
And the bit block is a super parameter and is used for adjusting the value of the whole bit block.
Thus, the adaptive bit block size in all the sub-vibration signal segments is obtained.
After binary conversion is carried out on the vibration signal data in each sub-vibration signal segment, DACs (digital audio coding) is carried out, and the coded data is transmitted to an analysis server, wherein the following specific process of DACs coding is as follows:
and carrying out binary coding on each vibration signal data in each sub vibration signal segment.
Each size is
Figure 736252DEST_PATH_IMAGE048
Bit block of
Figure DEST_PATH_IMAGE110
By using
Figure DEST_PATH_IMAGE112
Identifiers representing bit blocks, using
Figure DEST_PATH_IMAGE114
Representing the remainder of a block of bits
Figure 695111DEST_PATH_IMAGE014
One bit then has
Figure DEST_PATH_IMAGE116
. And starting to cut from the last bit, cutting 2 bits each time, and supplementing 0 if the bits are insufficient. For the identifier of the current bit block, if the current bit block is not the last bit block, the identifier is 1; if the current bit block is the last bit block, the identifier is 0.
As shown in figure 3 of the drawings,
Figure DEST_PATH_IMAGE118
(iv) a few data representing sub-vibration signal segments
Figure DEST_PATH_IMAGE120
A shaft),
Figure DEST_PATH_IMAGE122
represents a specific amplitude value (
Figure DEST_PATH_IMAGE124
Axis), first binary encoding and finally DACs encoding, to
Figure 208745DEST_PATH_IMAGE122
Taking data of 10 as an example, the binary code of the data is 1010, and the binary code is divided into 2, so that the binary code can be divided into 10, each bit block is composed of an identifier and a residual code, so that the divided code 10 needs to be added with the identifier, and the final coding result is 010 110; to be provided with
Figure 193144DEST_PATH_IMAGE122
For example, the data 21 is binary coded to 10101, and is divided into 2, then the binary coded data can be divided into 01 (complement to 0) 01, each bit block is composed of an identifier and the remaining codes, so that the identifier needs to be added to the divided code 0101, and the final coding result is 001 101. The rest of the data and so on.
The specific decoding process is shown in figure 4,
Figure 993609DEST_PATH_IMAGE112
for the identifier, the correspondence is coded as
Figure 903797DEST_PATH_IMAGE114
，
Figure DEST_PATH_IMAGE126
Indicating the number of decoding layers by data
Figure 458056DEST_PATH_IMAGE122
For example 18, the data to be decoded is the 7 th data, then
Figure 744680DEST_PATH_IMAGE126
The layer finds the 7 th position, then the corresponding identifier
Figure DEST_PATH_IMAGE128
Corresponding code
Figure DEST_PATH_IMAGE130
The identifier is 1 to indicate that the current bit block is not the last bit block; continue to in
Figure DEST_PATH_IMAGE132
Layer finding due to data being in
Figure 150385DEST_PATH_IMAGE126
Of a layer
Figure 497053DEST_PATH_IMAGE128
Is located at the sixth (i.e., sixth 1), and is therefore at
Figure 470432DEST_PATH_IMAGE132
Finding the 6 th position in the layer, the corresponding identifier
Figure DEST_PATH_IMAGE134
Corresponding code
Figure DEST_PATH_IMAGE136
The identifier is 1, which indicates that the current bit block is not the last bit block; continue to in
Figure DEST_PATH_IMAGE138
Layer finding due to data being in
Figure 249163DEST_PATH_IMAGE132
Of a layer
Figure 696325DEST_PATH_IMAGE134
Is located at the third (i.e. the third 1) and is therefore at
Figure 479473DEST_PATH_IMAGE138
Finding a third position in the layer, the corresponding identifier
Figure DEST_PATH_IMAGE140
Corresponding code
Figure DEST_PATH_IMAGE142
An identifier of 0 indicates that the current bit block is the last bit block. Thus, a binary of the 7 th data is obtained: 010010, the corresponding data is 18. The rest data are analogized in turn.
Thus, a coded vibration signal is obtained.
And transmitting the encoded vibration signal data to an analysis server, decoding the encoded vibration signal, extracting a kurtosis factor, a peak factor and a pulse factor of the vibration signal as characteristic indexes for training a neural network, and outputting the fault state of the current vibration signal.
The specific steps of training the neural network are as follows: the neural network adopted by the invention is a CNN convolutional neural network; the training data set is kurtosis factor, peak factor and pulse factor of the decoded bearing vibration signal, and the judgment result of the professional worker on the bearing vibration signal is output, namely 0 is normal, 1 is inner ring fault, 2 is outer ring fault and 3 is rolling body fault; the loss function is a cross entropy function.
Inputting the decoded vibration signal into a trained neural network, and outputting the fault state of the bearing, namely 0 is normal, 1 is inner ring fault, 2 is outer ring fault, and 3 is rolling body fault.
The invention has the beneficial effects that: the invention relates to a method for detecting faults of a bearing of an elevator, which divides a vibration signal into a plurality of sub vibration signal sections by obtaining the period of the vibration signal, calculates the smear degree and the importance degree of each sub vibration signal section, and obtains a self-adaptive bit block of DACs codes of each sub vibration signal section according to the smear degree and the importance degree of each sub vibration signal section, thereby achieving the purpose of ensuring the compression ratio and the speed of reading fault data, and finally utilizing a neural network to achieve the real-time and rapid detection of the faults.
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents, improvements and the like that fall within the spirit and principle of the present invention are intended to be included therein.

