CN110836775A - Rolling bearing fault identification method and system based on multi-dimensional entropy space distance - Google Patents
Rolling bearing fault identification method and system based on multi-dimensional entropy space distance Download PDFInfo
- Publication number
- CN110836775A CN110836775A CN201911352259.1A CN201911352259A CN110836775A CN 110836775 A CN110836775 A CN 110836775A CN 201911352259 A CN201911352259 A CN 201911352259A CN 110836775 A CN110836775 A CN 110836775A
- Authority
- CN
- China
- Prior art keywords
- entropy
- transformation
- working condition
- calculating
- rolling bearing
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
- G01M13/00—Testing of machine parts
- G01M13/04—Bearings
- G01M13/045—Acoustic or vibration analysis
Landscapes
- Physics & Mathematics (AREA)
- Acoustics & Sound (AREA)
- General Physics & Mathematics (AREA)
- Testing Of Devices, Machine Parts, Or Other Structures Thereof (AREA)
- Measurement Of Mechanical Vibrations Or Ultrasonic Waves (AREA)
Abstract
The application discloses a rolling bearing fault identification method and system based on multi-dimensional entropy space distance, which are used for collecting original vibration signals of normal working conditions and different fault types of a rolling bearing and selecting 10 groups of sample data of each working condition; performing noise reduction processing on the sample data by using EMD to obtain a plurality of IMF components; selecting the first 5 IMF components of each working condition, respectively calculating the information entropy, the energy entropy and the permutation entropy of IMF component transformation of each group of sample data of each working condition, and respectively calculating the average value; and constructing a three-dimensional space, randomly selecting a preset number of groups of test data, calculating three entropy values after transformation, calculating a spatial distance with an entropy value point of each working condition in the three-dimensional space in the previous step, judging the distance between the test data and each type of bearing working condition, and finally determining the working condition state of the rolling bearing. Therefore, the fault characteristics of the rolling bearing can be effectively extracted, and the accuracy of fault identification is improved.
Description
Technical Field
The application relates to the technical field of rolling bearing operation fault diagnosis, in particular to a rolling bearing fault identification method and system based on a multidimensional entropy space distance.
Background
The rolling bearing is one of the most important parts in the current rotating equipment, and the safety performance of industrial equipment can be directly influenced once the rolling bearing is damaged, so that how to utilize the acquired vibration signals of the rolling bearing to effectively extract the characteristic quantity which is obvious and can diagnose the fault type of the rolling bearing has important significance for the detection and fault diagnosis of the rolling bearing.
At present, time-frequency analysis methods such as wavelet transformation, Fourier decomposition and the like are commonly used at home and abroad to carry out fault diagnosis on the rolling bearing, but the methods have certain limitations. For example, wavelet transformation needs to select wavelet bases, and if the wavelet bases are different, the result of wavelet decomposition is also different, but currently, no better method exists for selecting the wavelet bases. The problems of inaccurate extracted characteristic quantity, low recognition rate and the like still exist in the existing fault diagnosis technology.
Disclosure of Invention
In order to solve the technical problems, the following technical scheme is provided:
in a first aspect, an embodiment of the present application provides a rolling bearing fault identification method based on a multidimensional entropy space distance, where the method includes: collecting original vibration signals of a rolling bearing under normal working conditions and different fault types, and selecting 10 groups of sample data of each working condition; performing noise reduction processing on the sample data by using Empirical Mode Decomposition (EMD) to obtain a plurality of Intrinsic Mode Function (IMF) components; selecting the first 5 IMF components of each working condition, respectively calculating the information entropy, the energy entropy and the permutation entropy of IMF component transformation of each group of sample data of each working condition, and respectively calculating the average value of the transformation information entropy, the average value of the transformation energy entropy and the average value of the transformation permutation entropy of 10 groups of sample data; constructing a three-dimensional space by taking the average value of the transformation permutation entropy as an X axis, the average value of the transformation information entropy as a Y axis and the average value of the transformation energy entropy as a Z axis; randomly selecting a preset number of groups of test data, calculating three entropy values after transformation according to the steps, calculating a space distance from an entropy value point of each working condition in the three-dimensional space of the previous step, and judging the distance between the test data and each type of bearing working condition.
By adopting the implementation mode, aiming at non-Gaussian and non-linear vibration signals of the movable bearing, firstly, the EMD is used for noise reduction, then, the first 5 IMF components of each working condition are selected, the information entropy, the energy entropy and the permutation entropy of each group of sample data IMF component transformation of each working condition are respectively calculated, and the average value of the information entropy, the energy entropy and the permutation entropy is calculated. And constructing a three-dimensional space, and finally determining the working condition state of the rolling bearing. Therefore, the fault characteristics of the rolling bearing can be effectively extracted, and the accuracy of fault identification is improved.
With reference to the first aspect, in a first possible implementation manner of the first aspect, the performing noise reduction processing on the sample data by using Empirical Mode Decomposition (EMD) to obtain a plurality of Intrinsic Mode Function (IMF) components includes:
h1＝x(t)-m1
r1＝x(t)-c1
rn＝rn-1-cn
wherein m is1Is the mean of the upper and lower envelope lines, x (t) is the original vibration signal, h1Is the difference between the two; r is1As a residual signal, rnFor the nth residual signal, c1(t)，c2(t)，…，cn(t) is the sum of n IMF components and X (t) is the sum of n IMF components and a residual signal.
