CN113505639A - TPE-XGboost-based rotating machine multi-parameter health state evaluation method - Google Patents
TPE-XGboost-based rotating machine multi-parameter health state evaluation method Download PDFInfo
- Publication number
- CN113505639A CN113505639A CN202110590308.6A CN202110590308A CN113505639A CN 113505639 A CN113505639 A CN 113505639A CN 202110590308 A CN202110590308 A CN 202110590308A CN 113505639 A CN113505639 A CN 113505639A
- Authority
- CN
- China
- Prior art keywords
- tpe
- xgboost
- model
- parameter
- factor
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 230000036541 health Effects 0.000 title claims abstract description 29
- 238000011156 evaluation Methods 0.000 title claims description 9
- 238000000034 method Methods 0.000 claims abstract description 75
- 238000012549 training Methods 0.000 claims abstract description 32
- 238000005457 optimization Methods 0.000 claims abstract description 18
- 238000001914 filtration Methods 0.000 claims abstract description 5
- 238000000605 extraction Methods 0.000 claims description 5
- 238000003745 diagnosis Methods 0.000 abstract description 4
- 238000013210 evaluation model Methods 0.000 abstract 2
- 238000010801 machine learning Methods 0.000 description 11
- 238000009826 distribution Methods 0.000 description 9
- 238000012544 monitoring process Methods 0.000 description 9
- 238000012360 testing method Methods 0.000 description 8
- 230000015556 catabolic process Effects 0.000 description 6
- 238000006731 degradation reaction Methods 0.000 description 6
- 230000008569 process Effects 0.000 description 6
- 230000003862 health status Effects 0.000 description 5
- 239000000203 mixture Substances 0.000 description 5
- 238000009827 uniform distribution Methods 0.000 description 5
- 230000006870 function Effects 0.000 description 4
- 230000007246 mechanism Effects 0.000 description 3
- 238000011160 research Methods 0.000 description 3
- 238000006243 chemical reaction Methods 0.000 description 2
- 238000010276 construction Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000002474 experimental method Methods 0.000 description 2
- 238000002372 labelling Methods 0.000 description 2
- 238000005070 sampling Methods 0.000 description 2
- 238000012706 support-vector machine Methods 0.000 description 2
- 230000001960 triggered effect Effects 0.000 description 2
- 238000013528 artificial neural network Methods 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 229920001971 elastomer Polymers 0.000 description 1
- 239000000806 elastomer Substances 0.000 description 1
- 230000004927 fusion Effects 0.000 description 1
- 239000010687 lubricating oil Substances 0.000 description 1
- 238000005461 lubrication Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 239000003921 oil Substances 0.000 description 1
- 239000004033 plastic Substances 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 238000010998 test method Methods 0.000 description 1
- 229920002725 thermoplastic elastomer Polymers 0.000 description 1
Images
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
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2415—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on parametric or probabilistic models, e.g. based on likelihood ratio or false acceptance rate versus a false rejection rate
-
- 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
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Signal Processing (AREA)
- Probability & Statistics with Applications (AREA)
- Testing Of Devices, Machine Parts, Or Other Structures Thereof (AREA)
Abstract
A rotating machinery multi-parameter health state assessment method based on TPE-XGboost, aiming at vibration signals collected on mechanical equipment, firstly, a Butterworth filter is used for filtering to obtain two parts higher than 1KHz and lower than 1KHz, and then five vibration sensitive parameters including an effective value, a kurtosis factor, a peak value factor, a skewness factor and a margin factor are extracted; then, training sample equalization is carried out by using a SMOTE-ENN method; and finally, using a TPE optimization algorithm to perform XGboost model to designate a hyper-parameter space for iterative optimization, and training under the optimal parameters to obtain an evaluation model. Two practical industrial cases prove that the evaluation model obtained by the method can output the state probability, can fully reflect the health state information of mechanical equipment, and improve the fault diagnosis accuracy.