Claims (9)

1. A method of detecting elevator bearing failure, comprising:
s1, collecting vibration signals of a bearing in the working state of an elevator, setting different time intervals for the vibration signals, obtaining the similarity of the amplitude values of corresponding signal points at the interval points of each time interval, obtaining the period of the vibration signals by using the time interval corresponding to the maximum similarity, and dividing the vibration signals into a plurality of sub vibration signal sections according to the period of the vibration signals;
s2, obtaining amplitudes of all signal points in the sub-vibration signal section, obtaining peak values and valley values of the sub-vibration signal section according to the amplitudes of all the signal points in the sub-vibration signal section, and obtaining effective peak value points and effective valley value points of the sub-vibration signal section by using the obtained peak values and valley values and corresponding threshold values;
s3, obtaining the smear degree of the corresponding sub-vibration signal section according to the distance between the effective peak points and the distance between the effective valley points of each sub-vibration signal section, and obtaining the importance degree of the corresponding sub-vibration signal section according to the amplitude values of the effective peak points and the effective valley points of the sub-vibration signal section;
and S4, obtaining the size of a bit block of the corresponding sub-vibration signal segment according to the importance and the smear degree of the sub-vibration signal segment, carrying out DACs (digital addressable Cs) coding on the corresponding sub-vibration signal segment according to the size of the bit block to obtain a coded vibration signal, transmitting the coded vibration signal to an analysis server, decoding the coded vibration signal by using the analysis server to obtain a decoded vibration signal, and obtaining the fault state of the bearing according to the decoded vibration signal.
2. The elevator bearing fault detection method of claim 1, wherein the period of the vibration signal is determined as follows:
acquiring all time intervals corresponding to the maximum similarity;
acquiring maximum value points corresponding to all time intervals;
and acquiring the mean value of the abscissa distances of the adjacent maximum value points, and taking the obtained mean value as the period of the vibration signal.
3. The method as claimed in claim 1, wherein the similarity of the signal amplitudes corresponding to each time interval point is expressed as:
Figure 398278DEST_PATH_IMAGE002
in the formula:
Figure DEST_PATH_IMAGE003
representing the similarity of the amplitudes of the corresponding signal points at the interval points of each time interval,
Figure 620181DEST_PATH_IMAGE004
representing time interval points
Figure DEST_PATH_IMAGE005
Of the amplitude of the vibration signal of (a),
Figure 954472DEST_PATH_IMAGE006
representing time interval points
Figure DEST_PATH_IMAGE007
Of the amplitude of the vibration signal of (a),
Figure 924086DEST_PATH_IMAGE008
the time interval is represented by a time interval,
Figure DEST_PATH_IMAGE009
representing the average of the amplitudes of all signal points of the vibration signal,
Figure 327516DEST_PATH_IMAGE010
representing the length of the time segment of the vibration signal.
4. The elevator bearing fault detection method of claim 1, wherein the method of obtaining the peak and valley values of the sub-oscillatory signal segments is:
if the amplitudes of two adjacent points of the signal points in the sub-vibration signal section are smaller than the amplitude of the signal point, the signal point is a peak point;
if the amplitudes of two adjacent points of the signal points in the sub-vibration signal section are larger than the amplitude of the signal point, the signal point is a wave valley point;
and taking the amplitudes corresponding to the wave peak point and the wave valley point as the peak value and the valley value of the sub-vibration signal segment.
5. The method of claim 1, wherein the step of using the obtained peaks and valleys and corresponding thresholds to obtain the effective peak and valley points of the sub-vibration signal segment comprises:
acquiring a first peak point of the sub-vibration signal segment;
acquiring an abscissa of a position corresponding to a first peak point of the sub-vibration signal segment in the sub-vibration signal segment of the next period adjacent to the sub-vibration signal segment;
acquiring the range of the abscissa after the addition and subtraction of the abscissa of the corresponding position by half a period;
