CN116698414A - Slurry pump bearing fault monitoring method based on multi-source data fusion - Google Patents
Slurry pump bearing fault monitoring method based on multi-source data fusion Download PDFInfo
- Publication number
- CN116698414A CN116698414A CN202310646135.4A CN202310646135A CN116698414A CN 116698414 A CN116698414 A CN 116698414A CN 202310646135 A CN202310646135 A CN 202310646135A CN 116698414 A CN116698414 A CN 116698414A
- Authority
- CN
- China
- Prior art keywords
- data
- bearing
- fuzzy
- layer
- slurry pump
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 42
- 230000004927 fusion Effects 0.000 title claims abstract description 34
- 238000012544 monitoring process Methods 0.000 title claims abstract description 22
- 239000002002 slurry Substances 0.000 title claims abstract description 22
- 230000009467 reduction Effects 0.000 claims abstract description 19
- 238000013528 artificial neural network Methods 0.000 claims abstract description 12
- 230000002159 abnormal effect Effects 0.000 claims abstract description 9
- 238000001914 filtration Methods 0.000 claims abstract description 8
- 238000012216 screening Methods 0.000 claims abstract description 7
- 230000003044 adaptive effect Effects 0.000 claims abstract description 6
- 238000012545 processing Methods 0.000 claims abstract description 5
- 239000011159 matrix material Substances 0.000 claims description 26
- 238000005096 rolling process Methods 0.000 claims description 21
- 239000013598 vector Substances 0.000 claims description 14
- 238000004364 calculation method Methods 0.000 claims description 8
- 230000006870 function Effects 0.000 claims description 8
- 230000015556 catabolic process Effects 0.000 claims description 3
- 238000006731 degradation reaction Methods 0.000 claims description 3
- 238000013507 mapping Methods 0.000 claims description 3
- 238000003062 neural network model Methods 0.000 claims description 3
- 238000010606 normalization Methods 0.000 claims description 3
- 238000005457 optimization Methods 0.000 claims description 3
- 238000012847 principal component analysis method Methods 0.000 claims description 3
- 230000004044 response Effects 0.000 claims description 3
- 230000011218 segmentation Effects 0.000 claims description 3
- 238000003745 diagnosis Methods 0.000 abstract description 14
- 230000000295 complement effect Effects 0.000 abstract description 5
- 238000005516 engineering process Methods 0.000 abstract description 5
- 230000008569 process Effects 0.000 abstract description 4
- 230000010354 integration Effects 0.000 description 5
- 239000007788 liquid Substances 0.000 description 5
- 238000004458 analytical method Methods 0.000 description 3
- 238000001514 detection method Methods 0.000 description 3
- 238000005259 measurement Methods 0.000 description 3
- 230000007547 defect Effects 0.000 description 2
- 239000012634 fragment Substances 0.000 description 2
- 230000001050 lubricating effect Effects 0.000 description 2
- 230000007257 malfunction Effects 0.000 description 2
- 238000012549 training Methods 0.000 description 2
- 230000008859 change Effects 0.000 description 1
- 238000001816 cooling Methods 0.000 description 1
- 230000007797 corrosion Effects 0.000 description 1
- 238000005260 corrosion Methods 0.000 description 1
- 238000005336 cracking Methods 0.000 description 1
- 238000000354 decomposition reaction Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000009776 industrial production Methods 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 238000005461 lubrication Methods 0.000 description 1
- 239000002184 metal Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000000737 periodic effect Effects 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
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)
Abstract
The invention discloses a slurry pump bearing fault monitoring method based on multi-source data fusion, which belongs to the technical field of fault monitoring and comprises the following steps: s1, acquiring vibration signals and temperature parameters of a bearing through a sensor to obtain sensing data information; s2, screening abnormal data by using a Laida criterion, and then performing data dimension reduction on the rest data; s3, performing adaptive filtering processing on the dimension reduced data; s4, calculating the weight of the sensor data by using a fuzzy integral fusion algorithm and taking out the maximum value; s5, inputting the fusion data into a fuzzy neural network to judge the running state of the bearing. By the mode, the invention applies the multi-source data fusion technology, comprehensively analyzes and processes various information of the bearing equipment, and performs fault diagnosis on the information, so that the information can furthest utilize the mutual complementary information in the multi-source data; to a certain extent, the characteristics of the fault are improved, and the fault is diagnosed.