Description
Technical Field
The invention relates to a health assessment method for a rotary machine, in particular to a health assessment method for a multi-parameter fusion rotary machine based on TPE-XGboost.
Background
Automatic identification of equipment faults is an important research field in the field of industrial internet. Generally, large-scale equipment includes many small and medium-sized rotating machine components, and it is essential for healthy and stable operation of the large-scale equipment to accurately predict the health state of the small and medium-sized rotating machines. The sensor is arranged on the rotary mechanical equipment, so that a plurality of parameters can be monitored, and whether the equipment has faults or not can be identified by monitoring certain specific parameters. When diagnosing equipment, fault diagnosis engineers observe a plurality of parameters, analyze and integrate information transmitted by the parameters, and accordingly accurately evaluate the health state of the equipment. Generally, the more parameters that are observed, the more accurate the health status of the device is reflected by these parameters. But this increases the difficulty of the engineers and reduces the monitoring efficiency to some extent. In addition, in order to find the tiny faults of the machine as early as possible, the health state monitoring method can also predict the operation data in a short time in the future according to the previous operation data so as to achieve early warning in time.
In order to solve the problems of increased working difficulty and reduced monitoring efficiency caused by simultaneously observing a large number of parameters, some methods construct a specific health index representing the health state of equipment according to a plurality of observed parameters, and the method often has higher requirements on expert knowledge of engineers. Generally, the more parameters are observed, the more accurate the health status of the equipment is reflected, but this further increases the difficulty of work and reduces the efficiency and effectiveness of monitoring.
To ameliorate the problems of increased operational difficulty and reduced monitoring efficiency resulting from the simultaneous observation of a large number of parameters, some methods attempt to construct a specified index to reflect the operational state of the plant. Such methods include both simple and complex exponential methods. The simple index method mainly refers to the fact that specific characteristic parameters are extracted from original vibration signals to construct a health index model, the complex index method is mostly developed based on experimental data, for example, a test method is adopted, the cross-correlation coefficient of normal bearing vibration signals and known performance degradation bearing signals is calculated, and the cross-correlation coefficient is used as a health index to represent the health state of the bearing and is used for predicting the residual life of the bearing. However, the data sets used for modeling and comparison are experimental data, and thus model generalization may be low.
In recent years, machine learning methods have proven effective in solving the above problems. Detection and diagnosis of faults, such as for degradation of components in the production system and wear of machinery, is achieved using ANN techniques; and identifying the fault mode by using a support vector machine and a clustering method, and the like. The above methods are well validated in various industrial cases. However, the neural network algorithm is based on traditional statistics of progressive theory, and only when the number of learning samples tends to infinity, the recognition performance can be theoretically guaranteed. Machine learning methods such as support vector machines have high requirements on model hyper-parameters, and the problem of unbalanced training samples cannot be effectively solved.
Disclosure of Invention
The invention aims to solve the problems of poor model performance, low efficiency of hyper-parameter optimization and low prediction precision when training samples are unbalanced in the prior art. The method provided by the invention is an automatic machine learning method which integrates an SMOTE-ENN method, a TPE (Tree-structured park Estimator) optimization algorithm and XGboost (eXtreme Gradient Boosting Tree). The method can effectively solve the problems of unbalanced training samples and few fault samples in the research in the field, a high-performance training set is constructed, the TPE optimization algorithm is used for optimizing the model hyper-parameters, and the complex and low-efficiency hyper-parameter debugging is avoided. The method can evaluate the health state of the equipment in service on line by off-line training of the existing full-life-cycle data sample.
The method mainly comprises the following specific processes in practical application:
(1) off-line training:
a) and acquiring full life cycle operation data of the equipment, and performing a small amount of labeling on the obvious equipment state by combining an expert experience method.
b) And (2) separating two parts of waveform data which are lower than 1KHz and higher than 1KHz from each period of waveform by using a high-low pass filtering method, and extracting five vibration sensitive time domain parameters from the two parts of data: effective value, kurtosis factor, peak factor, skewness factor, margin factor.
c) And expanding the labeled data set by using a SMOTE-ENN method to obtain a training set and a test set of the model.
d) And (3) specifying a target hyper-parameter space for XGboost optimization, and optimizing the XGboost model on a training set and a testing set by using a TPE (thermoplastic elastomer) method.
e) And selecting an optimal hyper-parameter combination to construct an XGboost model, and training on a training set and a test set.
f) And storing the obtained optimal model structure.