obtaining the amplitudes of all signal points in the range of the abscissa, and calculating to obtain the average value of the amplitudes of all the signal points;
and if the difference value is smaller than the threshold value, the first peak point of the sub-vibration signal segment is the effective peak point of the sub-vibration signal segment, and in the same way, the effective valley point of the sub-vibration signal segment is obtained.
6. The method according to claim 1, wherein the concrete expression of the smear degree of each sub-vibration signal segment is as follows:
Figure 369422DEST_PATH_IMAGE012
in the formula:
Figure DEST_PATH_IMAGE013
is shown as
Figure 915810DEST_PATH_IMAGE014
The extent of smearing of the segment vibration signal segments,
Figure DEST_PATH_IMAGE015
is shown as
Figure 601612DEST_PATH_IMAGE014
The number of the effective peak points of the segment vibration signal segment,
Figure 397399DEST_PATH_IMAGE016
is shown as
Figure 774154DEST_PATH_IMAGE014
In the sub-vibration signal section
Figure DEST_PATH_IMAGE017
A significant peak point and
Figure 801147DEST_PATH_IMAGE018
the distance on the abscissa between the effective peak points,
Figure DEST_PATH_IMAGE019
is shown as
Figure 503523DEST_PATH_IMAGE014
In the sub-vibration signal section
Figure 431946DEST_PATH_IMAGE018
The abscissa of the point of the effective peak,
Figure 64922DEST_PATH_IMAGE020
is shown as
Figure 398951DEST_PATH_IMAGE014
In the sub-vibration signal section
Figure 226224DEST_PATH_IMAGE017
The abscissa of each valid peak point.
7. The method according to claim 1, wherein the importance of each sub-vibration signal segment is expressed by:
Figure 747335DEST_PATH_IMAGE022
in the formula:
Figure 184001DEST_PATH_IMAGE013
is shown as
Figure 106958DEST_PATH_IMAGE014
The extent of smearing of the segment vibration signal segments,
Figure 367782DEST_PATH_IMAGE015
is shown as
Figure 376189DEST_PATH_IMAGE014
The number of effective peak points/effective valley points of the segment vibration signal segment,
Figure 616546DEST_PATH_IMAGE016
is shown as
Figure 659589DEST_PATH_IMAGE014
In the sub-vibration signal section
Figure 828664DEST_PATH_IMAGE017
Effective peak/valley point and
Figure 58788DEST_PATH_IMAGE018
the distance of the abscissa between the effective peak points/effective valley points,
Figure 102837DEST_PATH_IMAGE019
is shown as
Figure 617695DEST_PATH_IMAGE014
In the sub-vibration signal section
Figure 206939DEST_PATH_IMAGE018
The abscissa of the individual effective peak/effective valley points,
Figure 439206DEST_PATH_IMAGE020
is shown as
Figure 772099DEST_PATH_IMAGE014
Sub-vibration signal section one
Figure 274887DEST_PATH_IMAGE017
The abscissa of each valid peak/valid valley point.
8. The method according to claim 1, wherein the specific expression of the adaptive bit block of each sub-vibration signal segment is as follows:
Figure 35032DEST_PATH_IMAGE024
in the formula:
Figure DEST_PATH_IMAGE025
is as follows
Figure 957858DEST_PATH_IMAGE014
The initial block of bits of the sub-vibration signal segment,
Figure 576664DEST_PATH_IMAGE013
is shown as
Figure 183226DEST_PATH_IMAGE014
The extent of smearing of the segment vibration signal segments,
Figure 894699DEST_PATH_IMAGE026
is shown as
Figure 212810DEST_PATH_IMAGE014
The importance of a sub-vibration signal segment,
Figure DEST_PATH_IMAGE027
is as follows
Figure 340035DEST_PATH_IMAGE014
The over-parameters of the sub-vibration signal segments,
Figure 66683DEST_PATH_IMAGE028
is shown as
Figure 187872DEST_PATH_IMAGE014
A block of bits of a sub-vibration signal segment.
9. The method of claim 1, wherein the method of obtaining the fault status of the bearing according to the decoded vibration signal comprises:
acquiring a kurtosis factor, a peak factor and a pulse factor of a vibration signal;
taking a kurtosis factor, a peak value factor and a pulse factor of a vibration signal as the input of a neural network, taking the fault state of a bearing as the output of the neural network, and training the neural network to obtain a trained neural network;
acquiring a kurtosis factor, a peak factor and a pulse factor of the decoded vibration signal;
and taking the kurtosis factor, the peak factor and the pulse factor of the decoded vibration signal as the input of the trained neural network, and outputting the fault state of the bearing.
CN202211081144.5A 2022-09-06 2022-09-06 Elevator bearing fault detection method Pending CN115184016A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202211081144.5A CN115184016A (en) 2022-09-06 2022-09-06 Elevator bearing fault detection method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202211081144.5A CN115184016A (en) 2022-09-06 2022-09-06 Elevator bearing fault detection method