Description
Technical Field
The invention relates to the technical field of fault monitoring, in particular to a slurry pump bearing fault monitoring method based on multi-source data fusion.
Background
Bearings are an important component of rotary machines and have numerous advantages, such as rapid lubrication and cooling, high efficiency, etc., and are therefore widely used in the machine industry. During the operation of the machine, the outer and inner rings in the bearing often have problems of wear and fracture, and if these faults are not handled timely and effectively, the bearing eventually fails. In the fault detection of mechanical equipment, part of the equipment faults are caused by bearing faults, so the fault state monitoring research of the rolling bearing has been paid attention to. Bearings have become one of the indispensable components in mechanical equipment as a driving device, and play a large role in the operation of mechanical equipment, so that the bearings are liable to undergo wear, cracking, peeling and other failures during periodic load operation and impact. In rolling bearings, most of the bearing failures are manifested as local defects on the inner and outer race raceways, and the operating state of the rolling bearing directly affects the performance of the overall machine.
At present, the prior art mainly detects bearing faults through a vibration signal analysis method and an oil liquid detection method. The vibration signal analysis method is to obtain vibration signals of equipment and amplified fault time domain signals by using contact measurement, further perform frequency domain analysis, and determine characteristic values of different fault states of the slurry pump bearing according to fault characteristic frequencies of related bearing components so as to realize fault diagnosis. Because the composition of the lubricating liquid can change to a certain extent in the long-term running process of the bearing, and some worn metal fragments can be generated at the same time, the oil liquid detection method is to collect a lubricating liquid sample of the bearing and evaluate the running state of the bearing according to a plurality of indexes such as the oil liquid composition in the sample, the shape, the size, the number and the like of the fragments, thereby realizing fault diagnosis.
The traditional bearing fault diagnosis generally adopts a single signal source and a single model, and equipment fault diagnosis is realized by utilizing characteristic decomposition of vibration signals. In the case of using only one signal source, although a malfunction of the mechanical device can be determined, in many cases, the diagnosis result obtained is not reliable. This is because of the complexity and variety of the malfunctions of the mechanical equipment. For example, in high temperature and high corrosion environments, contact measurement of vibration signals is not satisfactory for industrial production.
Based on the problems, the invention designs a slurry pump bearing fault monitoring method based on multi-source data fusion to solve the problems.
Disclosure of Invention
Aiming at the defects existing in the prior art, the invention provides a slurry pump bearing fault monitoring method based on multi-source data fusion.
In order to achieve the above purpose, the invention is realized by the following technical scheme:
a slurry pump bearing fault monitoring method based on multi-source data fusion comprises the following steps:
s1, acquiring vibration signals and temperature parameters of a bearing through a sensor to obtain sensing data information;
s2, screening abnormal data by using a Laida criterion, and then performing data dimension reduction on the rest data;
s3, performing adaptive filtering processing on the dimension reduced data;
s4, calculating the weight of the sensor data by using a fuzzy integral fusion algorithm and taking out the maximum value;
s5, inputting the fusion data into a fuzzy neural network to judge the running state of the bearing.
In step S1, a temperature sensor is used to collect a temperature parameter, i.e. a temperature signal T, to reflect the internal running state of the bearing; the acquisition of the vibration signal of the bearing is to carry out numerical calculation on continuous signals within a certain period of time and analyze the degradation condition of the bearing by utilizing the acquired sample, wherein the failure frequency of the bearing is as follows:
wherein ,f0 For the failure frequency of the outer ring, f 1 For the failure frequency of the inner ring, f 2 For the failure frequency of the rolling bodies, f r The rolling element bearing is characterized in that the rolling element bearing is a high-speed shaft rotating frequency, D is the average diameter of the rolling elements, D is the pitch diameter of the bearing, alpha is the contact angle of the bearing, and x is the number of the rolling elements.
Further, the specific step S2 is as follows:
1. comparing the standard deviation with the sample error to judge whether the data is an abnormal value;
2. performing data dimension reduction on the rest data by using a principal component analysis method;
the calculation formula is as follows:
Y＝A'X；
wherein Y is the data after dimension reduction, X is the acquired original data, A' is the feature vector corresponding to the first X larger feature values, a new matrix is formed, and X, Y is composed of the main components of the dimension X.