(2) On-line assessment
a) And extracting the same characteristic parameters from the actually monitored operation data.
b) And inputting the characteristic parameters into a model trained in an off-line stage, and obtaining probability values x of the equipment in different states.
d) If the probability of x in the early warning state is rapidly increased and breaks through 80%, the device is gradually separated from the normal state, and an early warning mechanism is triggered when an early fault occurs; if the probability that x belongs to the fault state is rapidly increased and breaks through 80%, the equipment enters a late-stage fault state, an emergency alarm mechanism is triggered, and workers are further guided to operate the equipment.
2. The technical scheme adopted by the invention is as follows:
a rotating machinery multi-parameter health state assessment method based on TPE optimization and an XGboost model is characterized in that a SMOTE-ENN sample balancing method, a TPE optimization algorithm and the XGboost model are combined, and the known full-life-cycle samples of equipment are used for online assessment of the health state of the equipment.
(1) The SMOTE-ENN method is very suitable for the conditions of unbalanced samples, few fault samples and large normal samples in the field of fault diagnosis. The SMOTE-ENN method combines a SMOTE oversampling method and a nearest neighbor (KNN) algorithm, and is mainly used for solving the problem that the training effect of a model is poor due to unbalanced training samples in the field of machine learning.
In practical engineering applications, full-life samples of known equipment are easy to obtain, however, in this case, most samples are sample data of the equipment in a healthy state. For normal samples, the data occupation ratio of equipment degradation and serious fault states is small, and the model inerting can be caused by directly training the model on the training set, namely, the equipment is judged to be in a normal state, so that high precision can be obtained, and the machine learning model construction is not facilitated.
Therefore, the training sample construction method of the invention utilizes the SMOTE-ENN which is a good sample balance method to effectively solve the model inerting phenomenon caused by unbalanced training samples, can better construct a machine learning model and better evaluate the health state of equipment.
(2) The basic principle of the TPE optimization algorithm is as follows: the model hyper-parametric configuration space is converted into non-parametric density distribution, and the configuration space can be represented by uniform distribution, logarithmic uniform distribution, quantitative logarithmic uniform distribution and classification variables. The TPE algorithm converts it: uniform distribution → a truncated Gaussian mixture model, logarithmic uniform distribution → an exponential truncated Gaussian mixture model, classification → weight distribution classification, and continuously forming prior knowledge in the process of super-parameter optimization so as to more efficiently obtain the optimal super-parameter configuration of the model.
In the training phase, the main hyper-parameter search range of the model and the objective function optimized by the TPE algorithm are specified firstly. And continuously iterating the machine learning model in the training sample constructed by the SMOTE-ENN method, and finishing the optimization after the iteration times are finished or the specified conditions are met. And obtaining the optimal hyper-parameters of the corresponding machine learning model after the optimization is finished.
(2) The XGboost model belongs to an integrated model in a machine learning model, is an improved boosting algorithm of GBDT and has the advantages of high training speed, high prediction precision and the like.
And inputting the model hyperparameters obtained through the TPE optimization algorithm into the XGboost, and training and learning in training samples. At the same time, to prevent the model from being over-trained, resulting in overfitting, an early stop mechanism should also be introduced. After the model is learned, the XGboost model with the optimal structure can be obtained.
And inputting the monitoring data of the equipment in service into the model to obtain the probability values of the equipment in different health states. The smaller the probability of the normal state, the more the equipment operation deviates from the normal state.