Publications (1)

Publication Number Publication Date
CN115184016A true CN115184016A (en) 2022-10-14

Family

ID=83523217

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202211081144.5A Pending CN115184016A (en) 2022-09-06 2022-09-06 Elevator bearing fault detection method

Country Status (1)

Country Link
CN (1) CN115184016A (en)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115695564A (en) * 2022-12-30 2023-02-03 深圳市润信数据技术有限公司 Efficient transmission method for data of Internet of things
CN117743836A (en) * 2024-02-21 2024-03-22 聊城市产品质量监督检验所 Abnormal vibration monitoring method for bearing
CN117743836B (en) * 2024-02-21 2024-05-03 聊城市产品质量监督检验所 Abnormal vibration monitoring method for bearing

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1531814A (en) * 2001-09-04 2004-09-22 ��ʽ����Ntt����Ħ Encoding method selection method and terminal apparatus
US20110103432A1 (en) * 2008-10-02 2011-05-05 Texas Instruments Incorporated Enhanced cross correlation detection or mitigation circuits, processes, devices, receivers and systems
CN102256134A (en) * 2002-06-28 2011-11-23 杜比实验室认证公司 Improved interpolation of video compression frames
CN105900170A (en) * 2014-01-07 2016-08-24 哈曼国际工业有限公司 Signal quality-based enhancement and compensation of compressed audio signals
CN111412977A (en) * 2020-03-09 2020-07-14 华南理工大学 Preprocessing method for vibration sensing data of mechanical equipment
CN111899295A (en) * 2020-06-06 2020-11-06 东南大学 Monocular scene depth prediction method based on deep learning
CN113850174A (en) * 2021-09-22 2021-12-28 吉林化工学院 Unknown fault-containing wind turbine generator main bearing fault diagnosis method