Further, step one, comparing the standard deviation with the sample error to determine whether the data is an outlier; the method comprises the following specific steps: by taking the obtained vibration frequency and temperature as raw data, whenIf yes, then determine f i Is an outlier; when->If yes, then determine f i Is a normal value, should be preserved;
wherein ,fi The value of the ith data in the sample size, θ is the standard deviation,is the arithmetic mean of the sample volumes.
Further, the specific step S3 is as follows:
1. firstly, initializing a dimensionality reduction data;
noise cancellation in reduced-dimension data by adaptive filtering (RLS) algorithmThe method comprises the steps of carrying out a first treatment on the surface of the First, initializing the dimension reduction data, taking w (0) =0, and a (0) =m -1 I；
Wherein w is a weight, m is a smaller positive number, A is an autocorrelation matrix R xx The inverse of (n);
2. then inputting the data into a filter;
the method comprises the following specific steps: inputting the data into a filter to obtain d (n) and x (n);
wherein x (n) is an input signal vector and d (n) is a desired response signal;
3. and finally, acquiring weight vectors according to the data processed by the algorithm and updating the matrix.
Further, step three, finally, acquiring weight vectors according to the data processed by the algorithm and updating the matrix; the method comprises the following specific steps: acquiring weight vectors according to the data processed by the algorithm and updating a matrix A (n);
w(n)'＝w(n-1)+g(n)[d(n)-x T (n)w(n-1)]；
wherein w (n)' is an iterative filter parameter, and w (n-1) is a weight value at the moment of n-1;
A(n)＝λ -1 [A(n-1)-g(n)x T (n)A(n-1)]；
where g (n) is the gain factor, λ is the forgetting factor and <1.
Further, the specific step S4 is as follows:
1. fusing any fault data of the bearing to create a decision matrix;
2. after the fuzzy density is initialized, optimizing according to a decision matrix;
3. calculating the fuzzy measure by using the calculated fuzzy density;
wherein, the fuzzy measure Q (f U T) =Q (f) +Q (T) +aQ (f) Q (T), a is an arbitrary constant, Q (f) is the frequency fuzzy density, and Q (T) is the temperature fuzzy density;
4. through all fuzzy integration, the corresponding largest product is selectedThe fraction ε (f) i ) MAX As an output signal;
wherein ,ε(f i ) The output signal is represented by f (i/k) as i/k time integral, f (i/k-1) as i/k-1 time integral, Q (R) i/k ) The definition domain of i/k is denoted i=1, 2, 3..k=1, 2, 3..;
further, after initializing the fuzzy density, optimizing according to a decision matrix; the method comprises the following steps: for blur density P i/j After initialization, optimization is performed according to the decision matrix
wherein ,sigma and theta are set parameter values, +.>Representing the calculated blur density, M is the number of classifiers, i=1, 2, 3..j=1, 2, 3..k=1, 2, 3..n=1, 2, 3.;
further comprises: according to the corrected blurring densityInitializing the density again, and calculating sigma and theta;
wherein ,σ and θ are set parameter values, τ represents a fixed constant, i=1, 2,3., j=1, 2,3., k=1, 2,3., n=1, 2, 3.;
further, the specific step S5 is as follows:
establishing up-down connection for each fault signal by using a fuzzy neural network to establish a mapping relation between fusion data and bearing faults, and constructing a fuzzy neural network model to define a five-layer network;
wherein the first layer is an input layer and is used for converting f 0 Frequency of outer ring failure, f 1 Frequency of inner ring failure, f 2 Frequency of rolling element failure, f r The rotation frequency of the high-speed shaft where the bearing is positioned is input to the next layer, and the weight coefficient is
The second layer is a membership function layer, the input signals transmitted by the first layer are calculated to be relevant membership degree, and the weight coefficient is
The third layer is a release layer of fuzzy rule intensity, and multiplies each input signal after fuzzification;
the fourth layer is a rule intensity normalization layer, calculates the credibility of the rule, and the weight coefficient is
The fifth layer is used for calculating the total output of the input signals, and the weight coefficient is
t i and yi Representing the desired output and the actual output, respectively;
updating parameters by using error back propagation, wherein a learning algorithm for parameter adjustment is as follows:
wherein ,wi Weight coefficient representing network, m i Representing the fuzzy segmentation number of the network, C ij Andthe center and width of the membership function are represented, respectively, and β represents the learning rate.
Advantageous effects
The invention acquires the vibration signal and the temperature parameter of the bearing through the sensor to obtain the sensing data information; then screening and dimension reducing are carried out on the original data; calculating the weight of the sensor data by using a fuzzy integration algorithm and taking out the maximum value; and judging the running state of the bearing through a fuzzy neural network.
The invention applies the multi-source data fusion technology to comprehensively analyze and process various information of the bearing equipment and perform fault diagnosis on the information, so that the mutual complementary information in the multi-source data can be utilized to the maximum extent; to a certain extent, the characteristics of the fault are improved, and the fault is diagnosed.
The invention can monitor the running state of the slurry pump well; by adopting the multi-source data fusion technology, various information of the mechanical equipment can be comprehensively analyzed and processed and then used for fault diagnosis, the complementary fault information in the multi-source data is fully utilized, the diversity of fault characteristics is increased, the fault diagnosis is carried out, and the reliability of the diagnosis result can be increased.
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 evident that the drawings in the following description are only some embodiments of the present invention and that other drawings may be obtained from these drawings without inventive effort for a person of ordinary skill in the art.
FIG. 1 is a flow chart of a method for monitoring bearing faults of a slurry pump based on multi-source data fusion;
fig. 2 is a flowchart of a fuzzy integration algorithm in a method for monitoring bearing faults of a slurry pump based on multi-source data integration.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention more clear, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. It will be apparent that the described embodiments are some, but not all, embodiments of the invention. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
The invention is further described below with reference to examples.
Example 1
Referring to fig. 1-2 of the specification, a method for monitoring bearing faults of a slurry pump based on multi-source data fusion comprises the following steps:
s1, acquiring vibration signals and temperature parameters of a bearing through a sensor to obtain sensing data information;
the temperature parameter, namely a temperature signal T, is acquired by a temperature sensor and reflects the internal running state of the bearing; the acquisition of the vibration signal of the bearing is to carry out numerical calculation on continuous signals within a certain period of time and analyze the degradation condition of the bearing by utilizing the acquired sample, wherein the failure frequency of the bearing is as follows:
wherein ,f0 For the failure frequency of the outer ring, f 1 For the failure frequency of the inner ring, f 2 For the failure frequency of the rolling bodies, f r The rolling element bearing is characterized in that the rolling element bearing is a high-speed shaft with a rotating frequency, D is the average diameter of the rolling elements, D is the pitch diameter of the bearing, alpha is the contact angle of the bearing, and x is the number of the rolling elements;
s2, when data preprocessing is carried out, as the data volume is large, the measurement result has great difference from the actual data, and the data of the sensor are screened out by applying the Laida criterion; after screening abnormal data by using the Laida criterion, performing data dimension reduction on the rest data;
the method comprises the following specific steps:
1. comparing the standard deviation with the sample error to judge whether the data is an abnormal value;
the method comprises the following specific steps: the obtained vibration frequency and temperature are used as raw data to discriminate obvious problems or abnormal data whenIf yes, then determine f i Is an outlier; when->If yes, then determine f i Is a normal value, should be preserved;
wherein ,fi The value of the ith data in the sample size, θ is the standard deviation,is the arithmetic mean of the sample volumes;
2. performing data dimension reduction on the rest data by using a principal component analysis method;
the calculation formula is as follows:
Y＝A'X；
wherein Y is the data after dimension reduction, X is the acquired original data, A' is the feature vector corresponding to the first X larger feature values to form a new matrix, and X, Y is composed of the main components of the dimension X;
s3, after data screening and dimension reduction processing, noise signals in the data acquired by the sensor cannot be completely removed, so that the dimension reduced data is subjected to self-adaptive filtering processing;
the method comprises the following specific steps:
1. firstly, initializing a dimensionality reduction data;
removing noise in the dimension reduction data through an adaptive filtering (RLS) algorithm; first, initializing the dimension reduction data, taking w (0) =0, and a (0) =m -1 I；
Wherein w is a weight, m is a smaller positive number, A is an autocorrelation matrix R xx The inverse of (n);
2. then inputting the data into a filter;
the method comprises the following specific steps: inputting the data into a filter to obtain d (n) and x (n);
wherein x (n) is an input signal vector and d (n) is a desired response signal;
3. finally, acquiring a weight vector according to the data processed by the algorithm and updating a matrix;
the method comprises the following specific steps: acquiring weight vectors according to the data processed by the algorithm and updating a matrix A (n);
w(n)'＝w(n-1)+g(n)[d(n)-x T (n)w(n-1)]；
wherein w (n)' is an iterative filter parameter, and w (n-1) is a weight value at the moment of n-1;
A(n)＝λ -1 [A(n-1)-g(n)x T (n)A(n-1)]；
where g (n) is the gain factor, λ is the forgetting factor and <1, smaller λ means better tracking performance for non-stationarity of the signal;
according to the method, continuous self-adaptive filtering is carried out on the vibration signal and the temperature signal, so that the pretreatment of the signals is realized;
s4, calculating the weight of the sensor data by using a fuzzy integral fusion algorithm and taking out the maximum value;
as shown in fig. 2, fault signals acquired by the sensors are fused through a classifier and a decision matrix is established, all fuzzy integrals are calculated, and the maximum value is taken out;
the method comprises the following specific steps:
1. fusing any fault data of the bearing to create a decision matrix;
2. after the fuzzy density is initialized, optimizing according to a decision matrix;
the method comprises the following steps: for blur density P i/j After initialization, optimization is performed according to the decision matrix
wherein ,sigma and theta are set parameter values, +.>Representing the calculated blur density, M is the number of classifiers, i=1, 2, 3..j=1, 2, 3..k=1, 2, 3..n=1, 2, 3.;
further comprises: according to the corrected blurring densityInitializing the density again, and calculating sigma and theta;
wherein ,σ and θ are set parameter values, τ represents a fixed constant, i=1, 2,3., j=1, 2,3., k=1, 2,3., n=1, 2, 3.;
3. calculating the fuzzy measure by using the calculated fuzzy density;
wherein, the fuzzy measure Q (fuT) =q (f) +q (T) +aq (f) Q (T), a is an arbitrary constant, Q (f) is the frequency fuzzy density, and Q (T) is the temperature fuzzy density;
4. the corresponding maximum integral epsilon (f i ) MAX As an output signal;
wherein ,ε(f i ) The output signal is represented by f (i/k) as i/k time integral, f (i/k-1) as i/k-1 time integral, Q (R) i I/k) represents a domain of i/k, i=1, 2,3., k=1, 2, 3.;
s5, inputting the fusion data into a fuzzy neural network to judge the running state of the bearing;
the fused data is used as an input end of the fuzzy neural network, and the running state of the bearing is judged according to the result output by the fuzzy neural network and the parameters of the bearing during normal running through network training;
the method comprises the following specific steps:
establishing up-down connection for each fault signal by using a fuzzy neural network to establish a mapping relation between fusion data and bearing faults, and constructing a fuzzy neural network model to define a five-layer network;
wherein the first layer is an input layer and is used for converting f 0 Frequency of outer ring failure, f 1 Frequency of inner ring failure, f 2 Frequency of rolling element failure, f r The rotation frequency of the high-speed shaft where the bearing is positioned is input to the next layer, and the weight coefficient is
The second layer is a membership function layer, the input signals transmitted by the first layer are calculated to be relevant membership degree, and the weight coefficient is
The third layer is a release layer of fuzzy rule intensity, and multiplies each input signal after fuzzification;
the fourth layer is a rule intensity normalization layer, calculates the credibility of the rule, and the weight coefficient is
The fifth layer is used for calculating the total output of the input signals, and the weight coefficient is
t i and yi Representing the desired output and the actual output, respectively;
the loss function is used for measuring the degree of the prediction error, and the square loss function E which is convenient to calculate is selected because of the limited number of sample training sets, and the square difference loss is amplified by the distance between the predicted value and the true value in calculation, so that larger punishment is given to the output with larger error, and the calculation of error gradient is facilitated; updating parameters by using error back propagation, wherein a learning algorithm for parameter adjustment is as follows:
wherein ,wi Weight coefficient representing network, m i Representing the fuzzy segmentation number of the network, C ij Andthe center and width of the membership function are represented, respectively, and β represents the learning rate.
Judging according to the result output by the fuzzy neural network and parameters of the bearing during normal operation, and determining whether the bearing has faults or not;
the invention acquires the vibration signal and the temperature parameter of the bearing through the sensor to obtain the sensing data information; then screening and dimension reducing are carried out on the original data; calculating the weight of the sensor data by using a fuzzy integration algorithm and taking out the maximum value; judging the running state of the bearing through a fuzzy neural network;
the invention applies the multi-source data fusion technology to comprehensively analyze and process various information of the bearing equipment and perform fault diagnosis on the information, so that the mutual complementary information in the multi-source data can be utilized to the maximum extent; to a certain extent, the characteristics of the fault are improved, and the fault is diagnosed.
The invention can monitor the running state of the slurry pump well; by adopting the multi-source data fusion technology, various information of the mechanical equipment can be comprehensively analyzed and processed and then used for fault diagnosis, the complementary fault information in the multi-source data is fully utilized, the diversity of fault characteristics is increased, the fault diagnosis is carried out, and the reliability of the diagnosis result can be increased.
The above embodiments are only for illustrating the technical solution of the present invention, and are not limiting; although the invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit and scope of the technical solutions of the embodiments of the present invention.
Claims (9)
1. A slurry pump bearing fault monitoring method based on multi-source data fusion is characterized by comprising the following steps:
s1, acquiring vibration signals and temperature parameters of a bearing through a sensor to obtain sensing data information;
s2, screening abnormal data by using a Laida criterion, and then performing data dimension reduction on the rest data;
s3, performing adaptive filtering processing on the dimension reduced data;
s4, calculating the weight of the sensor data by using a fuzzy integral fusion algorithm and taking out the maximum value;
s5, inputting the fusion data into a fuzzy neural network to judge the running state of the bearing.
2. The method for monitoring bearing faults of a slurry pump based on multi-source data fusion according to claim 1, wherein in the step S1, temperature parameters, namely temperature signals T, are collected by a temperature sensor to reflect the internal running state of the bearing; the acquisition of the vibration signal of the bearing is to carry out numerical calculation on continuous signals within a certain period of time and analyze the degradation condition of the bearing by utilizing the acquired sample, wherein the failure frequency of the bearing is as follows:
wherein ,f0 For the failure frequency of the outer ring, f 1 For the failure frequency of the inner ring, f 2 For the failure frequency of the rolling bodies, f r The rolling element bearing is characterized in that the rolling element bearing is a high-speed shaft rotating frequency, D is the average diameter of the rolling elements, D is the pitch diameter of the bearing, alpha is the contact angle of the bearing, and x is the number of the rolling elements.
3. The method for monitoring bearing faults of a slurry pump based on multi-source data fusion according to claim 1, wherein the specific steps of S2 are as follows:
1. comparing the standard deviation with the sample error to judge whether the data is an abnormal value;
2. performing data dimension reduction on the rest data by using a principal component analysis method;
the calculation formula is as follows:
Y＝A′X；
wherein Y is the data after dimension reduction, X is the acquired original data, A' is the feature vector corresponding to the first X larger feature values, a new matrix is formed, and X, Y is composed of the main components of the dimension X.
4. The method for monitoring bearing faults of a slurry pump based on multi-source data fusion according to claim 3, wherein step one, comparing standard deviation with sample error to judge whether data is abnormal value; the method comprises the following specific steps: by taking the obtained vibration frequency and temperature as raw data, whenIf yes, then determine f i Is an outlier; when->If yes, then determine f i Is a normal value, should be preserved;
wherein ,fi The value of the ith data in the sample size, θ is the standard deviation,is the arithmetic mean of the sample volumes.
5. The method for monitoring bearing faults of a slurry pump based on multi-source data fusion according to claim 1, wherein the specific step of S3 is as follows:
1. firstly, initializing a dimensionality reduction data;
removing noise in the dimension reduction data through an adaptive filtering (RLS) algorithm; first, initializing the dimension reduction data, taking w (0) =0, and a (0) =m -1 I；
Wherein w is a weight, m is a smaller positive number, A is an autocorrelation matrix R xx The inverse of (n);
2. then inputting the data into a filter;
the method comprises the following specific steps: inputting the data into a filter to obtain d (n) and x (n);
wherein x (n) is an input signal vector and d (n) is a desired response signal;
3. and finally, acquiring weight vectors according to the data processed by the algorithm and updating the matrix.
6. The method for monitoring bearing faults of a slurry pump based on multi-source data fusion according to claim 5 and characterized in that step three, weight vectors are finally obtained according to data processed by an algorithm and a matrix is updated; the method comprises the following specific steps: acquiring weight vectors according to the data processed by the algorithm and updating a matrix A (n);
w(n)'＝w(n-1)+g(n)[d(n)-x T (n)w(n-1)]；
wherein w (n)' is an iterative filter parameter, and w (n-1) is a weight value at the moment of n-1;
A(n)＝λ -1 [A(n-1)-g(n)x T (n)A(n-1)]；
where g (n) is the gain factor, λ is the forgetting factor and <1.
7. The method for monitoring bearing faults of a slurry pump based on multi-source data fusion according to claim 1, wherein the specific step of S4 is as follows:
1. fusing any fault data of the bearing to create a decision matrix;
2. after the fuzzy density is initialized, optimizing according to a decision matrix;
3. calculating the fuzzy measure by using the calculated fuzzy density;
wherein, the fuzzy measure Q (f U T) =Q (f) +Q (T) +aQ (f) Q (T), a is an arbitrary constant, Q (f) is the frequency fuzzy density, and Q (T) is the temperature fuzzy density;
4. the corresponding maximum integral epsilon (f i ) MAX As an output signal;
wherein ,ε(f i ) The output signal is represented by f (i/k) as i/k time integral, f (i/k-1) as i/k-1 time integral, Q (R) i/k ) Representing the definition field of i/k.
8. The method for monitoring bearing faults of a slurry pump based on multi-source data fusion according to claim 7, and characterized in that the method comprises the following steps of optimizing according to a decision matrix after the density of fuzzy is initialized; the method comprises the following steps: for blur density P i/j After initialization, optimization is performed according to the decision matrix
wherein , sigma and theta are set parameter values, +.>Representing the calculated fuzzy density, wherein M is the number of classifiers;
further comprises: according to the corrected blurring densityInitializing the density again, and calculating sigma and theta;
wherein , sigma and theta are set parameter values, and τ represents a fixed constant.
9. The method for monitoring bearing faults of a slurry pump based on multi-source data fusion according to claim 7, wherein the specific step of S5 is as follows:
establishing up-down connection for each fault signal by using a fuzzy neural network to establish a mapping relation between fusion data and bearing faults, and constructing a fuzzy neural network model to define a five-layer network;
wherein the first layer is an input layer and is used for converting f 0 Frequency of outer ring failure, f 1 Frequency of inner ring failure, f 2 Frequency of rolling element failure, f r The rotation frequency of the high-speed shaft where the bearing is positioned is input to the next layer, and the weight coefficient is
The second layer is a membership function layer, the input signals transmitted by the first layer are calculated to be relevant membership degree, and the weight coefficient is
The third layer is a release layer of fuzzy rule intensity, and multiplies each input signal after fuzzification;
the fourth layer is a rule intensity normalization layer, calculates the credibility of the rule, and the weight coefficient is
The fifth layer is used for calculating the total output of the input signals, and the weight coefficient is
t i and yi Representing the desired output and the actual output, respectively;
updating parameters by using error back propagation, wherein a learning algorithm for parameter adjustment is as follows:
wherein ,wi Weight coefficient representing network, m i Representing the fuzzy segmentation number of the network, C ij Andthe center and width of the membership function are represented, respectively, and β represents the learning rate.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310646135.4A CN116698414A (en) | 2023-06-02 | 2023-06-02 | Slurry pump bearing fault monitoring method based on multi-source data fusion |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310646135.4A CN116698414A (en) | 2023-06-02 | 2023-06-02 | Slurry pump bearing fault monitoring method based on multi-source data fusion |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116698414A true CN116698414A (en) | 2023-09-05 |
Family
ID=87828577
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310646135.4A Pending CN116698414A (en) | 2023-06-02 | 2023-06-02 | Slurry pump bearing fault monitoring method based on multi-source data fusion |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116698414A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117329135A (en) * | 2023-10-13 | 2024-01-02 | 山东中探机械有限公司 | Method for monitoring running state of slurry pump based on data analysis |
| CN117331921A (en) * | 2023-09-28 | 2024-01-02 | 石家庄铁道大学 | Bearing monitoring multisource data processing method |
| CN117969092A (en) * | 2024-03-29 | 2024-05-03 | 山东天工岩土工程设备有限公司 | Fault detection method, equipment and medium for main bearing of shield tunneling machine |
| CN117990161A (en) * | 2024-04-03 | 2024-05-07 | 沈阳众创高科节能电机技术有限公司 | Mosaic type fusion sensing device and monitoring system for winding immersed motor |
-
2023
- 2023-06-02 CN CN202310646135.4A patent/CN116698414A/en active Pending
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117331921A (en) * | 2023-09-28 | 2024-01-02 | 石家庄铁道大学 | Bearing monitoring multisource data processing method |
| CN117329135A (en) * | 2023-10-13 | 2024-01-02 | 山东中探机械有限公司 | Method for monitoring running state of slurry pump based on data analysis |
| CN117329135B (en) * | 2023-10-13 | 2024-04-05 | 山东中探机械有限公司 | Method for monitoring running state of slurry pump based on data analysis |
| CN117969092A (en) * | 2024-03-29 | 2024-05-03 | 山东天工岩土工程设备有限公司 | Fault detection method, equipment and medium for main bearing of shield tunneling machine |
| CN117990161A (en) * | 2024-04-03 | 2024-05-07 | 沈阳众创高科节能电机技术有限公司 | Mosaic type fusion sensing device and monitoring system for winding immersed motor |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN116698414A (en) | Slurry pump bearing fault monitoring method based on multi-source data fusion | |
| Ren et al. | A novel model with the ability of few-shot learning and quick updating for intelligent fault diagnosis | |
| Chen et al. | Deep transfer learning for bearing fault diagnosis: A systematic review since 2016 | |
| Zhang et al. | Intelligent fault diagnosis under varying working conditions based on domain adaptive convolutional neural networks | |
| Kundu et al. | Multiple failure behaviors identification and remaining useful life prediction of ball bearings | |
| Jin et al. | Automatic feature extraction of waveform signals for in-process diagnostic performance improvement | |
| CN113392881B (en) | Rotary machine fault diagnosis method based on transfer learning | |
| Zhang et al. | Deep convolutional neural network using transfer learning for fault diagnosis | |
| CN114358124B (en) | New fault diagnosis method for rotary machinery based on deep countermeasure convolutional neural network | |
| Jiang et al. | Dual residual attention network for remaining useful life prediction of bearings | |
| Yan et al. | Fault diagnosis of rolling-element bearing using multiscale pattern gradient spectrum entropy coupled with Laplacian score | |
| Tian et al. | Deep learning-based open set multi-source domain adaptation with complementary transferability metric for mechanical fault diagnosis | |
| Joung et al. | Development and application of a method for real time motor fault detection | |
| Lazzerini et al. | Classifier ensembles to improve the robustness to noise of bearing fault diagnosis | |
| Zhao et al. | A fault diagnosis method for rotating machinery based on CNN with mixed information | |
| Sharma et al. | Use of feature ranking techniques for defect severity estimation of rolling element bearings | |
| CN113496061A (en) | SOM network clustering electromechanical device bearing fault analysis method based on transfer learning and manifold distance | |
| CN114486262B (en) | CNN-AT-LSTM-based bearing residual service life prediction method | |
| CN115146675B (en) | Rotary machine migration diagnosis method under variable working condition of depth multi-feature dynamic countermeasure | |
| CN115795350A (en) | Abnormal data information processing method in production process of blood rheology test cup | |
| CN114705427A (en) | Rolling bearing intelligent fault diagnosis method under strong noise environment | |
| Khayyat | ANN based Intelligent Mechanical Engineering Design: A Review | |
| Huang et al. | Bearing Fault Diagnosis With Incomplete Training Data: Fault Data With Partial Diameters | |
| Zhang et al. | Fault prediction model for rolling bearings based on double adaptive sliding time windows | |
| Shashank et al. | Intelligent fault diagnosis of Rolling-element bearing using 1-D Deep Convolutional Neural Networks. |
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 |