3. The invention has the following advantages and prominent effects:
the method provided by the invention can be effectively applied to equipment health monitoring under the conditions of small samples, few samples and unbalanced samples; the method can predict different states of the equipment and give probability values of the equipment in different states; the method provided by the invention can avoid a complex hyper-parameter optimization process of the machine learning model and avoid the problem of low prediction performance of the machine learning model caused by poor hyper-parameter selection; the method provided by the invention has novel thought, combines the advantages of SMOTE-ENN, TPE and XGboost models, and enables the equipment state evaluation to be more accurate; and high-frequency and low-frequency data are separated by using a filtering method, and corresponding vibration sensitive characteristic parameters are extracted, so that the model has strong robustness and generalization. In practical engineering application, the method can successfully predict the time from the normal state to the degradation period to the failure period of a certain practical rotating machine under the condition that an expert cannot efficiently and accurately predict through a traditional method.
In conclusion, the method is simple and feasible, high in flexibility, wide in application range and convenient to apply to engineering practice.
Drawings
FIG. 1 is a training flow chart of a rotating machine multi-parameter health status evaluation method based on TPE-XGboost.
Fig. 2 is a timing chart of characteristic parameters extracted from a pump 4005 according to an embodiment of a rotating machine multi-parameter health status evaluation method based on TPE-XGBoost.
Fig. 3 is a schematic diagram of a rotational machine multi-parameter health state evaluation method based on TPE-XGBoost according to the present invention, in which a SMOTE-ENN method is used to perform sample equalization.
FIG. 4 shows the result of the application of the rotating machine multi-parameter health status evaluation method based on TPE-XGboost in the embodiment of the invention, A4005 pump.
Fig. 5 is a timing chart of characteristic parameters extracted from three groups of bearings in an embodiment of a rotating machine multi-parameter health state evaluation method based on TPE-XGBoost in an IMS center double-row ball bearing full life experiment.
FIG. 6 shows the application results of three sets of three bearings in an IMS center double-row ball bearing life-cycle test in an embodiment of a TPE-XGboost-based rotating machine multi-parameter health state assessment method of the present invention.
Detailed Description
The present invention is further described below with reference to examples. The scope of the present invention is not limited by these examples. The present invention will be described in further detail with reference to the accompanying drawings.
The TPE-XGboost-based health prediction method designed by the invention is applied to an actual A4005 pump fault case, the whole offline training process is shown in figure 1, the data acquisition interval is 8 seconds, the sampling frequency is 25600Hz, and the data point length of each file is 16384.
Full life cycle vibration sensor data is first collected from an old pump of the same model as the pump. The data set is collected by sensors installed on the old pump, which contains all data information from the time the old pump is just put into operation to the time the old pump fails. Carrying out filtering and feature extraction work on the full life cycle data, and extracting five vibration sensitive time domain parameters of two parts (high and low) of data higher than 1KHz and lower than 1 KHz: the effective value (RMS), kurtosis factor (KUR), peak factor (CF), skewness factor (SKEW), and margin factor (CL) constitute a ten-dimensional feature vector, as shown in fig. 2. And (3) using expert experience to label a small number of samples in the data of the whole life cycle to obtain three state labels of normal state, degradation state and fault state. There are 4100 samples in the normal state, 300 samples in the degraded state, and 100 samples in the fault state, which belong to serious unbalanced samples.
Training samples were equalized using the SMOTE-ENN method. The SMOTE method (Synthetic priority updating technique) is to reduce the error of the classifier by using a self-help method and a K-nearest neighbor method and generating new data similar to Minority class observation based on a feature space. The method comprises the following specific steps:
let A denote a minority of classes, arbitrarilyXiE.g. A, calculating the distance from the sample to all samples in the minority class sample set A by taking the Euclidean distance as a standard to obtain XiK nearest neighbor samples, randomly selecting one sample from the nearest neighbor samples, namely Xij(j ═ 1,2, …, n); at XiAnd Xij(j is 1,2, …, n) to construct new minority sample YjThis process is equivalent to randomly selecting one on the line connecting the two samples.