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1531814A (en) * 2001-09-04 2004-09-22 ��ʽ����Ntt����Ħ Encoding method selection method and terminal apparatus
CN102256134A (en) * 2002-06-28 2011-11-23 杜比实验室认证公司 Improved interpolation of video compression frames
US20110103432A1 (en) * 2008-10-02 2011-05-05 Texas Instruments Incorporated Enhanced cross correlation detection or mitigation circuits, processes, devices, receivers and systems
CN105900170A (en) * 2014-01-07 2016-08-24 哈曼国际工业有限公司 Signal quality-based enhancement and compensation of compressed audio signals
CN111412977A (en) * 2020-03-09 2020-07-14 华南理工大学 Preprocessing method for vibration sensing data of mechanical equipment
CN111899295A (en) * 2020-06-06 2020-11-06 东南大学 Monocular scene depth prediction method based on deep learning
CN113850174A (en) * 2021-09-22 2021-12-28 吉林化工学院 Unknown fault-containing wind turbine generator main bearing fault diagnosis method

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115695564A (en) * 2022-12-30 2023-02-03 深圳市润信数据技术有限公司 Efficient transmission method for data of Internet of things
CN115695564B (en) * 2022-12-30 2023-03-10 深圳市润信数据技术有限公司 Efficient transmission method of Internet of things data
CN117743836A (en) * 2024-02-21 2024-03-22 聊城市产品质量监督检验所 Abnormal vibration monitoring method for bearing
CN117743836B (en) * 2024-02-21 2024-05-03 聊城市产品质量监督检验所 Abnormal vibration monitoring method for bearing

Similar Documents

Publication Publication Date Title
CN107316046B (en) Fault diagnosis method based on incremental compensation dynamic self-adaptive enhancement
CN110470477B (en) Rolling bearing fault diagnosis method based on SSAE and BA-ELM
CN109583386B (en) Intelligent rotating machinery fault depth network feature identification method
CN111931625B (en) Product key part residual life prediction method based on asymmetric loss neural network
CN115184016A (en) Elevator bearing fault detection method
CN112070208B (en) Tool wear prediction method based on encoder-decoder stage attention mechanism
CN113642414A (en) Method for predicting residual service life of rolling bearing based on Transformer model
CN116150901A (en) Rolling bearing residual life prediction method based on attention-enhanced time-frequency converter
CN110927497B (en) Point switch fault detection method and device
CN113343591A (en) Product key part life end-to-end prediction method based on self-attention network
CN116740523A (en) Slewing bearing fault diagnosis method based on image multi-level fusion convolution
CN117235557B (en) Electrical equipment fault rapid diagnosis method based on big data analysis
CN102353814B (en) Method for detecting output signal failure of wind direction sensor
CN115695564B (en) Efficient transmission method of Internet of things data
CN115114565B (en) Linear motor fault detection method and system based on big data analysis
CN116465623A (en) Gearbox service life prediction method based on sparse converter
CN114781450B (en) Train rolling bearing state identification method based on parameter optimization MOMEDA-MIA-CNN
CN117213856A (en) Bearing residual life prediction method based on Yun Bian double-data-source fusion
CN1422006A (en) Receiving circuit for receiving information signal
CN114367710B (en) Electric spark machining control method based on deep learning and acoustic emission signals
CN110598674B (en) Point switch fault identification method based on pattern identification
CN111832910B (en) Multi-index abnormal sound judgment threshold value determining method, system and computer equipment
CN115270895B (en) Fault detection method for diesel engine
CN114935676B (en) Digital circuit and method for preprocessing trigger data based on FPGA
CN117113107B (en) Rotor die casting abnormal error data processing method

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