And separating impacts of different scales in the vibration signal one by using Empirical Mode Decomposition (EMD) to obtain a plurality of intrinsic mode function components, namely IMF. Furthermore, the IMF must be a function that satisfies two conditions: firstly, in the time of an original vibration signal, the difference between the number of zero points and the number of extreme points cannot be larger than 1; and secondly, the mean values of the upper envelope line and the lower envelope line are both 0 at any time of the vibration signal. EMD can thus be regarded as a noise reduction method, reducing the effect of noise on the original vibration signal.
Judging h according to the two satisfied conditions of the IMF1If not, let h1Recalculate instead of x (t) as a new signal; if h1Satisfy the two conditions of IMF, then h1Is changed into c1。r1Is a difference valueSignals, also called residual signals, rnIs the nth residual signal.
With reference to the first possible implementation manner of the first aspect, in a second possible implementation manner of the first aspect, the calculating information entropy of each group of sample data IMF component transformations in each operating condition, and calculating an average value of the information entropy of 10 groups of sample data transformations includes:
carrying out normalization processing on the IMF components c (t), and then carrying out blocking; in the application, 10 blocks are used, for example, 0-10 blocks are used as first blocks, 10-20 blocks are used as second blocks, and the like.
Counting the number of data in each block range, and calculating the corresponding probability;
determining the average HX of the entropy of the information1*sum(-p.*log2(p))-a2Where p is the probability, HX is the information entropy of the transformation, a1，a2Respectively different positive integers.
With reference to the second possible implementation manner of the first aspect, in a third possible implementation manner of the first aspect, the calculating an energy entropy of each group of sample data IMF component transformations in each operating condition, and calculating an average value of the energy entropies of 10 groups of sample data transformations includes:
wherein, ci(t) is the IMF component, EcIs the total energy, pi＝Ei/EcThe ratio of the ith IMF energy to the total energy, b1，b2Is a positive integer.
With reference to the third possible implementation manner of the first aspect, in a fourth possible implementation manner of the first aspect, the calculating the permutation entropy of each set of sample data IMF component transformation under each working condition, and calculating an average value of the permutation entropies of 10 sets of sample data transformation includes:
one-dimensional discrete time series: reconstructing any element X (i) in the discrete time sequence by a phase space reconstruction delay coordinate method, and selecting continuous m sample points around the sampling point to obtain a reconstruction vector of an m-dimensional space of the sampling point X (i): x ═ X (i), X (i +1), ·, X (i + (m-1) ·, l), m and l are the dimension of the reconstruction space and the delay time, respectively;
and (3) performing corresponding ascending arrangement on the reconstructed vectors after the discrete time sequence x (i) is reconstructed to obtain: xi'＝{x(i+(j1-1)*l)≤x(i+(j2-1)*l)≤···≤x(i+(jm-1) × } results in a new permutation: i { j1,j2,···,jm}；
Counting the number of times of occurrence of various arrangement situations of the X sequence, which is equal to the m! As the probability p of occurrenceiCalculating the sequence permutation entropy:
wherein: h is the permutation entropy of the transform, piFor the probability of occurrence of various permutation conditions, d1，d2Is a positive integer.
In a second aspect, an embodiment of the present application provides a rolling bearing fault identification system based on a multidimensional entropy space distance, where the system includes: the system comprises an acquisition module, a storage module and a processing module, wherein the acquisition module is used for acquiring original vibration signals of a rolling bearing under normal working conditions and different fault types and selecting 10 groups of sample data of each working condition; the acquisition module is used for carrying out noise reduction processing on the sample data through EMD to obtain a plurality of intrinsic mode functions IMF components; the processing module is used for selecting the first 5 IMF components of each working condition, respectively calculating the information entropy, the energy entropy and the permutation entropy of the IMF component transformation of each group of sample data of each working condition, and respectively calculating the average value of the transformation information entropy, the average value of the transformation energy entropy and the average value of the transformation permutation entropy of 10 groups of sample data; the three-dimensional space construction module is used for constructing a three-dimensional space by taking the average value of the transformation arrangement entropy as an X axis, taking the average value of the transformation information entropy as a Y axis and taking the average value of the transformation energy entropy as a Z axis; the determining module is used for randomly selecting a preset number of groups of test data, calculating three entropy values after transformation according to the steps, calculating a space distance with an entropy value point of each working condition in a three-dimensional space of the last step, judging the distance between the test data and each type of bearing working condition, and if the distance between the test data and the first bearing working condition is minimum, the working condition state of the rolling bearing is the first bearing working condition, and the first bearing working condition is any one of all working condition states of the rolling bearing.
With reference to the second aspect, in a first possible implementation manner of the second aspect, the obtaining module includes:
h1＝x(t)-m1
r1＝x(t)-c1
rn＝rn-1-cn
wherein m is1Is the mean of the upper and lower envelope lines, x (t) is the original vibration signal, h1Is the difference between the two; r is1As a residual signal, rnFor the nth residual signal, c1(t)，c2(t)，…，cn(t) is the sum of n IMF components and X (t) is the sum of n IMF components and a residual signal.
With reference to the first possible implementation manner of the second aspect, in a second possible implementation manner of the second aspect, the processing module includes:
the normalization processing unit is used for performing normalization processing on the IMF components c (t) and then performing blocking;
the statistical unit is used for counting the number of data in each block range and calculating the corresponding probability;
a determination unit for determining an average HX a of the information entropy1*sum(-p.*log2(p))-a2Where p is the probability, HX is the information entropy of the transformation, a1，a2Respectively different positive integers.
With reference to the second possible implementation manner of the second aspect, in a third possible implementation manner of the second aspect, the processing module includes:
wherein, ci(t) is the IMF component, EcIs the total energy, pi＝Ei/EcThe ratio of the ith IMF energy to the total energy, b1，b2Is a positive integer.
With reference to the third possible implementation manner of the second aspect, in a fourth possible implementation manner of the second aspect, the processing module includes:
a reconstruction unit for reconstructing the one-dimensional discrete-time series: reconstructing any element X (i) in the discrete time sequence by a phase space reconstruction delay coordinate method, and selecting continuous m sample points around the sampling point to obtain a reconstruction vector of an m-dimensional space of the sampling point X (i): x ═ X (i), X (i +1), ·, X (i + (m-1) ·, l), m and l are the dimension of the reconstruction space and the delay time, respectively;
an arranging unit, configured to perform corresponding ascending arrangement on the reconstructed vectors after the discrete time sequence x (i) is reconstructed, so as to obtain: xi'＝{x(i+(j1-1)*l)≤x(i+(j2-1)*l)≤···≤x(i+(jm-1) × } results in a new permutation: i { j1,j2,···,jm}；
A calculating unit for counting the occurrence times of various arrangement situations of the X sequence, which is equal to the full arrangement m! As the probability p of occurrenceiCalculating the sequence permutation entropy:
wherein: h is the permutation entropy of the transform, piFor the probability of occurrence of various permutation conditions, d1，d2Is a positive integer.
Drawings
Fig. 1 is a schematic flowchart of a rolling bearing fault identification method based on a multidimensional entropy space distance according to an embodiment of the present application;
fig. 2 is a time domain signal diagram of four working conditions of a first set of sample data provided in the embodiment of the present application;
fig. 3 is a time domain diagram of the first 5 IMF components after the EMD decomposition of the first set of four working conditions of sample data provided by the embodiment of the present application;
FIG. 4 is a 3-dimensional space formed by the average values of three entropies of 10 sets of sample data provided in the embodiment of the present application;
FIG. 5 is an image of a set of test data closest to an outer ring fault provided by an embodiment of the present application;
fig. 6 is a schematic diagram of a rolling bearing fault identification system based on a multidimensional entropy space distance according to an embodiment of the present application.
Detailed Description
The present invention will be described with reference to the accompanying drawings and embodiments.
Fig. 1 is a schematic flow diagram of a rolling bearing fault identification method based on a multidimensional entropy space distance provided in an embodiment of the present application, and referring to fig. 1, the rolling bearing fault identification method based on the multidimensional entropy space distance provided in the embodiment of the present application includes:
s101, acquiring original vibration signals of the rolling bearing under normal working conditions and different fault types, and selecting 10 groups of sample data of each working condition.
The application is operated and completed in an MATLAB 2014a software environment. This application data is for placing the vibration acceleration signal of the antifriction bearing of the acceleration sensor collection in motor drive end bearing frame top, and it carries out the single-point damage to the bearing with the electric spark processing technique, and drive end damage diameter does respectively: 0.1778 mm, 0.3556 mm, 0.5334 mm, the diameter chosen for this application is 0.1778 mm. The bearing model of the driving end is 6205-2RS JEM SKF deep groove ball bearing, the sampling frequency of the bearing is 12Khz and 48Khz, the sampling frequency selected by the application is 12Khz, and the rotating speed of the motor is 1750 r/min.
The method comprises the steps of collecting original vibration signals of normal working conditions of a rolling bearing, ball faults, outer ring faults and inner ring faults, and selecting 10 groups of sample data of each working condition, wherein the data volume of the sample data of each working condition is 4000. A time domain image of the first set of sample data is shown in figure 2.
S102, performing noise reduction processing on the sample data by using Empirical Mode Decomposition (EMD) to obtain a plurality of Intrinsic Mode Functions (IMF) components.
Preprocessing the signals, and separating impacts of different scales in the vibration signals one by Empirical Mode Decomposition (EMD) to obtain a plurality of intrinsic mode function components, which are called IMF. Furthermore, the IMF must be a function that satisfies two conditions: firstly, in the time of an original vibration signal, the difference between the number of zero points and the number of extreme points cannot be larger than 1; and secondly, the mean values of the upper envelope line and the lower envelope line are both 0 at any time of the vibration signal. EMD can thus be regarded as a noise reduction method, reducing the effect of noise on the original vibration signal. The concrete formula is as follows:
h1＝x(t)-m1
r1＝x(t)-c1
rn＝rn-1-cn
wherein m is1Is the mean of the upper and lower envelope lines, x (t) is the original vibration signal, h1The difference between the two. Judging h according to the two satisfied conditions of the IMF1If not, let h1Recalculate instead of x (t) as a new signal; if h1Satisfy the two conditions of IMF, then h1Is changed into c1。r1Is a difference signal, also called residual signal, rnFor the nth residual signal, c1(t)，c2(t)，…，cn(t) is the sum of n IMF components and a residual signal, and X (t) is the sum of n IMF components and a residual signal.
S103, selecting the first 5 IMF components of each working condition, respectively calculating the information entropy, the energy entropy and the permutation entropy of the IMF component transformation of each group of sample data of each working condition, and respectively calculating the average value of the transformation information entropy, the average value of the transformation energy entropy and the average value of the transformation permutation entropy of 10 groups of sample data.
Selecting the first 5 IMF components of each working condition, i.e. selecting c obtained in the previous step1(t)，c2(t)，c3(t)，c4(t)，c5(t) of (d). The results for the first set of sample data are shown in fig. 3.
And calculating the information entropy of IMF component transformation of each group of sample data under each working condition, and calculating the average value of the information entropy of transformation of 10 groups of sample data. Firstly, the IMF component c (t) is normalized and then partitioned, in this application, 10 partitions are used, for example, 0-10 are used as the first partition, 10-20 are used as the second partition, and so on. Then, the number of data in each block range is counted, and the corresponding probability is calculated. And finally, solving the information entropy. The concrete formula is as follows:
HX＝a1*sum(-p.*log2(p))-a2
wherein: p is probability, HX is information entropy of transformation, a1，a2Respectively different positive integers. A is obtained by a large number of experiments1＝2，a21. 5 layers of IMF components of each group of sample data can obtain a transformation information entropy, and the average value of the transformation information entropy of each layer of IMF components of 10 groups of sample data is calculated. The results are shown in Table 1 below.
TABLE 1 mean value of the entropy of the transformation information for four operating modes
And calculating the energy entropy of each IMF component transformation of each working condition. The concrete formula is as follows:
wherein, ci(t) is the IMF component, EcIs the total energy, pi＝Ei/EcThe ratio of the energy of the ith IMF to the total energy, i.e. the energy entropy of the required transformation, b1，b2Is a positive integer. B was obtained from a number of experiments1＝10，b 20. According to the formula, when the energy entropy of transformation is obtained, the energy of each layer of signal is obtained firstly, and then the energy entropy of transformation is obtained. The results are shown in Table 2 below.
TABLE 2 mean value of the energy entropy of the four operating modes
And calculating the permutation entropy of each IMF component transformation of each working condition. The specific calculation steps are as follows:
first, a one-dimensional discrete time series: x (1), X (2), ·, X (n) reconstructs a phase space of any element X (i) in the discrete time series by a phase space reconstruction delay coordinate method. And selecting continuous m sample points around the sampling point to obtain a reconstruction vector of the m-dimensional space of the sampling point x (i): x ═ X (i), X (i +1), ·, X (i + (m-1) · l). A phase space matrix of the discrete time series X can be obtained. M and l in the matrix are respectively the dimension and delay time of the reconstruction space, and m is 2 and l is 6 in the application. Then, performing corresponding ascending arrangement on the reconstructed vectors Xi after the discrete time sequence x (i) is reconstructed, so as to obtain: xi'＝{x(i+(j1-1)*l)≤x(i+(j2-1)*l)≤···≤x(i+(jm-1) × l). Thus, a new arrangement can be obtained: i { j1,j2,···,jm}. The arrangement is a full arrangement m! The number of times of occurrence of various arrangement conditions of the X sequence is counted, which is compared with m! As the probability p of occurrenceiCalculating the sequence permutation entropy:
where H is the permutation entropy of the transform, piFor the probability of occurrence of various permutation conditions, d1，d2Is a positive integer. D was obtained from a number of experiments1＝10000，d26900. The results are shown in Table 3 below.
TABLE 3 mean value of the permutation entropy for the four operating modes
And S104, constructing a three-dimensional space by taking the average value of the transformation arrangement entropy as an X axis, the average value of the transformation information entropy as a Y axis and the average value of the transformation energy entropy as a Z axis.
And constructing a three-dimensional space, wherein the average value of the transformation arrangement entropy is used as an X axis, the average value of the transformation information entropy is used as a Y axis, and the average value of the transformation energy entropy is used as a Z axis to form a point in a three-dimensional space. The points of the first 5 IMF components for each condition are connected to form a line graph in three-dimensional space. The results are shown in FIG. 4.
S105, randomly selecting a preset number of groups of test data, calculating three entropy values after transformation according to the steps, calculating a space distance from an entropy value point of each working condition in the three-dimensional space of the previous step, judging the distance between the test data and each type of bearing working condition, and if the distance between the test data and the first bearing working condition is minimum, determining that the working condition state of the rolling bearing is the first bearing working condition, wherein the first bearing working condition is any one of all working condition states of the rolling bearing.
And randomly selecting a group of data with the data quantity of 4000, and calculating the information entropy, the energy entropy and the permutation entropy after the linear transformation of the first 5 IMF components after EMD decomposition. And drawing a three-dimensional space diagram of the bearing, and then respectively calculating the space distance between the bearing and the three-dimensional space point of the sample data of the four working conditions, so that the minimum space distance between the bearing and the outer ring fault is 2.4665, and the bearing state can be identified as the outer ring fault. The arrangement entropy, the information entropy and the energy entropy after the linear transformation are shown in the following table 4, the space distance between the linear transformation and the sample data of four working conditions is shown in the following table 5, and the result is shown in fig. 5.
TABLE 4 set of three entropy values after an exemplary Linear transformation
Table 5 illustrates the spatial distances between sample data for four operating conditions
The data volume of each collected working condition is 120000, the data volume of each group of sample data is 4000, the data volume of the whole sample data is 40000, 190 groups of test data are randomly selected from the following 80000, and the data volume of each group of test data is 4000. . And calculating information entropy, permutation entropy and energy entropy after linear transformation of the first 5 IMF components after EMD decomposition of the test data, and then calculating a space distance, namely the distance between two points in the space, with the average entropy value point of each working condition in the three-dimensional space in the previous step respectively. And if the distance between the test data and a certain working condition in the sample data is minimum, the bearing is in which fault. The accuracy of the rolling bearing fault identification method based on the multidimensional entropy space distance is 93.16%.
According to the embodiment, the rolling bearing fault identification method based on the multi-dimensional entropy space distance is provided, aiming at non-Gaussian and non-linear dynamic bearing vibration signals, firstly, noise reduction is carried out through EMD, then, the first 5 IMF components of each working condition are selected, the information entropy, the energy entropy and the permutation entropy of each group of sample data IMF component transformation of each working condition are respectively calculated, and the average value of the information entropy, the energy entropy and the permutation entropy is calculated. And constructing a three-dimensional space, and finally determining the working condition state of the rolling bearing. Therefore, the fault characteristics of the rolling bearing can be effectively extracted, and the accuracy of fault identification is improved.
Corresponding to the rolling bearing fault identification method based on the multidimensional entropy space distance provided by the above embodiment, the present application also provides an embodiment of a rolling bearing fault identification system based on the multidimensional entropy space distance, and referring to fig. 6, the rolling bearing fault identification system 20 based on the multidimensional entropy space distance includes: the system comprises an acquisition module 201, an acquisition module 202, a processing module 203, a three-dimensional space construction module 204 and a determination module 205.
The acquisition module 201 is configured to acquire original vibration signals of a rolling bearing under normal conditions and different fault types, and select 10 sets of sample data of each condition. The obtaining module 202 is configured to perform noise reduction processing on the sample data through empirical mode decomposition EMD to obtain multiple intrinsic mode functions IMF components. The processing module 203 is configured to select the first 5 IMF components of each operating condition, calculate the information entropy, the energy entropy, and the permutation entropy of the IMF component transformation of each group of sample data of each operating condition, and calculate the average value of the transformation information entropy, the average value of the transformation energy entropy, and the average value of the transformation permutation entropy of 10 groups of sample data. The three-dimensional space construction module 204 is configured to construct a three-dimensional space by using the average value of the transformation permutation entropy as an X axis, the average value of the transformation information entropy as a Y axis, and the average value of the transformation energy entropy as a Z axis. The determining module 205 is configured to randomly select a preset number of groups of test data, calculate three entropy values after transformation according to the above steps, calculate a spatial distance with an entropy value point of each working condition in the three-dimensional space of the previous step, and determine a distance between the test data and each type of bearing working condition.
Further, the obtaining module 202 preprocesses the signal, and separates impacts of different scales in the vibration signal one by Empirical Mode Decomposition (EMD) to obtain a plurality of intrinsic mode function components IMF. Furthermore, the IMF must be a function that satisfies two conditions: firstly, in the time of an original vibration signal, the difference between the number of zero points and the number of extreme points cannot be larger than 1; and secondly, the mean values of the upper envelope line and the lower envelope line are both 0 at any time of the vibration signal. EMD can thus be regarded as a noise reduction method, reducing the effect of noise on the original vibration signal. The concrete formula comprises:
h1＝x(t)-m1
r1＝x(t)-c1
rn＝rn-1-cn
wherein m is1Is the mean of the upper and lower envelope lines, x (t) is the original vibration signal, h1Is the difference between the two; r is1As a residual signal, rnFor the nth residual signal, c1(t)，c2(t)，…，cn(t) is the sum of n IMF components and X (t) is the sum of n IMF components and a residual signal.
The processing module 203 comprises: the device comprises a normalization processing unit, a statistical unit and a determination unit.
And the normalization processing unit is used for performing normalization processing on the IMF components c (t) and then performing blocking. And the statistical unit is used for counting the number of data in each block range and calculating the corresponding probability. The determining unit is used for determining the average value HX of the information entropy1*sum(-p.*log2(p))-a2Where p is the probability, HX is the information entropy of the transformation, a1，a2Respectively different positive integers.
Further, the processing module 203 calculates an energy entropy of each IMF component transformation for each condition. The concrete formula is as follows:
wherein, ci(t) is the IMF component, EcIs the total energy, pi＝Ei/EcThe ratio of the ith IMF energy to the total energy, b1，b2Is a positive integer.
The processing module 203 further comprises: a reconstruction unit, an arrangement unit and an arrangement unit.
The reconstruction unit is configured to apply the one-dimensional discrete time series: reconstructing any element X (i) in the discrete time sequence by a phase space reconstruction delay coordinate method, and selecting continuous m sample points around the sampling point to obtain a reconstruction vector of an m-dimensional space of the sampling point X (i): x ═ X (i), X (i +1), ·, X (i + (m-1) ·, l), and m and l are the dimension and delay time of the reconstruction space, respectively. An arranging unit, configured to perform corresponding ascending arrangement on the reconstructed vectors after the discrete time sequence x (i) is reconstructed, so as to obtain: xi'＝{x(i+(j1-1)*l)≤x(i+(j2-1)*l)≤···≤x(i+(jm-1) × } results in a new permutation: i { j1,j2,···,jm}. A permutation unit for counting the occurrence times of various permutation conditions of the X sequence, which is equal to the full permutation m! As the probability p of occurrenceiCalculating the sequence permutation entropy:
wherein: h is the permutation entropy of the transform, piFor the probability of occurrence of various permutation conditions, d1，d2Is a positive integer.
It should be noted that, in this document, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other identical elements in a process, method, article, or apparatus that comprises the element.
Of course, the above description is not limited to the above examples, and technical features that are not described in this application may be implemented by or using the prior art, and are not described herein again; the above embodiments and drawings are only for illustrating the technical solutions of the present application and not for limiting the present application, and the present application is only described in detail with reference to the preferred embodiments instead, it should be understood by those skilled in the art that changes, modifications, additions or substitutions within the spirit and scope of the present application may be made by those skilled in the art without departing from the spirit of the present application, and the scope of the claims of the present application should also be covered.
Claims (10)
1. A rolling bearing fault identification method based on multi-dimensional entropy space distance is characterized by comprising the following steps:
collecting original vibration signals of a rolling bearing under normal working conditions and different fault types, and selecting 10 groups of sample data of each working condition;
performing noise reduction processing on the sample data by using Empirical Mode Decomposition (EMD) to obtain a plurality of Intrinsic Mode Function (IMF) components;
selecting the first 5 IMF components of each working condition, respectively calculating the information entropy, the energy entropy and the permutation entropy of IMF component transformation of each group of sample data of each working condition, and respectively calculating the average value of the transformation information entropy, the average value of the transformation energy entropy and the average value of the transformation permutation entropy of 10 groups of sample data;
constructing a three-dimensional space by taking the average value of the transformation permutation entropy as an X axis, the average value of the transformation information entropy as a Y axis and the average value of the transformation energy entropy as a Z axis;
randomly selecting a preset number of groups of test data, calculating three entropy values after transformation according to the steps, calculating a space distance from an entropy value point of each working condition in the three-dimensional space of the previous step, and judging the distance between the test data and each type of bearing working condition.
2. The rolling bearing fault identification method based on the multidimensional entropy space distance as claimed in claim 1, wherein the performing noise reduction processing on the sample data by using Empirical Mode Decomposition (EMD) to obtain a plurality of Intrinsic Mode Functions (IMF) components comprises:
h1＝x(t)-m1
r1＝x(t)-c1
rn＝rn-1-cn
wherein m is1Is the mean of the upper and lower envelope lines, x (t) is the original vibration signal, h1Is the difference between the two; r is1As a residual signal, rnFor the nth residual signal, c1(t)，c2(t)，…，cn(t) is the sum of n IMF components and X (t) is the sum of n IMF components and a residual signal.
3. The rolling bearing fault identification method based on the multidimensional entropy space distance is characterized in that the calculating the information entropy of each group of sample data IMF component transformation under each working condition and the calculating the average value of 10 groups of sample data transformation information entropies comprises the following steps:
carrying out normalization processing on the IMF components c (t), and then carrying out blocking;
counting the number of data in each block range, and calculating the corresponding probability;
determining the average HX of the entropy of the information1*sum(-p.*log2(p))-a2Where p is the probability, HX is the information entropy of the transformation, a1，a2Respectively different positive integers.
4. The rolling bearing fault identification method based on the multidimensional entropy space distance is characterized in that the calculating the energy entropy of each group of sample data IMF component transformation under each working condition and the calculating the average value of the energy entropy of 10 groups of sample data transformation comprises the following steps:
wherein, ci(t) is the IMF component, EcIs the total energy, pi＝Ei/EcThe ratio of the ith IMF energy to the total energy, b1，b2Is a positive integer.
5. The rolling bearing fault identification method based on the multidimensional entropy space distance is characterized in that the calculating of the permutation entropy of each group of sample data IMF component transformation under each working condition and the calculating of the average value of the permutation entropy of 10 groups of sample data transformation comprises the following steps:
one-dimensional discrete time series: reconstructing any element X (i) in the discrete time sequence by a phase space reconstruction delay coordinate method, and selecting continuous m sample points around the sampling point to obtain a reconstruction vector of an m-dimensional space of the sampling point X (i): x ═ { X (i), X (i +1), …, X (i + (m-1) × l) }, m and l being the dimension of the reconstruction space and the delay time, respectively;
and (3) performing corresponding ascending arrangement on the reconstructed vectors after the discrete time sequence x (i) is reconstructed to obtain: x'i＝{x(i+(j1-1)*l)≤x(i+(j2-1)*l)≤…≤x(i+(jm-1) × } results in a new permutation: i { j1,j2,…,jm}；
Counting the number of times of occurrence of various arrangement situations of the X sequence, which is equal to the m! As the probability p of occurrenceiCalculating the sequence permutation entropy:wherein: h is the permutation entropy of the transform, piFor the probability of occurrence of various permutation conditions, d1，d2Is a positive integer.
6. A rolling bearing fault identification system based on multidimensional entropy space distance, the system comprising:
the system comprises an acquisition module, a storage module and a processing module, wherein the acquisition module is used for acquiring original vibration signals of a rolling bearing under normal working conditions and different fault types and selecting 10 groups of sample data of each working condition;
the acquisition module is used for carrying out noise reduction processing on the sample data through EMD to obtain a plurality of intrinsic mode functions IMF components;
the processing module is used for selecting the first 5 IMF components of each working condition, respectively calculating the information entropy, the energy entropy and the permutation entropy of the IMF component transformation of each group of sample data of each working condition, and respectively calculating the average value of the transformation information entropy, the average value of the transformation energy entropy and the average value of the transformation permutation entropy of 10 groups of sample data;
the three-dimensional space construction module is used for constructing a three-dimensional space by taking the average value of the transformation arrangement entropy as an X axis, taking the average value of the transformation information entropy as a Y axis and taking the average value of the transformation energy entropy as a Z axis;
the determining module is used for randomly selecting a preset number of groups of test data, calculating three entropy values after transformation according to the steps, calculating a space distance with an entropy value point of each working condition in a three-dimensional space of the last step, judging the distance between the test data and each type of bearing working condition, and if the distance between the test data and the first bearing working condition is minimum, the working condition state of the rolling bearing is the first bearing working condition, and the first bearing working condition is any one of all working condition states of the rolling bearing.
7. The rolling bearing fault identification system based on multidimensional entropy space distance as claimed in claim 6, wherein the obtaining module comprises:
h1＝x(t)-m1
r1＝x(t)-c1
rn＝rn-1-cn
wherein m is1Is the mean of the upper and lower envelope lines, x (t) is the original vibration signal，h1Is the difference between the two; r is1As a residual signal, rnFor the nth residual signal, c1(t)，c2(t)，…，cn(t) is the sum of n IMF components and X (t) is the sum of n IMF components and a residual signal.
8. The rolling bearing fault identification system based on multidimensional entropy space distance of claim 7, wherein the processing module comprises:
the normalization processing unit is used for performing normalization processing on the IMF components c (t) and then performing blocking;
the statistical unit is used for counting the number of data in each block range and calculating the corresponding probability;
a determination unit for determining an average HX a of the information entropy1*sum(-p.*log2(p))-a2Where p is the probability, HX is the information entropy of the transformation, a1，a2Respectively different positive integers.
9. The rolling bearing fault identification system based on multidimensional entropy space distance of claim 8, wherein the processing module comprises:
wherein, ci(t) is the IMF component, EcIs the total energy, pi＝Ei/EcThe ratio of the ith IMF energy to the total energy, b1，b2Is a positive integer.
10. The rolling bearing fault identification system based on multidimensional entropy space distance of claim 9, wherein the processing module comprises:
a reconstruction unit for reconstructing the one-dimensional discrete-time series: reconstructing any element X (i) in the discrete time sequence by a phase space reconstruction delay coordinate method, and selecting continuous m sample points around the sampling point to obtain a reconstruction vector of an m-dimensional space of the sampling point X (i): x ═ { X (i), X (i +1), …, X (i + (m-1) × l) }, m and l being the dimension of the reconstruction space and the delay time, respectively;
an arranging unit, configured to perform corresponding ascending arrangement on the reconstructed vectors after the discrete time sequence x (i) is reconstructed, so as to obtain: x'i＝{x(i+(j1-1)*l)≤x(i+(j2-1)*l)≤…≤x(i+(jm-1) × } results in a new permutation: i { j1,j2,…,jm}；
A calculating unit for counting the occurrence times of various arrangement situations of the X sequence, which is equal to the full arrangement m! As the probability p of occurrenceiCalculating the sequence permutation entropy:
wherein: h is the permutation entropy of the transform, piFor the probability of occurrence of various permutation conditions, d1，d2Is a positive integer.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911352259.1A CN110836775B (en) | 2019-12-25 | 2019-12-25 | Rolling bearing fault identification method and system based on multi-dimensional entropy space distance |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911352259.1A CN110836775B (en) | 2019-12-25 | 2019-12-25 | Rolling bearing fault identification method and system based on multi-dimensional entropy space distance |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110836775A true CN110836775A (en) | 2020-02-25 |
| CN110836775B CN110836775B (en) | 2021-07-27 |
Family
ID=69578680
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911352259.1A Expired - Fee Related CN110836775B (en) | 2019-12-25 | 2019-12-25 | Rolling bearing fault identification method and system based on multi-dimensional entropy space distance |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110836775B (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110595745A (en) * | 2019-04-26 | 2019-12-20 | 深圳市豪视智能科技有限公司 | Detection method for abnormality of fixing screw of equipment and related product |
| CN111401168A (en) * | 2020-03-06 | 2020-07-10 | 上海神添实业有限公司 | Multi-layer radar feature extraction and selection method for unmanned aerial vehicle |
| CN112801026A (en) * | 2021-02-08 | 2021-05-14 | 上海电机学院 | Signal feature extraction method based on multi-dimensional entropy features |
| CN113029567A (en) * | 2021-02-26 | 2021-06-25 | 上海电机学院 | Fault signal diagnosis method based on feature fusion equipotential planet map |
| CN113468818A (en) * | 2021-07-13 | 2021-10-01 | 上海电气风电集团股份有限公司 | Method and device for monitoring winding running state of wind generating set |
| CN114427972A (en) * | 2022-02-09 | 2022-05-03 | 中国人民解放军战略支援部队航天工程大学士官学校 | Rolling bearing performance degradation feature extraction method and system |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106092574B (en) * | 2016-05-30 | 2018-04-17 | 西安工业大学 | Based on the Method for Bearing Fault Diagnosis for improving EMD decomposition and sensitive features selection |
| CN108168890A (en) * | 2018-02-10 | 2018-06-15 | 北京工业大学 | A kind of method of rolling bearing fault diagnosis |
| CN108760268A (en) * | 2018-06-08 | 2018-11-06 | 济南大学 | A kind of Vertical Mill operation data both phase step fault diagnostic method based on comentropy |
| CN109000926A (en) * | 2018-10-25 | 2018-12-14 | 中国计量大学 | Rolling bearing sound emission signal characteristic extracting method based on EMD and approximate entropy |
| CN108168886B (en) * | 2017-12-22 | 2019-08-30 | 合肥工业大学 | Rolling Bearing Fault Character extracts and method for monitoring operation states |
| CN110595765A (en) * | 2019-08-26 | 2019-12-20 | 西安理工大学 | Wind turbine generator gearbox fault diagnosis method based on VMD and FA _ PNN |
-
2019
- 2019-12-25 CN CN201911352259.1A patent/CN110836775B/en not_active Expired - Fee Related
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106092574B (en) * | 2016-05-30 | 2018-04-17 | 西安工业大学 | Based on the Method for Bearing Fault Diagnosis for improving EMD decomposition and sensitive features selection |
| CN108168886B (en) * | 2017-12-22 | 2019-08-30 | 合肥工业大学 | Rolling Bearing Fault Character extracts and method for monitoring operation states |
| CN108168890A (en) * | 2018-02-10 | 2018-06-15 | 北京工业大学 | A kind of method of rolling bearing fault diagnosis |
| CN108760268A (en) * | 2018-06-08 | 2018-11-06 | 济南大学 | A kind of Vertical Mill operation data both phase step fault diagnostic method based on comentropy |
| CN109000926A (en) * | 2018-10-25 | 2018-12-14 | 中国计量大学 | Rolling bearing sound emission signal characteristic extracting method based on EMD and approximate entropy |
| CN110595765A (en) * | 2019-08-26 | 2019-12-20 | 西安理工大学 | Wind turbine generator gearbox fault diagnosis method based on VMD and FA _ PNN |
Non-Patent Citations (1)
| Title |
|---|
| 孙抗等: ""基于EEMD和马氏距离的高压断路器机械故障诊断研究"", 《高压电器》 * |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110595745A (en) * | 2019-04-26 | 2019-12-20 | 深圳市豪视智能科技有限公司 | Detection method for abnormality of fixing screw of equipment and related product |
| CN110595745B (en) * | 2019-04-26 | 2021-05-07 | 深圳市豪视智能科技有限公司 | Detection method for abnormality of fixing screw of equipment and related product |
| CN111401168A (en) * | 2020-03-06 | 2020-07-10 | 上海神添实业有限公司 | Multi-layer radar feature extraction and selection method for unmanned aerial vehicle |
| CN111401168B (en) * | 2020-03-06 | 2023-11-17 | 上海神添实业有限公司 | Multilayer radar feature extraction and selection method for unmanned aerial vehicle |
| CN112801026A (en) * | 2021-02-08 | 2021-05-14 | 上海电机学院 | Signal feature extraction method based on multi-dimensional entropy features |
| CN113029567A (en) * | 2021-02-26 | 2021-06-25 | 上海电机学院 | Fault signal diagnosis method based on feature fusion equipotential planet map |
| CN113468818A (en) * | 2021-07-13 | 2021-10-01 | 上海电气风电集团股份有限公司 | Method and device for monitoring winding running state of wind generating set |
| CN114427972A (en) * | 2022-02-09 | 2022-05-03 | 中国人民解放军战略支援部队航天工程大学士官学校 | Rolling bearing performance degradation feature extraction method and system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110836775B (en) | 2021-07-27 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110836775B (en) | Rolling bearing fault identification method and system based on multi-dimensional entropy space distance | |
| CN105115594B (en) | Gear-box vibration signal fault signature extracting method based on Wavelet Entropy and information fusion | |
| CN112798280B (en) | Rolling bearing fault diagnosis method and system | |
| CN111160167A (en) | Spindle fault classification and identification method based on S-transform deep convolutional neural network | |
| CN109765055B (en) | Rolling bearing fault detection method and system based on EWT, spectrum effective value and KNN | |
| CN110118657B (en) | Rolling bearing fault diagnosis method and system based on relative entropy and K nearest neighbor algorithm | |
| Hao et al. | Underdetermined source separation of bearing faults based on optimized intrinsic characteristic-scale decomposition and local non-negative matrix factorization | |
| CN113485302B (en) | Vehicle operation process fault diagnosis method and system based on multivariate time sequence data | |
| CN111307438B (en) | Rotary machine vibration fault diagnosis method and system based on information entropy | |
| CN104807534B (en) | Equipment eigentone self study recognition methods based on on-line vibration data | |
| KR20090078075A (en) | Fault diagnosis of inductirn motors by dft and wavelet | |
| CN108549078B (en) | Cross-channel combination and detection method for radar pulse signals | |
| CN108195584B (en) | A kind of Fault Diagnosis of Roller Bearings based on accuracy spectrogram | |
| CN113468760B (en) | Motor weak fault detection method and system based on dictionary learning | |
| CN112857804B (en) | Rolling bearing fault diagnosis method, device, medium and computer equipment | |
| CN113538353B (en) | Five-phase asynchronous motor rolling bearing fault diagnosis method based on single-channel diagram data enhancement and migration training residual error network | |
| CN111458146B (en) | Rolling bearing multi-measuring-point vibration signal compression sampling and synchronous reconstruction method | |
| CN115392301A (en) | Converter transformer state identification method, converter transformer state identification device, converter transformer state identification equipment, converter transformer state identification medium and program product | |
| CN115031967A (en) | Fault diagnosis method and device for rolling bearing and electronic equipment | |
| CN114021275B (en) | Rolling bearing fault diagnosis method based on deep convolution fuzzy system | |
| CN112345247A (en) | Fault diagnosis method and device for rolling bearing | |
| CN116086807A (en) | Fault judging method for rolling bearing and related product | |
| CN113219333B (en) | Spectrum parameter processing method during motor fault diagnosis | |
| Chen et al. | Rotating machinery diagnosis using wavelet packets-fractal technology and neural networks | |
| CN116595324B (en) | Method for extracting signal transient impact starting point |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20210727 Termination date: 20211225 |