Yj＝Xi+rand(0,1)×(Xij-Xi)
In the formula, rand (0,1) represents a random number in the interval (0, 1). The schematic diagram of the synthesized sample is shown in FIG. 3. The SMOTE-ENN method is a combination of the SMOTE and ENN methods. Generating a new few samples by using a SMOTE method, obtaining a new data set ND, predicting each sample in the new data set ND by using a K neighbor (K is 5 in the text) method, and if the prediction result is different from the actual class label, rejecting the sample to obtain a training set. In the obtained training set, there are 3350 normal samples, 4058 degraded samples and 4077 failed samples.
And (4) specifying a hyper-parameter space of the XGboost model, and performing iterative optimization on the XGboost model by using a TPE (thermal plastic elastomer) method. TPE spatially converts the hyper-parameters into a nonparametric density distribution, modeling the p (x | y) process. The conversion mode includes 3 kinds of conversion from even distribution to truncated Gaussian mixture distribution, logarithmic even distribution to exponential stage Gaussian mixture distribution and discrete distribution to re-weighted discrete distribution. By using different observations (x) in nonparametric densities1,x2,…,xk) Alternatively, the super parameter set of the TPE may use learning algorithms with different densities. Its density is defined as:
wherein l (x) is represented by an observed value { x }iIs less than y*G (x) is the observed value { x }iAn objective function F (x) of y or more*The density composition of (a).TPE Algorithm Using y*As quantile γ for the observed value y. By maintaining an ordered list of observation data in the observation domain H, the run time of each iteration of the TPE algorithm can be scaled linearly in | H | and the optimized feature dimension, where the desired boost (EI) is:
finally, let γ equal p (y)<y*) And the two equations above are brought into the following equation, so that at each iteration, x is returned where the maximum EI point is obtained*。
p(x)＝∫p(x|y)p(y)dy＝γl(x)+(1-γ)g(x)
The XGboost model is specified to have an objective function of each iteration as follows:
wherein:
where T is a tree of leaf nodes, j is the index of each leaf node, i represents the ith sample in the data set, yiIn order to be the true value of the value,
for the predicted value in t-1 iterations, the set of samples contained on the leaf with index j is defined as IjAnd gamma is a self-defined parameter controlling the number of leaves. The formula for carrying out the hyper-parametric optimization by using the TPE method is as follows:
wherein, F (x) represents an objective function of XGboost; x is the parameter at which the best result is obtained, x represents the hyper-parameter space.
After the optimal model structure and parameters in the offline training state are obtained, the same characteristics of the operating data of the service equipment are extracted and input into the model, and the probability values of the model in three health states can be obtained. The experimental results obtained are shown in fig. 4. In the pump source monitoring system, the low alarm is displayed about 1 hour before the failure occurs (sampling point 8700). The results obtained with this method showed a degraded signal at sample point 5434, about 8 hours earlier.
Meanwhile, the method is applied to the IMS center double-row ball bearing life-cycle experiment. The experimental data of the group of bearings in the whole life is provided by the IMS center of the university of Cincinnati. The test bearing model was RexnordZA-2115 double row ball bearing. The bearing life test divides into three groups altogether, four bearings of every group, and the test condition is: the shaft speed was 2000rpm and the radial load was 6000 lbs. Lubricating oil is adopted for lubrication, and a magnetic plug is arranged on the oil return pipeline. The test was stopped when the deposit adhering to the magnetic plug exceeded a certain level. The data acquisition interval was 10 minutes, with 20480 data points per file.
The third group of bearing with the third number (#3-3) is used for analyzing the full-life data, 6321 data files are used in total, and the collected characteristic parameters are shown in fig. 5, so that the outer ring fault finally occurs. And (3) labeling the model by using an expert experience method, and training the TPE-XGboost model by using the method to obtain an optimal model.
The model trained on the bearing full life cycle sample #3-3 was applied to the full life data of the second group of bearings #2-1 for analysis, and the analysis results are shown in fig. 6 (black line indicates low alarm). The system sends out degradation alarm signals in advance around the sample point 400, sends out strong fault signals at the sample point 560, and then always gives an alarm, which is consistent with the actual situation.
