CN104634603A - Early fault diagnosis method for complex equipment - Google Patents
Early fault diagnosis method for complex equipment Download PDFInfo
- Publication number
- CN104634603A CN104634603A CN201510114270.XA CN201510114270A CN104634603A CN 104634603 A CN104634603 A CN 104634603A CN 201510114270 A CN201510114270 A CN 201510114270A CN 104634603 A CN104634603 A CN 104634603A
- Authority
- CN
- China
- Prior art keywords
- fault diagnosis
- neural network
- msub
- kernel function
- mrow
- 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
- 238000003745 diagnosis Methods 0.000 title claims abstract description 78
- 238000000034 method Methods 0.000 title claims abstract description 36
- 238000013528 artificial neural network Methods 0.000 claims abstract description 35
- 230000006870 function Effects 0.000 claims description 25
- 238000012423 maintenance Methods 0.000 claims description 16
- 238000012549 training Methods 0.000 claims description 14
- 230000009466 transformation Effects 0.000 claims description 7
- 238000012545 processing Methods 0.000 claims description 5
- 238000000605 extraction Methods 0.000 claims description 3
- 238000007781 pre-processing Methods 0.000 claims description 3
- 238000012706 support-vector machine Methods 0.000 abstract description 5
- 230000008901 benefit Effects 0.000 abstract description 3
- 238000012360 testing method Methods 0.000 abstract description 3
- 238000010586 diagram Methods 0.000 description 9
- 238000004458 analytical method Methods 0.000 description 6
- 238000001514 detection method Methods 0.000 description 4
- 238000010606 normalization Methods 0.000 description 4
- 238000013507 mapping Methods 0.000 description 3
- 238000013178 mathematical model Methods 0.000 description 3
- 238000013473 artificial intelligence Methods 0.000 description 2
- 238000004422 calculation algorithm Methods 0.000 description 2
- 238000010276 construction Methods 0.000 description 2
- 238000000354 decomposition reaction Methods 0.000 description 2
- 238000012544 monitoring process Methods 0.000 description 2
- 238000003062 neural network model Methods 0.000 description 2
- 210000002569 neuron Anatomy 0.000 description 2
- OAICVXFJPJFONN-UHFFFAOYSA-N Phosphorus Chemical compound [P] OAICVXFJPJFONN-UHFFFAOYSA-N 0.000 description 1
- 230000003044 adaptive effect Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 230000036541 health Effects 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000003909 pattern recognition Methods 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 230000035945 sensitivity Effects 0.000 description 1
- 230000002123 temporal effect Effects 0.000 description 1
Landscapes
- Testing And Monitoring For Control Systems (AREA)
Abstract
The invention discloses an early fault diagnosis method for complex equipment. The early fault diagnosis method for the complex equipment comprises the specific steps: extracting collection signals through a sensor; carrying out wavelet transform on the extracted collection signals for de-noising; and establishing a fault diagnosis model based on a BP (Back Propagation) neural network and a fault diagnosis model based on an SVM (Support Vector Machine) according to mechanical features and circuit features, and carrying out fault diagnosis. A kernel function in the fault diagnosis model based on the SVM is a polynomial kernel function, a radial basis function (RBF) or a Sigmoid kernel function. According to the early fault diagnosis method, the fault diagnosis of the complex equipment is divided into a fault diagnosis with the mechanical features and a fault diagnosis with the circuit features, and the fault diagnosis models are established respectively according to the mechanical features and the circuit features, so that mechanical tests are easy, many samples can be obtained, quick convergence can be realized by using the BP neural network, and accuracy is higher; furthermore, the circuit sample data is less. With the adoption of the advantages of small samples of the SVM, the fault diagnosis of the complex system, such as the complex equipment, can be realized.
Description
Technical Field
The invention relates to a fault diagnosis method, in particular to an early fault diagnosis method for complex equipment.
Background
The fault diagnosis mainly refers to monitoring the state of the equipment and judging faults, not only needs to make correct judgment on the reason, position and degree of the equipment fault and further prevent the equipment fault and reduce the fault loss, but also needs to monitor the health state of the equipment, especially monitors early faults, makes early warning judgment as soon as possible to reduce the occurrence of sudden faults, and can also predict the occurrence time of future faults, so that the maintenance resources can be fully saved, the best maintenance decision basis is provided for equipment maintenance and maintenance, and especially the possibility is provided for realizing the maintenance strategy according to the circumstances.
Technically, fault diagnosis is actually implemented by automatically detecting monitoring and diagnosing equipment under the guidance of a certain fault judgment strategy, namely, a fault model of the equipment is obtained through analysis and construction, fault characteristics are extracted, monitored information is comprehensively evaluated according to a preset strategy and principle, and finally necessary maintenance measures are prompted to maintenance personnel. Therefore, fault diagnosis requires not only modern control theory, computer science, artificial intelligence, signal processing, pattern recognition, statistical mathematics, and the like. The common methods are mainly divided into two categories, namely a mathematical model-based method and an artificial intelligence-based method, and the mathematical model-based method is the most widely applied fault detection method at present. The method is based on monitored signals, extracts characteristic information of the signals through signal processing and analysis, and then compares the characteristics under normal conditions to judge whether the analysis equipment has faults, and a specific diagnosis schematic diagram is shown in fig. 1.
The detection analysis for a single signal is simple, for example: a simple system output signal is y (t), as shown in fig. 2, under an actual normal operating condition, the amplitude of the output signal y (t) is within an interval [ min (y (t)) and max (y (t)) ], as shown in fig. 2, a red dotted line is an amplitude interval range, which is under the normal operating condition of the system. Theoretically, under a normal working condition, the output signal y (t) is as a red solid line in fig. 2, the actually measured output signal is compared with the amplitude interval range, and if the actually measured output signal exceeds the amplitude interval range, the fault state is determined. The measured output signal is shown as a blue solid line in fig. 2.
The interval range of the detection of a single signal is easy to establish, if a complex system monitors multiple signals, a specific envelope model is difficult to establish, and the method is difficult to carry out fault diagnosis.
Disclosure of Invention
The invention aims to provide a complex equipment early fault diagnosis method with convenient use and high accuracy, so as to solve the problems in the background technology.
In order to achieve the purpose, the invention provides the following technical scheme:
a method for diagnosing early faults of complex equipment comprises the following specific steps:
(1) extracting an acquisition signal through a sensor;
(2) carrying out wavelet transformation denoising on the extracted and collected signals;
(3) and respectively establishing a fault diagnosis model based on the BP neural network and a fault diagnosis model based on the SVM according to the mechanical characteristics and the circuit characteristics, and performing fault diagnosis.
As a further scheme of the invention: the method for diagnosing the fault based on the fault diagnosis model of the BP neural network comprises the following specific steps:
(1) the sensor is connected to mechanical equipment to extract an acquisition signal;
(2) carrying out wavelet transformation denoising processing on the extracted and collected signals, then respectively carrying out original data diagnosis and sample data training, and then respectively carrying out preprocessing and feature selection/extraction in a BP neural network;
(3) the original data is preprocessed and feature-selected/extracted and then is diagnosed by a BP neural network to obtain a diagnosis result, and the sample data is preprocessed and feature-selected/extracted and then is subjected to learning training;
(4) and carrying out diagnosis decision on the obtained diagnosis result, and then respectively carrying out regular maintenance, situation maintenance and after-the-fact maintenance.
As a still further scheme of the invention: and the kernel function in the SVM-based fault diagnosis model is a polynomial kernel function, a radial basis kernel function (RBF) or a Sigmoid kernel function.
Compared with the prior art, the invention has the beneficial effects that:
according to the invention, the fault diagnosis of the complex equipment is divided into the fault diagnosis of the mechanical characteristic and the fault diagnosis of the circuit characteristic, and the fault diagnosis models are respectively established according to the mechanical characteristic and the circuit characteristic, so that the mechanical test is easy, more samples can be obtained, the convergence is fast by applying a BP neural network, and the accuracy is higher; and the circuit sample data is less, and the advantage of the small sample of the SVM is utilized, so that the fault diagnosis of complex systems such as complex equipment is realized.
Drawings
FIG. 1 is a diagram of a mathematical model-based method specific diagnostic principle.
FIG. 2 is a schematic diagram of a simple system for single signal detection and analysis.
Fig. 3 is a schematic diagram of a fault diagnosis method according to the present invention.
Fig. 4 is a schematic diagram of a fault diagnosis method based on a fault diagnosis model of a BP neural network in the present invention.
FIG. 5 is a topology structure diagram of the BP neural network of the present invention.
FIG. 6 is a schematic diagram of learning and training of the BP neural network according to the present invention.
Fig. 7 is a signal waveform diagram after denoising and extracting the vibration signal of the mechanical device in the invention.
FIG. 8 is a non-linear mapping from an SVM-based fault diagnosis model input space to a high-dimensional feature space in accordance with the present invention.
Fig. 9 is a schematic diagram of a fault diagnosis method based on the fault diagnosis model of the SVM in the present invention.
Detailed Description
The technical solution of the present patent will be described in further detail with reference to the following embodiments.
Referring to fig. 3, a method for diagnosing early failure of complex equipment includes the following steps:
(1) extracting an acquisition signal through a sensor;
(2) carrying out wavelet transformation denoising on the extracted and collected signals;
(3) and respectively establishing a fault diagnosis model based on the BP neural network and a fault diagnosis model based on the SVM according to the mechanical characteristics and the circuit characteristics, and performing fault diagnosis.
Referring to fig. 4, the method for performing fault diagnosis based on the fault diagnosis model of the BP neural network includes the following specific steps:
(1) the sensor is connected to mechanical equipment to extract an acquisition signal;
(2) carrying out wavelet transformation denoising processing on the extracted and collected signals, then respectively carrying out original data diagnosis and sample data training, and then respectively carrying out preprocessing and feature selection/extraction in a BP neural network;
(3) the original data is preprocessed and feature-selected/extracted and then is diagnosed by a BP neural network to obtain a diagnosis result, and the sample data is preprocessed and feature-selected/extracted and then is subjected to learning training;
(4) and carrying out diagnosis decision on the obtained diagnosis result, and then respectively carrying out regular maintenance, situation maintenance and after-the-fact maintenance.
Referring to fig. 5, the BP neural network includes an input layer, an intermediate layer, and an output layer, the learning algorithm in the learning training is a supervised learning algorithm, and the BP neural network is adjusted according to the correct input and output, so that the BP neural network can make a correct response. The learning training samples are represented as: (p)i,di) 1,2, n, wherein p isiInput data for the sample, diOutputting data for the sample, adjusting parameters of each neuron through learning to enable the BP neural network to generate expected results, and inputting the sample p after learning and training of the BP neural network modeliThe output of the BP neural network is as much as possible compared with diAnd (4) approaching.
Referring to fig. 6, the learning of the BP neural network is divided into two stages: inputting a determined learning sample, setting the structure of a BP neural network, calculating a weight and a threshold obtained by previous iteration, calculating backwards from a first layer, and solving the output of each neuron; modifying the weight and the threshold, and respectively calculating and modifying each weight and threshold from the last layer to the front according to the total error influence gradient. And the two stages are repeatedly alternated, and the weight and the threshold between layers are modified according to the error until the error meets the requirement after convergence.
Taking an engine in a diesel generator as an example to explain the construction of a fault diagnosis model of a mechanical equipment rotation system based on a BP neural network, please refer to FIG. 6, three-layer wavelet decomposition is performed on an output signal y (t) extracted by denoising a vibration signal of the engine in the diesel generator, and a decomposition coefficient vector of a third layer is [ y ] from low to high respectively30,y31,y32,y33,y34,y35,y36,y37]The total signal y:
y＝S30+S31+S32+S33+S34+S35+S36+S37equation 1
Wherein S is30Representing node y30Reconstructing the recovered signal, S31Representing node y31Reconstructing the recovered signal, S32Representing node y32Reconstructing the recovered signal, S33Representing node y33Reconstructing the recovered signal, S34Representing node y34Reconstructing the recovered signal, S33Representing node y35Reconstructing the recovered signal, S36Representing node y36Reconstructing the recovered signal, S37Representing node y37The reconstruction of (2) restores the signal.
Assuming that the lowest frequency in the output signal y (t) is 0 and the highest frequency is 1, the output signal is divided into eight frequency ranges, as shown in table 1:
TABLE 1 frequency Range of wavelet decomposed signals
| Signal component | Frequency interval | Signal component | Frequency interval |
| S30 | 0～0.125 | S34 | 0.500～0.625 |
| S31 | 0.125～0.250 | S35 | 0.625～0.750 |
| S32 | 0.250～0.375 | S36 | 0.750～0.875 |
| S33 | 0.375～0.500 | S37 | 0.875～1 |
Determination of the input signal: the input signal includes band signal energy and temporal abruptness.
Wavelet decomposing the output signal y (t) into eight frequency bands, and calculating the signal energy of each frequency band:
Wherein E is3jIs a frequency band S3jCorresponding to the capability, Ts is the analysis period of the output signal y (t), j is 0, 1, 7, k is 1,2, N is the number of samples, y is the number of samplesjkIs S3jThe k-th discrete point amplitude of the signal.
In order to improve the accuracy of the BP neural network, the energy of each frequency band is normalized:
Thus, eight eigenvectors of the signal energy of each band are obtained:
The probability distribution of the vibration signals of the mechanical equipment rotating system is close to the positive distribution, the time domain abruptness is the quantitative degree of describing the deviation of the actually measured signals from the positive distribution, and the time domain abruptness calculation method comprises the following steps:
Adopting a most value normalization treatment:
Wherein,for data after normalization, giFor data before normalization, gi min、gi maxThe minimum and maximum of the data before normalization, respectively.
Thus, the variables of the BP neural network input parameter P are 10:
Determination of output parameters:
for engine fault diagnosis, the vibration faults comprise dozens of types, and the occurrence rate of the vibration faults is about more than 95 of the total number of the vibration faults, and the typical faults comprise rotor unbalance, rotor rubbing, rotor misalignment, loose bearings and the like.
Therefore, the four typical faults are taken as the output parameters of the BP neural network fault diagnosis model and are expressed as
D＝[d1，d2，d3，d4]Di is more than 0 and less than or equal to 1, formula 9
The ideal output state and meaning are shown in table 2:
TABLE 3.2 output State and Definitions
| Type of failure | d1 | d2 | d3 | d4 |
| Rotor unbalance | 1 | 0 | 0 | 0 |
| Rotor rubbing | 0 | 1 | 0 | 0 |
| Misalignment of rotor | 0 | 0 | 1 | 0 |
| Bearing looseness | 0 | 0 | 0 | 1 |
Therefore, in the BP neural network fault diagnosis model, the number of input nodes in an input layer is ten, the input nodes comprise eight frequency band energy signals and two time domain steepnesses, the number of output nodes in an output layer is four, and twenty-one hidden layer nodes in the middle layer are twenty-one according to the principle of 2M + 1.
The BP neural network model has good nonlinear capability, and is simple in structure and good in performance for an actual nonlinear system. The key of the fault diagnosis is to implement mapping from the fault feature space to the fault space, so as to achieve the purpose of fault diagnosis. The BP neural network has adaptive capacity, the network can not only learn in a self-adaptive mode, but also adjust the size of the network, and in addition, the BP neural network has certain fault tolerance, which is shown in incompleteness to input mode information or low sensitivity to defect of characteristics, but is obvious to mechanical characteristics in complex equipment, a large number of fault samples are easily obtained through experiments, and the reliability of the BP neural network is ensured. Therefore, the BP neural network can effectively realize the fault diagnosis of the complex equipment in the complex system.
Referring to fig. 8-9, the SVM-based fault diagnosis model outputs a feature sample (x) through N-dimensional input of the denoised wavelet signali,yi) The mapping F from the original space to the high-dimensional feature space is achieved by selecting a non-linear transformation θ (-), where xi∈RnAs input vector, yi∈[-1,1]As an output failure class, i 1, 2.
Constructing an optimal linear classification function:
(x) sign [ (w · θ (x) + b) ] formula 10
Wherein w is a weight vector; b is the deviation or a classification threshold.
To satisfy the minimization of structural risk, we need to find w and b to make equation 11 hold:
Wherein, wTw is the complexity of a control model, and C is a predetermined constant, so that the sample control punishment degree of the output error is realized; rPIs an error control function.
The error control function is RPA two-norm of the error ξ, so the fault diagnosis model can be optimized as:
st.yi[wT·θ(xi)+b]＝1-ξ
Since the weight vector may be infinite in dimension, equation 12 is converted to the following system of equations:
Wherein 1 ═ 1, 1., 1]T，Y＝[y1,y2,...,yN]T，a＝[a1,a2,...,aN]TI is a unit matrix, K ═ θ (x)i)Tθ(xj) Is a kernel function.
The kernel function is a polynomial kernel function, a radial basis kernel function RBF or a Sigmoid kernel function. The specific functional form is as follows: polynomial kernel function K (x, x ') (x · x' +1)dWherein d is the order of the kernel function. Radial basis kernel function RBF: k (x, x ') -exp (- γ | | | x-x' | | non-conducting phosphor2) Wherein gamma is more than 0, and gamma is a coefficient for controlling the radius. Sigmoid kernel function: k (x, x') ═ tanh (vx)Txi+α)。
When the kernel function is a radial basis kernel function, the values of b and a are obtained by adopting a least square method, and finally, a fault classification function subjected to wavelet denoising can be obtained:
When the SVM-based fault diagnosis model is trained, training samples (x) are input1,y1)，...，(xN,yN) Wherein x isi∈X＝RnN is the number of training samples, N is the number of input variables, and m is the number of fault categories; determining the type of the kernel function; solving the optimal solution of the target function formula by using a quadratic programming method to obtain the optimal a; substituting a support vector machine X in the sample library into the formula 13 to obtain a deviation value b.
The classification method of the SVM comprises a one-to-one method and a one-to-many method, the training speed of the one-to-one method is much higher than that of the one-to-many method, the global optimization capability is strong, in the training effect, the one-to-many method needs to distinguish samples of a certain class from other sample classes, and the classification hyperplane is more complex than the one-to-one method and is more prone to the problem of excessive matching. Thus, the one-to-one method is more suitable for engineering applications.
According to the invention, the fault diagnosis of the complex equipment is divided into the fault diagnosis of the mechanical characteristic and the fault diagnosis of the circuit characteristic, and the fault diagnosis models are respectively established according to the mechanical characteristic and the circuit characteristic, so that the mechanical test is easy, more samples can be obtained, the convergence is fast by applying a BP neural network, and the accuracy is higher; and the circuit sample data is less, and the advantage of the small sample of the SVM is utilized, so that the fault diagnosis of complex systems such as complex equipment is realized.
Although the preferred embodiments of the present patent have been described in detail, the present patent is not limited to the above embodiments, and various changes can be made without departing from the spirit of the present patent within the knowledge of those skilled in the art.
Claims (3)
1. A method for diagnosing early faults of complex equipment is characterized by comprising the following specific steps:
(1) extracting an acquisition signal through a sensor;
(2) carrying out wavelet transformation denoising on the extracted and collected signals;
(3) and respectively establishing a fault diagnosis model based on the BP neural network and a fault diagnosis model based on the SVM according to the mechanical characteristics and the circuit characteristics, and performing fault diagnosis.
2. The early fault diagnosis method for the complex equipment according to claim 1, wherein the fault diagnosis method based on the fault diagnosis model of the BP neural network comprises the following specific steps:
(1) the sensor is connected to mechanical equipment to extract an acquisition signal;
(2) carrying out wavelet transformation denoising processing on the extracted and collected signals, then respectively carrying out original data diagnosis and sample data training, and then respectively carrying out preprocessing and feature selection/extraction in a BP neural network;
(3) the original data is preprocessed and feature-selected/extracted and then is diagnosed by a BP neural network to obtain a diagnosis result, and the sample data is preprocessed and feature-selected/extracted and then is subjected to learning training;
(4) and carrying out diagnosis decision on the obtained diagnosis result, and then respectively carrying out regular maintenance, situation maintenance and after-the-fact maintenance.
3. The early fault diagnosis method for the complex equipment according to claim 1, wherein the kernel function in the SVM based fault diagnosis model is a polynomial kernel function, a radial basis kernel function (RBF) or a Sigmoid kernel function.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510114270.XA CN104634603A (en) | 2015-03-16 | 2015-03-16 | Early fault diagnosis method for complex equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510114270.XA CN104634603A (en) | 2015-03-16 | 2015-03-16 | Early fault diagnosis method for complex equipment |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN104634603A true CN104634603A (en) | 2015-05-20 |
Family
ID=53213598
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510114270.XA Pending CN104634603A (en) | 2015-03-16 | 2015-03-16 | Early fault diagnosis method for complex equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104634603A (en) |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109406949A (en) * | 2018-12-14 | 2019-03-01 | 国网山东省电力公司电力科学研究院 | Power distribution network incipient fault detection method and device based on support vector machines |
| CN109580230A (en) * | 2018-12-11 | 2019-04-05 | 中国航空工业集团公司西安航空计算技术研究所 | A kind of Fault Diagnosis of Engine and device based on BP neural network |
| CN111259927A (en) * | 2020-01-08 | 2020-06-09 | 西北工业大学 | A Fault Diagnosis Method of Rocket Engine Based on Neural Network and Evidence Theory |
| CN111413030A (en) * | 2019-01-07 | 2020-07-14 | 哈尔滨工业大学 | Measurement and neural network learning control method and device for large-scale high-speed rotary equipment based on spatial projection maximization of stiffness vector |
| CN111539516A (en) * | 2020-04-22 | 2020-08-14 | 谭雄向 | Power grid fault diagnosis system and method based on big data processing |
| RU2753151C1 (en) * | 2020-09-23 | 2021-08-12 | Федеральное государственное бюджетное образовательное учреждение высшего образования "Орловский государственный университет имени И.С. Тургенева" (ФГБОУ ВО "ОГУ имени И.С. Тургенева") | Method for vibration diagnostics of rotary systems |
| CN116993003A (en) * | 2023-07-01 | 2023-11-03 | 武汉爱科森网络科技有限公司 | Equipment failure prediction system and method |
| CN117978612A (en) * | 2024-03-28 | 2024-05-03 | 成都格理特电子技术有限公司 | Network fault detection method, storage medium and electronic device |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5566092A (en) * | 1993-12-30 | 1996-10-15 | Caterpillar Inc. | Machine fault diagnostics system and method |
| CN101158873A (en) * | 2007-09-26 | 2008-04-09 | 东北大学 | A Fault Diagnosis Method for Nonlinear Process |
| CN101614787A (en) * | 2009-07-07 | 2009-12-30 | 南京航空航天大学 | Fault Diagnosis Method of Analog Electronic Circuit Based on M-ary Structure Classifier |
| KR20120027733A (en) * | 2010-09-13 | 2012-03-22 | 한국수력원자력 주식회사 | Rotating machinery fault diagnostic method and system using support vector machines |
| CN103512765A (en) * | 2013-09-13 | 2014-01-15 | 中国科学院苏州生物医学工程技术研究所 | Fault detection method for variable learning rate wavelet BP neural network of blood type centrifugal machine |
| CN103995237A (en) * | 2014-05-09 | 2014-08-20 | 南京航空航天大学 | Satellite power supply system online fault diagnosis method |
-
2015
- 2015-03-16 CN CN201510114270.XA patent/CN104634603A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5566092A (en) * | 1993-12-30 | 1996-10-15 | Caterpillar Inc. | Machine fault diagnostics system and method |
| CN101158873A (en) * | 2007-09-26 | 2008-04-09 | 东北大学 | A Fault Diagnosis Method for Nonlinear Process |
| CN101614787A (en) * | 2009-07-07 | 2009-12-30 | 南京航空航天大学 | Fault Diagnosis Method of Analog Electronic Circuit Based on M-ary Structure Classifier |
| KR20120027733A (en) * | 2010-09-13 | 2012-03-22 | 한국수력원자력 주식회사 | Rotating machinery fault diagnostic method and system using support vector machines |
| CN103512765A (en) * | 2013-09-13 | 2014-01-15 | 中国科学院苏州生物医学工程技术研究所 | Fault detection method for variable learning rate wavelet BP neural network of blood type centrifugal machine |
| CN103995237A (en) * | 2014-05-09 | 2014-08-20 | 南京航空航天大学 | Satellite power supply system online fault diagnosis method |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109580230A (en) * | 2018-12-11 | 2019-04-05 | 中国航空工业集团公司西安航空计算技术研究所 | A kind of Fault Diagnosis of Engine and device based on BP neural network |
| CN109406949A (en) * | 2018-12-14 | 2019-03-01 | 国网山东省电力公司电力科学研究院 | Power distribution network incipient fault detection method and device based on support vector machines |
| CN111413030A (en) * | 2019-01-07 | 2020-07-14 | 哈尔滨工业大学 | Measurement and neural network learning control method and device for large-scale high-speed rotary equipment based on spatial projection maximization of stiffness vector |
| CN111259927A (en) * | 2020-01-08 | 2020-06-09 | 西北工业大学 | A Fault Diagnosis Method of Rocket Engine Based on Neural Network and Evidence Theory |
| CN111259927B (en) * | 2020-01-08 | 2022-08-05 | 西北工业大学 | A Fault Diagnosis Method of Rocket Engine Based on Neural Network and Evidence Theory |
| CN111539516A (en) * | 2020-04-22 | 2020-08-14 | 谭雄向 | Power grid fault diagnosis system and method based on big data processing |
| RU2753151C1 (en) * | 2020-09-23 | 2021-08-12 | Федеральное государственное бюджетное образовательное учреждение высшего образования "Орловский государственный университет имени И.С. Тургенева" (ФГБОУ ВО "ОГУ имени И.С. Тургенева") | Method for vibration diagnostics of rotary systems |
| CN116993003A (en) * | 2023-07-01 | 2023-11-03 | 武汉爱科森网络科技有限公司 | Equipment failure prediction system and method |
| CN117978612A (en) * | 2024-03-28 | 2024-05-03 | 成都格理特电子技术有限公司 | Network fault detection method, storage medium and electronic device |
| CN117978612B (en) * | 2024-03-28 | 2024-06-04 | 成都格理特电子技术有限公司 | Network fault detection method, storage medium and electronic equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113935406A (en) | Mechanical equipment unsupervised fault diagnosis method based on counter current model | |
| CN110110809B (en) | Fuzzy automaton construction method based on machine fault diagnosis | |
| CN111964908A (en) | A bearing fault diagnosis method under variable working conditions based on MWDCNN | |
| CN104502103A (en) | Bearing fault diagnosis method based on fuzzy support vector machine | |
| CN114386452A (en) | Method for detecting faults of sun wheel of nuclear power circulating water pump | |
| CN114118138A (en) | A Bearing Composite Fault Diagnosis Method Based on Multi-label Domain Adaptive Model | |
| Ye et al. | Multiscale weighted morphological network based feature learning of vibration signals for machinery fault diagnosis | |
| Zhang et al. | A novel intelligent fault diagnosis method based on variational mode decomposition and ensemble deep belief network | |
| CN106596116A (en) | Vibration fault diagnosis method of wind generating set | |
| KR102404498B1 (en) | Industrial gearbox failure diagnosis apparatus and method using convolutional neural network based on adaptive time-frequency representation | |
| CN104809722A (en) | Electrical device fault diagnosis method based on infrared thermography | |
| CN103995237A (en) | Satellite power supply system online fault diagnosis method | |
| CN114964778A (en) | Bearing fault diagnosis method based on wavelet time-frequency graph and deep learning | |
| Han et al. | An intelligent fault diagnosis method of variable condition gearbox based on improved DBN combined with WPEE and MPE | |
| CN105678343A (en) | Adaptive-weighted-group-sparse-representation-based diagnosis method for noise abnormity of hydroelectric generating set | |
| Deng et al. | Application of BP neural network and convolutional neural network (CNN) in bearing fault diagnosis | |
| Li et al. | Gear pitting fault diagnosis using raw acoustic emission signal based on deep learning | |
| CN114897138A (en) | System fault diagnosis method based on attention mechanism and depth residual error network | |
| CN115795275A (en) | Bearing Fault Diagnosis Method Based on Variational Mode Decomposition and Wavelet Joint Denoising | |
| Desmet et al. | Leak detection in compressed air systems using unsupervised anomaly detection techniques | |
| CN106301610A (en) | The adaptive failure detection of a kind of superhet and diagnostic method and device | |
| CN104156628A (en) | A ship radiation signal recognition method based on multi-core learning discriminant analysis | |
| Rahimilarki et al. | Time-series deep learning fault detection with the application of wind turbine benchmark | |
| Wang et al. | Multi-scale convolutional neural network fault diagnosis based on attention mechanism | |
| CN113657664A (en) | System and method for predicting state of equipment in marine engine room |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20150520 |