CN105335698A - Gear failure diagnosis method based on adaptive genetic algorithm and SOM (Self-Organizing Map) network - Google Patents
Gear failure diagnosis method based on adaptive genetic algorithm and SOM (Self-Organizing Map) network Download PDFInfo
- Publication number
- CN105335698A CN105335698A CN201510610800.XA CN201510610800A CN105335698A CN 105335698 A CN105335698 A CN 105335698A CN 201510610800 A CN201510610800 A CN 201510610800A CN 105335698 A CN105335698 A CN 105335698A
- Authority
- CN
- China
- Prior art keywords
- som network
- gear
- vibration signal
- neuron
- follows
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/08—Feature extraction
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
- G01M13/00—Testing of machine parts
- G01M13/02—Gearings; Transmission mechanisms
- G01M13/021—Gearings
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
- G01M13/00—Testing of machine parts
- G01M13/02—Gearings; Transmission mechanisms
- G01M13/028—Acoustic or vibration analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/12—Classification; Matching
Landscapes
- Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Theoretical Computer Science (AREA)
- Acoustics & Sound (AREA)
- Artificial Intelligence (AREA)
- Signal Processing (AREA)
- General Engineering & Computer Science (AREA)
- Testing Of Devices, Machine Parts, Or Other Structures Thereof (AREA)
Abstract
The invention provides a gear failure diagnosis method based on an adaptive genetic algorithm and an SOM (Self-Organizing Map) network. The method comprises the following steps: obtaining a vibration signal of a gear, carrying out wavelet packet analysis on the vibration signal, extracting the feature vectors of the vibration signal, dividing the feature vectors into training data and test data, firstly, utilizing the training data to train the SOM network optimized by the adaptive genetic algorithm, continuously updating the weight and the threshold value of the SOM network until an error output by the SOM network meets an accuracy requirement or achieves a maximum iteration, then, adopting the trained SOM network to diagnose a fault type of the test data, and outputting a fault diagnosis result of the gear. The gear failure diagnosis method has the characteristics of being high in precision, high in reliability and the like, and can be widely applied to the field of the fault diagnosis of mechanical equipment.
Description
Technical field
The present invention relates to fault diagnosis technology field, particularly relate to a kind of gear failure diagnosing method based on self-adapted genetic algorithm and SOM network.
Background technology
Gear is as the parts of transmission power requisite in plant equipment kinematic train, and its running status directly has influence on the work efficiency of whole plant equipment, reliability and life-span.For the troubleshooting issue of gear, at present, Chinese scholars proposes different fault diagnosis method.The methods such as main employing rough set theory, support vector machine, Bayes Method, fuzzy logic, neural network carry out fault diagnosis to gear.Rough set theory has larger superiority on the fuzzy and uncertain information of process, but its decision rule is very unstable, and accuracy is poor, and is based on complete infosystem, during process data, often can run into loss of data phenomenon.Support vector machine has advantage in solution small sample, non-linear and high dimensional pattern identification problem, but recognition capability is subject to inherent parameters impact.Bayes Method needs known difference probability definitely, and in fact can not provide definite difference probability.Fuzzy logic needs certain priori, has stronger dependence to Selecting parameter.Neural network has simple structure and very strong problem solving ability, and can process noise data preferably, but algorithm exists local optimum problem, and convergence is poor, limited reliability.
As can be seen here, in the prior art, there is the problems such as precision is low, poor reliability in gear failure diagnosing method.
Summary of the invention
In view of this, fundamental purpose of the present invention is the gear failure diagnosing method providing a kind of high precision, good reliability.
In order to achieve the above object, the technical scheme that the present invention proposes is:
Based on a gear failure diagnosing method for self-adapted genetic algorithm and SOM network, described gear failure diagnosing method comprises the steps:
Step 1, adopt acceleration transducer to gather gear case speed change system data, obtain the vibration signal of normal, tooth surface abrasion, flank of tooth cut and broken teeth four kinds of state gears.
Step 2, respectively employing db5 small echo carry out three layers of WAVELET PACKET DECOMPOSITION and reconstruct to the vibration signal of described gear, and with third layer 8 frequency bands separately energy for element structural attitude vector, extract the fault signature of gear, obtain the proper vector T of the vibration signal of described gear.
Step 3, using the neuron of the proper vector T of the vibration signal of described gear as SOM network input layer, the input layer of SOM network is one dimension, the output layer of SOM network is a two-dimensional network having 6 × 6 output neurons, the neuron of output layer lines up a neighbour structure, each neuron connects with other neuron side direction around it, and each input neuron is connected to all output neurons: the initial connection weights of SOM network are IW=[w
1, w
2..., w
8]
36 × 8, the initial threshold of SOM network is
wherein, w
1, w
2... w
8for less non-zero random number.
The coded system of step 4, self-adapted genetic algorithm adopts real coding mode, and each individuality is the chromosome of a real number string composition, is the initial population of some chromosome composition by the initial connection weights of SOM network and threshold coding.
Step 5, adopt the inverse of error sum of squares as the fitness function of each individuality, concrete calculating formula is as follows:
Wherein, f
ibe i-th individual fitness value, y
ireal output value, F is error sum of squares,
for idea output, according to selected fitness function, calculate the fitness value of each individuality.
Step 6, fitness value according to each individuality, select the high chromosome of ideal adaptation angle value to carry out copying, crossover and mutation operation, produces new population.
Step 7, the SOM network weight obtaining optimum and threshold value.
Step 8, SOM network learnt and trains, upgrading weights and threshold.
Whether step 9, discrimination precision meet the demands or reach maximum iteration time and set up; If set up, then perform step 10; If be false, then perform step 8.
Step 10, using the SOM network after adopting self-adapted genetic algorithm to optimize as sorter, fault diagnosis is carried out to the gear of normal, tooth surface abrasion, flank of tooth cut and broken teeth four kinds of states, and exports diagnostic result.
In sum, gear failure diagnosing method based on self-adapted genetic algorithm and SOM network of the present invention by the vibration signal of described gear that obtains after wavelet packet analysis and reconstruct, obtain the proper vector of the vibration signal of described gear, proper vector is divided into training data and test data, SOM network after first utilizing training data to train employing self-adapted genetic algorithm to optimize, constantly update the weights and threshold of SOM network, until the error that SOM network exports meets accuracy requirement or reaches maximum iteration time, then the SOM network trained is adopted to remove the fault type of diagnostic test data, the fault diagnosis result of output gear, thus improve precision and the reliability of Gear Fault Diagnosis.
Accompanying drawing explanation
Fig. 1 is the process flow diagram of a kind of gear failure diagnosing method based on self-adapted genetic algorithm and SOM network of the present invention.
Fig. 2 is that proper vector of the present invention solves schematic diagram.
Fig. 3 is SOM network topology structure schematic diagram of the present invention.
Fig. 4 is the SOM network fault diagnosis process flow diagram of the embodiment of the present invention.
Fig. 5 is the time domain beamformer of four kinds of state Gearbox vibration signals that the embodiment of the present invention gathers.
Embodiment
For making the object, technical solutions and advantages of the present invention clearly, the present invention is described in further detail below in conjunction with the accompanying drawings and the specific embodiments.
Fig. 1 is the process flow diagram of a kind of gear failure diagnosing method based on self-adapted genetic algorithm and SOM network of the present invention.As shown in Figure 1, gear failure diagnosing method of the present invention, comprises the steps:
Step 1, adopt acceleration transducer to gather gear case speed change system data, obtain the vibration signal of normal, tooth surface abrasion, flank of tooth cut and broken teeth four kinds of state gears.
Step 2, respectively employing db5 small echo carry out three layers of WAVELET PACKET DECOMPOSITION and reconstruct to the vibration signal of described gear, and with third layer 8 frequency bands separately energy for element structural attitude vector, extract the fault signature of gear, obtain the proper vector T of the vibration signal of described gear.
Step 3, using the neuron of the proper vector T of the vibration signal of described gear as SOM network input layer, the input layer of SOM network is one dimension, the output layer of SOM network is a two-dimensional network having 6 × 6 output neurons, the neuron of output layer lines up a neighbour structure, each neuron connects with other neuron side direction around it, and each input neuron is connected to all output neurons; The initial connection weights of SOM network are IW=[w
1, w
2..., w
8]
36 × 8, the initial threshold of SOM network is
wherein, w
1, w
2... w
8for less non-zero random number.
The coded system of step 4, self-adapted genetic algorithm adopts real coding mode, and each individuality is the chromosome of a real number string composition, is the initial population of some chromosome composition by the initial connection weights of SOM network and threshold coding.
Step 5, adopt the inverse of error sum of squares as the fitness function of each individuality, concrete calculating formula is as follows:
Wherein, f
ibe i-th individual fitness value, y
ireal output value, F is error sum of squares,
for idea output, according to selected fitness function, calculate the fitness value of each individuality.
Step 6, fitness value according to each individuality, select the high chromosome of ideal adaptation angle value to carry out copying, crossover and mutation operation, produces new population.
Step 7, the SOM network weight obtaining optimum and threshold value.
Step 8, SOM network learnt and trains, upgrading weights and threshold.
Whether step 9, discrimination precision meet the demands or reach maximum iteration time and set up; If set up, then perform step 10; If be false, then perform step 8.
Step 10, using the SOM network after adopting self-adapted genetic algorithm to optimize as sorter, fault diagnosis is carried out to the gear of normal, tooth surface abrasion, flank of tooth cut and broken teeth four kinds of states, and exports diagnostic result.
In a word, gear failure diagnosing method based on self-adapted genetic algorithm and SOM network of the present invention by the vibration signal of described gear that obtains after wavelet packet analysis and reconstruct, obtain the proper vector of the vibration signal of described gear, proper vector is divided into training data and test data, SOM network after first utilizing training data to train employing self-adapted genetic algorithm to optimize, constantly update the weights and threshold of SOM network, until the error that SOM network exports meets accuracy requirement or reaches maximum iteration time, then the SOM network trained is adopted to remove the fault type of diagnostic test data, the fault diagnosis result of output gear, thus improve precision and the reliability of Gear Fault Diagnosis.
In the inventive method, described step 2 comprises the steps:
Step 21, employing db5 small echo carry out three layers of wavelet decomposition to the vibration signal of described gear, extract the Gearbox vibration signal feature of the 3rd layer of 8 frequency content from low to high respectively, obtain coefficient of wavelet decomposition sequence from low to high, with (i in decomposition, j) a jth node of i-th layer is represented, each node represents certain Gearbox vibration signal feature, node (0,0) original signal S is represented, node (1,0) represents the 1st layer of low frequency coefficient X of WAVELET PACKET DECOMPOSITION
10, node (1,1) represents the 1st layer of high frequency coefficient X of WAVELET PACKET DECOMPOSITION
11, other by that analogy.
Step 22, to WAVELET PACKET DECOMPOSITION coefficient reconstruct, extract the Gearbox vibration signal feature of each frequency band range, WAVELET PACKET DECOMPOSITION calculating formula is as follows:
Wherein, d
lfor frequency band function, d
kfor wavelet packet basis, a
k, b
kbe respectively the coefficient of wavelet decomposition conjugate filter, the calculating formula of wavelet package reconstruction is as follows:
Wherein, p
k, q
kbe respectively wavelet reconstruction conjugate filter coefficient, use S
30represent X
30reconstruction signal, S
31represent X
31reconstruction signal, other by that analogy, analyze all nodes of the 3rd layer, the calculating formula of resultant signal S is as follows:
S=S
30+S
31+...+S
37
Step 23, calculate the gross energy of each band signal, S
3jcorresponding energy is E
3j(j=0 ..., 7), the calculating formula of gross energy is as follows:
Wherein, x
jk(j=0 ..., 7; K=1 ..., n) be reconstruction signal S
3jthe amplitude of discrete point.
Whole energy meter formulas of the vibration signal of step 24, described gear are as follows:
The relative wavelet-packet energy calculating formula of certain frequency range is as follows:
Wavelet Packet Energy Eigenvector is T=(T
30, T
31..., T
37), then selected characteristic vector T is the proper vector of the vibration signal of described gear.
In step 3 of the present invention, described coded system is specially:
The long calculating formula of chromosome string of real coding mode is as follows:
L=L
1×L
o+L
o
Wherein, L is chromosome length, L
ifor the neuron number of SOM network input layer, L
0for the neuron number of SOM network output layer.
In the inventive method, described step 6 comprises the steps:
Step 61, from current population, adopt the chromosome that the method choice ideal adaptation angle value of roulette is high, chromosomal fitness value is higher, larger by the chance selected, and the select probability calculating formula of roulette method is as follows:
Wherein, p
ibe i-th individual select probability, n is the size of population.
Step 62, chromosome high for ideal adaptation angle value is carried out adaptive crossover and mutation operation, crossover probability P
csize with fitness value carries out adaptive adjustment, adopts larger crossover probability P to the individuality of poor-performing
c, to the individuality of function admirable according to the size of ideal adaptation angle value, adopt suitable P
c, crossover probability P
ccalculating formula as follows:
Mutation probability P
malso with fitness value adaptively modifying, mutation probability P
mcalculating formula as follows:
Wherein, f
mfor the maximum adaptation angle value in colony, f
cfor treating fitness value the greater in intersection individuality, f
afor the average fitness value in colony, f is the ideal adaptation angle value that will make a variation, k
1, k
2, P
c1and P
m1be the random number on (0,1) interval.
In the inventive method, described step 8 comprises the steps:
Step 81, calculate the proper vector of vibration signal of described gear in the distance of moment t to all output nodes, adopt Eucliden distance, calculating formula is as follows:
Wherein, T
i(t) for proper vector is in the value of t, w
ijbe the connection weights between i-th input neuron node and a jth output neuron node.
Step 82, selection produce minor increment d
jnode as the neuron mated most,
neuron i (x) is triumph neuron.
Step 83, to triumph neuron, upgrade the weights and threshold of SOM network respectively, the calculating formula of weights is as follows:
w
ij(t+1)=w
ij(t)+η(t)h
j,i(x)[T
i(t)-w
ij(t)]
Wherein, η (t) is learning efficiency, 0 < η (t) < 1, and t dullness reduces in time, h
j, t (t)t () is the neighborhood function around triumph neuron, calculating formula is as follows:
Wherein, r
j, r
i (x)sOM network output node j respectively, the position of i (x);
The calculating formula of SOM network threshold is as follows:
b=e
1-log[(I-β)e-log(b)+β×α]
Wherein, β is the learning rate of threshold value, and 0 < β < 1, α is the neuronic output of output layer,
In the inventive method, described step 10 is specially:
The proper vector of the vibration signal of described gear is divided into training data and test data, SOM network after first utilizing training data to train employing self-adapted genetic algorithm to optimize, constantly update the weights and threshold of SOM network, until the error that SOM network exports meets accuracy requirement or reaches maximum iteration time, then, the SOM network trained is adopted to remove the fault type of diagnostic test data, the fault diagnosis result of output gear.
Embodiment
Gear case speed change system is normal, the vibration signal data of tooth surface abrasion, flank of tooth cut and broken teeth four kinds of state gears to adopt acceleration transducer to gather respectively, collect 300 groups of gear-type data are divided into 200 groups of training datas, 100 groups of test datas, utilize the SOM network trained to go diagnostic test data, Fig. 4 is SOM network fault diagnosis process flow diagram.The part training data of SOM network is as shown in table 1.The diagnostic result of SOM network is as shown in table 2.
Table 1SOM network portion training data
The diagnostic result of table 2SOM network
From table 2 data, when train epochs is 10, training data 1,2 is divided into a class, 3,4,5,6,7,8 be divided into another kind of, SOM network has carried out preliminary classification to data, when train epochs is 100,1 and 2,3 and 4,5 and 6,7 and 8 are divided into same class, and at this moment SOM network is to data Further Division, correctly can classify to the fault type of gear.From the test result of SOM network, adopt the SOM network after self-adapted genetic algorithm optimization can judge the fault type of gear exactly, reliability is high.
In sum, these are only preferred embodiment of the present invention, be not intended to limit protection scope of the present invention.Within the spirit and principles in the present invention all, any amendment done, equivalent replacement, improvement etc., all should be included within protection scope of the present invention.
Claims (6)
1., based on a gear failure diagnosing method for self-adapted genetic algorithm and SOM network, it is characterized in that, described gear failure diagnosing method comprises the steps:
Step 1, adopt acceleration transducer to gather gear case speed change system data, obtain the vibration signal of normal, tooth surface abrasion, flank of tooth cut and broken teeth four kinds of state gears;
Step 2, respectively employing db5 small echo carry out three layers of WAVELET PACKET DECOMPOSITION and reconstruct to the vibration signal of described gear, and with third layer 8 frequency bands separately energy for element structural attitude vector, extract the fault signature of gear, obtain the proper vector T of the vibration signal of described gear;
Step 3, using the neuron of the proper vector T of the vibration signal of described gear as SOM network input layer, the input layer of SOM network is one dimension, the output layer of SOM network is a two-dimensional network having 6 × 6 output neurons, the neuron of output layer lines up a neighbour structure, each neuron connects with other neuron side direction around it, and each input neuron is connected to all output neurons; The initial connection weights of SOM network are IW=[w
1, w
2..., w
8]
36 × 8, the initial threshold of SOM network is
wherein, w
1, w
2... w
8for less non-zero random number;
The coded system of step 4, self-adapted genetic algorithm adopts real coding mode, and each individuality is the chromosome of a real number string composition, is the initial population of some chromosome composition by the initial connection weights of SOM network and threshold coding;
Step 5, adopt the inverse of error sum of squares as the fitness function of each individuality, concrete calculating formula is as follows:
Wherein, f
ibe i-th individual fitness value, y
ireal output value, F is error sum of squares,
for idea output, according to selected fitness function, calculate the fitness value of each individuality;
Step 6, fitness value according to each individuality, select the high chromosome of ideal adaptation angle value to carry out copying, crossover and mutation operation, produces new population;
Step 7, the SOM network weight obtaining optimum and threshold value;
Step 8, SOM network learnt and trains, upgrading weights and threshold;
Whether step 9, discrimination precision meet the demands or reach maximum iteration time and set up; If set up, then perform step 10; If be false, then perform step 8;
Step 10, using the SOM network after adopting self-adapted genetic algorithm to optimize as sorter, fault diagnosis is carried out to the gear of normal, tooth surface abrasion, flank of tooth cut and broken teeth four kinds of states, and exports diagnostic result.
2. gear failure diagnosing method according to claim 1, is characterized in that, described step 2 comprises the steps:
Step 21, employing db5 small echo carry out three layers of wavelet decomposition to the vibration signal of described gear, extract the Gearbox vibration signal feature of the 3rd layer of 8 frequency content from low to high respectively, obtain coefficient of wavelet decomposition sequence from low to high, with (i in decomposition, j) a jth node of i-th layer is represented, each node represents certain Gearbox vibration signal feature, node (0,0) original signal S is represented, node (1,0) represents the 1st layer of low frequency coefficient X of WAVELET PACKET DECOMPOSITION
10, node (1,1) represents the 1st layer of high frequency coefficient X of WAVELET PACKET DECOMPOSITION
11, other by that analogy;
Step 22, to WAVELET PACKET DECOMPOSITION coefficient reconstruct, extract the Gearbox vibration signal feature of each frequency band range, WAVELET PACKET DECOMPOSITION calculating formula is as follows:
Wherein, d
lfor frequency band function, d
kfor wavelet packet basis, a
k, b
kbe respectively the coefficient of wavelet decomposition conjugate filter, the calculating formula of wavelet package reconstruction is as follows:
Wherein, p
k, q
kbe respectively wavelet reconstruction conjugate filter coefficient, use S
30represent X
30reconstruction signal, S
31represent X
31reconstruction signal, other by that analogy, analyze all nodes of the 3rd layer, the calculating formula of resultant signal S is as follows:
S=S
30+S
31+...+S
37
Step 23, calculate the gross energy of each band signal, S
3jcorresponding energy is E
3j(j=0 ..., 7), the calculating formula of gross energy is as follows:
Wherein, x
jk(j=0 ..., 7; K=1 ..., n) be reconstruction signal S
3jthe amplitude of discrete point;
Whole energy meter formulas of the vibration signal of step 24, described gear are as follows:
The relative wavelet-packet energy calculating formula of certain frequency range is as follows:
Wavelet Packet Energy Eigenvector is T=(T
30, T
31..., T
37), then selected characteristic vector T is the proper vector of the vibration signal of described gear.
3. gear failure diagnosing method according to claim 1, is characterized in that, in step 4, the long calculating formula of chromosome string of described real coding mode is as follows:
L=L
i×L
o+L
o
Wherein, L is chromosome length, L
ifor the neuron number of SOM network input layer, L
0for the neuron number of SOM network output layer.
4. gear failure diagnosing method according to claim 1, is characterized in that, described step 6 comprises following concrete steps:
Step 61, from current population, adopt the chromosome that the method choice ideal adaptation angle value of roulette is high, chromosomal fitness value is higher, larger by the chance selected, and the select probability calculating formula of roulette method is as follows:
Wherein, p
ibe i-th individual select probability, n is the size of population;
Step 62, chromosome high for ideal adaptation angle value is carried out adaptive crossover and mutation operation, crossover probability P
csize with fitness value carries out adaptive adjustment, adopts larger crossover probability P to the individuality of poor-performing
c, to the individuality of function admirable according to the size of ideal adaptation angle value, adopt suitable P
c, crossover probability P
ccalculating formula as follows:
Mutation probability P
malso with fitness value adaptively modifying, mutation probability P
mcalculating formula as follows:
Wherein, f
mfor the maximum adaptation angle value in colony, f
cfor treating fitness value the greater in intersection individuality, f
afor the average fitness value in colony, f is the ideal adaptation angle value that will make a variation, k
1, k
2, P
c1and P
m1be the random number on (0,1) interval.
5. gear failure diagnosing method according to claim 1, is characterized in that, described step 8 comprises the steps:
Step 81, calculate the proper vector of vibration signal of described gear in the distance of moment t to all output nodes, adopt Eucliden distance, calculating formula is as follows:
Wherein, T
i(t) for proper vector is in the value of t, w
ijbe the connection weights between i-th input neuron node and a jth output neuron node;
Step 82, selection produce minor increment d
jnode as the neuron mated most,
neuron i (x) is triumph neuron;
Step 83, to triumph neuron, upgrade the weights and threshold of SOM network respectively, the calculating formula of weights is as follows:
w
ij(t+1)=w
ij(t)+η(t)h
j,i(x)[T
i(t)-w
ij(t)]
Wherein, η (t) is learning efficiency, 0 < η (t) < 1, and t dullness reduces in time, h
j, i (t)t () is the neighborhood function around triumph neuron, calculating formula is as follows:
Wherein, r
j, r
i (x)sOM network output node j respectively, the position of i (x);
The calculating formula of SOM network threshold is as follows:
b=e
1-log[(1-β)e-log(b)+β×α]
Wherein, β is the learning rate of threshold value, and 0 < β < 1, α is the neuronic output of output layer, α=[α
1, α
2..., α
36],
6. gear failure diagnosing method according to claim 1, it is characterized in that, described step 10 is specially: the proper vector of the vibration signal of described gear is divided into training data and test data, SOM network after first utilizing training data to train employing self-adapted genetic algorithm to optimize, constantly update the weights and threshold of SOM network, until the error that SOM network exports meets accuracy requirement or reaches maximum iteration time, then, the SOM network trained is adopted to remove the fault type of diagnostic test data, the fault diagnosis result of output gear.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510610800.XA CN105335698A (en) | 2015-09-15 | 2015-09-15 | Gear failure diagnosis method based on adaptive genetic algorithm and SOM (Self-Organizing Map) network |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510610800.XA CN105335698A (en) | 2015-09-15 | 2015-09-15 | Gear failure diagnosis method based on adaptive genetic algorithm and SOM (Self-Organizing Map) network |
Publications (1)
Publication Number | Publication Date |
---|---|
CN105335698A true CN105335698A (en) | 2016-02-17 |
Family
ID=55286214
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201510610800.XA Pending CN105335698A (en) | 2015-09-15 | 2015-09-15 | Gear failure diagnosis method based on adaptive genetic algorithm and SOM (Self-Organizing Map) network |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN105335698A (en) |
Cited By (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105651512A (en) * | 2016-04-21 | 2016-06-08 | 重庆理工大学 | Bump detection method of main speed-reducing gear of automobile driving axle |
CN106529725A (en) * | 2016-11-10 | 2017-03-22 | 河南理工大学 | Gas outburst prediction method based on firefly algorithm and SOM network |
CN107576491A (en) * | 2017-10-18 | 2018-01-12 | 河海大学 | A kind of breaker mechanical fault recognition method |
CN107607303A (en) * | 2017-09-13 | 2018-01-19 | 河海大学 | Mechanical Failure of HV Circuit Breaker recognition methods based on wavelet packet Yu SOM networks |
CN108268892A (en) * | 2017-12-29 | 2018-07-10 | 英特尔产品(成都)有限公司 | Fault in production management analysis method |
CN109508248A (en) * | 2018-11-14 | 2019-03-22 | 上海交通大学 | The detection method of fuel system failure based on self-organizing map neural network |
CN109993183A (en) * | 2017-12-30 | 2019-07-09 | 中国移动通信集团四川有限公司 | Network failure appraisal procedure, calculates equipment and storage medium at device |
CN110688809A (en) * | 2019-09-05 | 2020-01-14 | 西安理工大学 | Box transformer substation fault diagnosis method based on VPRS-RBF neural network |
CN110703078A (en) * | 2019-09-26 | 2020-01-17 | 河海大学 | GIS fault diagnosis method based on spectral energy analysis and self-organizing competition algorithm |
CN110726957A (en) * | 2019-11-05 | 2020-01-24 | 国网江苏省电力有限公司宜兴市供电分公司 | Fault identification method of dry-type reactor |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102024179A (en) * | 2010-12-07 | 2011-04-20 | 南京邮电大学 | Genetic algorithm-self-organization map (GA-SOM) clustering method based on semi-supervised learning |
CN104390776A (en) * | 2014-12-10 | 2015-03-04 | 北京航空航天大学 | Fault detection, diagnosis and performance evaluation method for redundant aileron actuator |
CN104850889A (en) * | 2014-11-19 | 2015-08-19 | 北京航空航天大学 | Airplane rotation actuator drive unit adaptive fault detection, isolation and confidences assessment method |
-
2015
- 2015-09-15 CN CN201510610800.XA patent/CN105335698A/en active Pending
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102024179A (en) * | 2010-12-07 | 2011-04-20 | 南京邮电大学 | Genetic algorithm-self-organization map (GA-SOM) clustering method based on semi-supervised learning |
CN104850889A (en) * | 2014-11-19 | 2015-08-19 | 北京航空航天大学 | Airplane rotation actuator drive unit adaptive fault detection, isolation and confidences assessment method |
CN104390776A (en) * | 2014-12-10 | 2015-03-04 | 北京航空航天大学 | Fault detection, diagnosis and performance evaluation method for redundant aileron actuator |
Non-Patent Citations (3)
Title |
---|
T KOHONEN: ""The self-organizing map"", 《PROCEEDINGS OF THE IEEE》 * |
孙士慧: ""基于小波神经网络的设备故障诊断方法研究"", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
闻新等: "《应用MATLAB实现神经网络》", 30 June 2015 * |
Cited By (13)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105651512B (en) * | 2016-04-21 | 2018-05-18 | 重庆理工大学 | A kind of automobile drive axle master subtracts gear enclosed mass detection method |
CN105651512A (en) * | 2016-04-21 | 2016-06-08 | 重庆理工大学 | Bump detection method of main speed-reducing gear of automobile driving axle |
CN106529725A (en) * | 2016-11-10 | 2017-03-22 | 河南理工大学 | Gas outburst prediction method based on firefly algorithm and SOM network |
CN107607303A (en) * | 2017-09-13 | 2018-01-19 | 河海大学 | Mechanical Failure of HV Circuit Breaker recognition methods based on wavelet packet Yu SOM networks |
CN107576491A (en) * | 2017-10-18 | 2018-01-12 | 河海大学 | A kind of breaker mechanical fault recognition method |
CN108268892A (en) * | 2017-12-29 | 2018-07-10 | 英特尔产品(成都)有限公司 | Fault in production management analysis method |
CN109993183A (en) * | 2017-12-30 | 2019-07-09 | 中国移动通信集团四川有限公司 | Network failure appraisal procedure, calculates equipment and storage medium at device |
CN109508248A (en) * | 2018-11-14 | 2019-03-22 | 上海交通大学 | The detection method of fuel system failure based on self-organizing map neural network |
CN109508248B (en) * | 2018-11-14 | 2021-08-20 | 上海交通大学 | Fuel system fault detection method based on self-organizing mapping neural network |
CN110688809A (en) * | 2019-09-05 | 2020-01-14 | 西安理工大学 | Box transformer substation fault diagnosis method based on VPRS-RBF neural network |
CN110688809B (en) * | 2019-09-05 | 2023-06-09 | 西安理工大学 | Box transformer fault diagnosis method based on VPRS-RBF neural network |
CN110703078A (en) * | 2019-09-26 | 2020-01-17 | 河海大学 | GIS fault diagnosis method based on spectral energy analysis and self-organizing competition algorithm |
CN110726957A (en) * | 2019-11-05 | 2020-01-24 | 国网江苏省电力有限公司宜兴市供电分公司 | Fault identification method of dry-type reactor |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN105335698A (en) | Gear failure diagnosis method based on adaptive genetic algorithm and SOM (Self-Organizing Map) network | |
Han et al. | Multi-level wavelet packet fusion in dynamic ensemble convolutional neural network for fault diagnosis | |
CN103728535B (en) | A kind of extra-high-voltage direct-current transmission line fault location based on wavelet transformation transient state energy spectrum | |
Cuadra et al. | Computational intelligence in wave energy: Comprehensive review and case study | |
CN105550700B (en) | A kind of time series data cleaning method based on association analysis and principal component analysis | |
CN103728551B (en) | A kind of analog-circuit fault diagnosis method based on cascade integrated classifier | |
CN108062572A (en) | A kind of Fault Diagnosis Method of Hydro-generating Unit and system based on DdAE deep learning models | |
CN106124212A (en) | Based on sparse coding device and the Fault Diagnosis of Roller Bearings of support vector machine | |
CN110738010A (en) | Wind power plant short-term wind speed prediction method integrated with deep learning model | |
CN109213121B (en) | Method for diagnosing clamping cylinder fault of fan braking system | |
CN102487343B (en) | Diagnosis and prediction method for hidden faults of satellite communication system | |
CN107941537A (en) | A kind of mechanical equipment health state evaluation method | |
CN108089099A (en) | The diagnostic method of distribution network failure based on depth confidence network | |
CN104751229A (en) | Bearing fault diagnosis method capable of recovering missing data of back propagation neural network estimation values | |
CN108960303A (en) | A kind of unmanned plane during flying data exception detection method based on LSTM | |
CN103926526A (en) | Analog circuit fault diagnosis method based on improved RBF neural network | |
CN103413174A (en) | Short-term wind speed multi-step prediction method based on deep learning method | |
CN103995237A (en) | Satellite power supply system online fault diagnosis method | |
CN101587155A (en) | Oil soaked transformer fault diagnosis method | |
CN105628425A (en) | Rotation machinery early stage fault diagnosis method based on heredity annealing optimization multi-core support vector machine | |
CN101819253A (en) | Probabilistic neural network-based tolerance-circuit fault diagnosis method | |
CN109214356A (en) | A kind of fan transmission system intelligent fault diagnosis method based on DCNN model | |
CN106250935A (en) | The rotary machinery fault diagnosis method that genetic programming and weighted evidence theory merge | |
Xie et al. | Learning features from high speed train vibration signals with deep belief networks | |
CN105005708B (en) | A kind of broad sense load Specialty aggregation method based on AP clustering algorithms |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20160217 |
|
WD01 | Invention patent application deemed withdrawn after publication |