The practical case research shows that the method has good generalization performance, can early warn the failure trend in advance, and has practical application value.
Claims (5)
1. A rotating machinery multi-parameter health state assessment method based on TPE-XGboost mainly comprises multi-parameter extraction, sample equalization and equipment health state assessment of a TPE-XGboost model, and is characterized in that: the multi-parameter extraction method comprises five vibration sensitive parameters obtained after high-low pass filtering, wherein sample equalization is to balance training samples by using a SMOTE-ENN method, and the TPE-XGboost model is to firstly use a TPE algorithm to optimize super parameters of a specified parameter space and then train an XGboost model with an optimal super parameter structure to complete equipment health state evaluation.
2. The multi-parameter extraction method of claim 1, wherein: the waveform is filtered by a Butterworth filter to obtain two parts of a part higher than 1KHz and a part lower than 1KHz, and then five vibration sensitive parameters including an effective value, a kurtosis factor, a peak value factor, a skewness factor and a margin factor are extracted.
3. A method of sample equalization as claimed in claim 1, characterized by: generating a few samples by using a SMOTE oversampling method, judging the generated samples by using an ENN (edited KNN) method, and removing the samples if a prediction result is different from an actual class label to obtain a balanced sample.
4. The TPE-XGBoost method of claim 1, wherein: firstly, a hyper-parameter space of an XGboost (eXtreme Gradient Boosting Tree) model is appointed, then, a TPE (Tree-structured park Estimator) is used for optimizing hyper-parameters of the model, and model training is carried out under the condition of the optimal hyper-parameters.
5. A method for assessing the health of a device using the method of claim 1, wherein: the method comprises the steps of multi-parameter extraction, sample equalization and automatic optimization of the XGboost model hyper-parameters by adopting a TPE optimization algorithm.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110590308.6A CN113505639B (en) | 2021-05-28 | 2021-05-28 | Rotary machine multi-parameter health state assessment method based on TPE-XGBoost |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110590308.6A CN113505639B (en) | 2021-05-28 | 2021-05-28 | Rotary machine multi-parameter health state assessment method based on TPE-XGBoost |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113505639A true CN113505639A (en) | 2021-10-15 |
| CN113505639B CN113505639B (en) | 2024-03-22 |
Family
ID=78008596
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110590308.6A Active CN113505639B (en) | 2021-05-28 | 2021-05-28 | Rotary machine multi-parameter health state assessment method based on TPE-XGBoost |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113505639B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114707706A (en) * | 2022-03-17 | 2022-07-05 | 广西科学院 | Diesel engine fault diagnosis method based on unbalanced data |
| CN114969990A (en) * | 2022-08-02 | 2022-08-30 | 中国电子科技集团公司第十研究所 | Multi-model fused avionic product health assessment method |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107941537A (en) * | 2017-10-25 | 2018-04-20 | 南京航空航天大学 | A kind of mechanical equipment health state evaluation method |
| CN109063308A (en) * | 2018-07-26 | 2018-12-21 | 北京航空航天大学 | A kind of health evaluating method based on depth quantum learning |
| CN110685868A (en) * | 2019-10-25 | 2020-01-14 | 长沙理工大学 | Wind turbine generator fault detection method and device based on improved gradient elevator |
| CN111695209A (en) * | 2020-05-13 | 2020-09-22 | 东南大学 | Rotary machine small sample health assessment method driven by meta-depth learning |
| CN112507479A (en) * | 2020-11-19 | 2021-03-16 | 电子科技大学 | Oil drilling machine health state assessment method based on manifold learning and softmax |
-
2021
- 2021-05-28 CN CN202110590308.6A patent/CN113505639B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107941537A (en) * | 2017-10-25 | 2018-04-20 | 南京航空航天大学 | A kind of mechanical equipment health state evaluation method |
| US20190285517A1 (en) * | 2017-10-25 | 2019-09-19 | Nanjing Univ. Of Aeronautics And Astronautics | Method for evaluating health status of mechanical equipment |
| CN109063308A (en) * | 2018-07-26 | 2018-12-21 | 北京航空航天大学 | A kind of health evaluating method based on depth quantum learning |
| CN110685868A (en) * | 2019-10-25 | 2020-01-14 | 长沙理工大学 | Wind turbine generator fault detection method and device based on improved gradient elevator |
| CN111695209A (en) * | 2020-05-13 | 2020-09-22 | 东南大学 | Rotary machine small sample health assessment method driven by meta-depth learning |
| CN112507479A (en) * | 2020-11-19 | 2021-03-16 | 电子科技大学 | Oil drilling machine health state assessment method based on manifold learning and softmax |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114707706A (en) * | 2022-03-17 | 2022-07-05 | 广西科学院 | Diesel engine fault diagnosis method based on unbalanced data |
| CN114969990A (en) * | 2022-08-02 | 2022-08-30 | 中国电子科技集团公司第十研究所 | Multi-model fused avionic product health assessment method |
| CN114969990B (en) * | 2022-08-02 | 2022-10-14 | 中国电子科技集团公司第十研究所 | Multi-model fused avionic product health assessment method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113505639B (en) | 2024-03-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Wen et al. | A new two-level hierarchical diagnosis network based on convolutional neural network | |
| CN102112933B (en) | Error detection method and system | |
| Ren et al. | Multi-bearing remaining useful life collaborative prediction: A deep learning approach | |
| CN112418277B (en) | Method, system, medium and equipment for predicting residual life of rotating machine parts | |
| Sugumaran et al. | Fault diagnostics of roller bearing using kernel based neighborhood score multi-class support vector machine | |
| Sugumaran et al. | Automatic rule learning using decision tree for fuzzy classifier in fault diagnosis of roller bearing | |
| CN115034248A (en) | Automatic diagnostic method, system and storage medium for equipment | |
| Hu et al. | Dynamic degradation observer for bearing fault by MTS–SOM system | |
| Pan et al. | A deep learning network via shunt-wound restricted Boltzmann machines using raw data for fault detection | |
| Patro et al. | A hybrid approach estimates the real-time health state of a bearing by accelerated degradation tests, Machine learning | |
| CN110110768A (en) | Fault Diagnosis of Roller Bearings based on Concurrent Feature study and multi-categorizer | |
| CN112765890B (en) | Dynamic domain adaptive network-based multi-working-condition rotating machine residual life prediction method | |
| CN113505639B (en) | Rotary machine multi-parameter health state assessment method based on TPE-XGBoost | |
| CN114295377A (en) | CNN-LSTM bearing fault diagnosis method based on genetic algorithm | |
| CN114358124B (en) | New fault diagnosis method for rotary machinery based on deep countermeasure convolutional neural network | |
| Mosallam et al. | Component based data-driven prognostics for complex systems: Methodology and applications | |
| CN116593157A (en) | Complex working condition gear fault diagnosis method based on matching element learning under small sample | |
| CN115859077A (en) | Multi-feature fusion motor small sample fault diagnosis method under variable working conditions | |
| Techane et al. | Rotating machinery prognostics and application of machine learning algorithms: Use of deep learning with similarity index measure for health status prediction | |
| Hu et al. | Fault diagnosis based on multi-scale redefined dimensionless indicators and density peak clustering with geodesic distances | |
| Kizito et al. | The application of random forest to predictive maintenance | |
| Yang et al. | Self-attention parallel fusion network for wind turbine gearboxes fault diagnosis | |
| CN115774955A (en) | Intelligent health degree assessment method for wind power gear box | |
| Yang et al. | Evaluation method of rotating machinery health state based on TPE-XGBoost | |
| wahhab Lourari et al. | An ANFIS-based Framework for the Prediction of Bearing’s Remaining Useful Life |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |