CN109508745A - The detection method of gas turbine gascircuit fault based on Bayesian network model - Google Patents
The detection method of gas turbine gascircuit fault based on Bayesian network model Download PDFInfo
- Publication number
- CN109508745A CN109508745A CN201811354221.3A CN201811354221A CN109508745A CN 109508745 A CN109508745 A CN 109508745A CN 201811354221 A CN201811354221 A CN 201811354221A CN 109508745 A CN109508745 A CN 109508745A
- Authority
- CN
- China
- Prior art keywords
- bayesian network
- gas turbine
- network model
- parameter
- gascircuit
- 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
- 238000001514 detection method Methods 0.000 title claims abstract description 20
- 238000005457 optimization Methods 0.000 claims abstract description 23
- 238000000034 method Methods 0.000 claims abstract description 12
- 238000007689 inspection Methods 0.000 claims description 14
- 238000004140 cleaning Methods 0.000 claims description 3
- 238000003708 edge detection Methods 0.000 claims description 3
- 230000002068 genetic effect Effects 0.000 claims description 3
- 238000004064 recycling Methods 0.000 claims description 3
- 238000005070 sampling Methods 0.000 claims description 3
- 235000015170 shellfish Nutrition 0.000 claims description 3
- 230000010354 integration Effects 0.000 claims 1
- 238000005259 measurement Methods 0.000 abstract description 6
- 238000011897 real-time detection Methods 0.000 abstract description 3
- 230000002159 abnormal effect Effects 0.000 abstract description 2
- 238000007621 cluster analysis Methods 0.000 abstract 1
- 238000007781 pre-processing Methods 0.000 abstract 1
- 239000007789 gas Substances 0.000 description 37
- 241001269238 Data Species 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 2
- 239000000567 combustion gas Substances 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 238000002485 combustion reaction Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000010338 mechanical breakdown Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
- G06F18/232—Non-hierarchical techniques
- G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
- G06F18/23213—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions with fixed number of clusters, e.g. K-means clustering
-
- 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
- G06F18/24155—Bayesian classification
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/06—Energy or water supply
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/44—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Business, Economics & Management (AREA)
- Economics (AREA)
- Probability & Statistics with Applications (AREA)
- Health & Medical Sciences (AREA)
- Multimedia (AREA)
- Public Health (AREA)
- Water Supply & Treatment (AREA)
- General Health & Medical Sciences (AREA)
- Human Resources & Organizations (AREA)
- Marketing (AREA)
- Primary Health Care (AREA)
- Strategic Management (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Testing And Monitoring For Control Systems (AREA)
Abstract
一种基于贝叶斯网络模型的燃气轮机气路故障的检测方法，将实时信号采集得到的燃气轮机气路部件的数据生成数据集，从数据集中得到正常工况参数、待测异常参数、训练集和测试集，经预处理和聚类分析得到离散化后的优化训练集和优化测试集，然后通过优化测试集对经初始化以及参数优化后的贝叶斯网络进行测试，得到用于实时检测燃气轮机气路系统当前的工况状态的优化贝叶斯网络模型，从而对系统故障进行检测。本发明提出了具体的贝叶斯网络模型结构学习和参数学习的方法，进而建立测量参数与燃气轮机正常工况参数之间的相关模型，实现了燃气轮机气路系统的在线故障检测。
A method for detecting gas turbine gas circuit faults based on a Bayesian network model, generating a data set from the data of gas turbine gas circuit components acquired by real-time signal acquisition, and obtaining normal operating condition parameters, abnormal parameters to be measured, training set and The test set, the discretized optimized training set and the optimized test set are obtained after preprocessing and cluster analysis, and then the Bayesian network after initialization and parameter optimization is tested through the optimized test set, and the results are obtained for real-time detection of gas turbine gas. The optimal Bayesian network model of the current working condition of the road system is used to detect system faults. The invention proposes a specific Bayesian network model structure learning and parameter learning method, and further establishes a correlation model between the measurement parameters and the gas turbine normal operating condition parameters, and realizes the on-line fault detection of the gas turbine gas circuit system.
Description
Technical field
The present invention relates to a kind of technology of field of thermal power, specifically a kind of combustion based on Bayesian network model
The detection method of gas-turbine gas path failure.
Background technique
Structure is complicated for gas turbine, and longtime running is in the state of high revolving speed, high temperature and pressure and high stress, building ring
Border is severe, Yi Fasheng mechanical breakdown and gas path failure.Gas turbine is made of gas path component and subsystem, wherein gas turbine
If gas path component breaks down, it will seriously affect the availability of gas turbine, it is therefore desirable to the gas path component of gas turbine
Fault detection is carried out, discovering device is abnormal in advance, the major accident caused by escalation of fault is effectively prevent, to guarantee combustion gas
Turbine safe and stable operation.For gas turbine, gas path failure occurs generally in the form of single failure, is in normal in engine
When working condition, each component and complete machine have the corresponding feature of unfaulty conditions, and thermal parameter is in normal range；It is working
When breaking down in the case that environment is constant, the variation of component capabilities will lead to the variation of measurement parameter.Therefore, it is surveyed by analysis
The variation of amount parameter can estimate the variation of gas turbine nominal situation parameter, and then analyze the nominal situation of gas turbine gas circuit
Situation.The prior art can only mostly detect fault mode, or lack targetedly Structure learning and parametric learning method, not from
The actual demand of turbine engine failure detection is set out.
Summary of the invention
The present invention In view of the above shortcomings of the prior art, proposes a kind of gas turbine based on Bayesian network model
The detection method of gas path failure establishes the correlation model between measurement parameter and gas turbine nominal situation parameter, realizes combustion gas
The online fault detection of turbine air-channel system.
The present invention is achieved by the following technical solutions:
The data for the gas turbine gas path component that the present invention obtains real time signal aquisition generate data set, from data set
Nominal situation parameter, anomaly parameter to be measured, training set and test set are obtained, after preprocessed and clustering obtains discretization
Optimize training set and optimal inspection collection, then by optimal inspection collection to the Bayesian network after initialized and parameter optimization
It is tested, obtains the optimization Bayesian network model of the work condition state current for real-time detection gas turbine air-channel system,
To be detected to the system failure.
The data set operates normally data and fault data comprising multiple period gas turbine gas path components, through drawing
Get nominal situation parameter, anomaly parameter to be measured.
The data that the training set and test set is randomly selected in nondimensionalization treated data set, are preferably instructed
Practice and integrates with the ratio of test set as 4:1.
The pretreatment includes: cleaning exceptional value to delete invalid data, iterative median filter to reduce measuring instrument
Noise and combination gradient and the edge detection of Laplacian operator are to reduce wrong early warning.
The clustering refers to: being based on nominal situation parameter and anomaly parameter to be measured to instruction using K-Means algorithm
Practice collection and carry out continuous variable clustering, make training set discretization and obtain optimization training set and optimal inspection collection, specifically includes
Following steps are as follows:
A. k sample is randomly selected from data set as initial cluster centre C={ c1, c2 ..., ck }；
B. for data sets in each sample, calculate the distance for arriving k cluster centre, and it is nearest to assign it to distance
Cluster centre corresponding in class；
C. after all samples are assigned, the center of k cluster is recalculated；
D. k cluster centre being calculated for the second time is calculated gained with first time to be compared, if cluster centre is sent out
Changing then jumps to step b, and otherwise algorithm terminates.
The Bayesian network, using being based on, genetic algorithm is combined with K2 (GA-K2) algorithm and Bayesian Information is quasi-
Then the method for (BIC) is established to obtain, specifically includes the following steps:
I. initial population is generated, each individual is a kind of ordering structure of all-network node；
Ii. K2 algorithm is implemented to each of initial population individual, learns corresponding bayesian network structure out；
Iii. BIC scoring is carried out under current data set to the network structure that step ii is obtained, the expression formula of BIC scoring isWherein: first item is metrology structure model ζ and dataFitting
Degree, Section 2 are the penalty terms of structural model complexity；
Iiii. it regard BIC scoring as fitness function, is selected in original population, generate next-generation new population；
Iiiii. judge whether to reach default the number of iterations, be, carry out in next step, otherwise return step ii；
Iiiiii. the highest individual of fitness in current population is obtained final as node sequence with K2 algorithm
Bayesian network structure.
The initialization refers to: optimization training set being based on historical experience and expertise is modeled to know expert
Knowledge is converted into prior probability distribution, specifically includes the following steps:
1) Posterior probability distribution p (θ | D) is calculated based on data set；
2) the conditional probability distribution p (D of next sample is calculatedm+1|D)。
The parameter optimization refers to: after generating analog sample using importance sampling method, recycling Bayesian Estimation whole
Priori knowledge optimization network parameter study is closed, specifically includes the following steps:
1., according to network condition probability tables, n are obtained by way of generating random number mutually independently from root node
Sample；
2. approximate calculation Posterior Mean；
3. the conditional probability θ between network node is considered as stochastic variable, it is assumed that θ obeys known distribution p (θ), finds out survey
The sample Joint Distribution of the data set D obtained
4. Posterior distrbutionp p (θ | D) can be calculated according to the distribution of θ；
5. calculating p (x | D)=∫ p (x | θ) p (θ | D) d θ according to Bayesian Estimation formula；
6. p (x | D) is substituted into posterior probability p (wi| x, D) to obtain the optimal estimation of conditional probability θ.
The test refers to: carrying out Bayesian network model test based on optimal inspection collection, output infers accuracy rate most
Excellent Bayesian network model.
Technical effect
Compared with prior art, the invention proposes the sides of specific Bayesian network model Structure learning and parameter learning
Method, and then the correlation model between measurement parameter and gas turbine nominal situation parameter is established, realize gas turbine gas circuit system
The online fault detection of system.
Detailed description of the invention
Fig. 1 is the principle of the present invention figure；
Fig. 2 is schematic diagram of the invention；
Fig. 3 is present system schematic diagram.
Specific embodiment
As shown in figure 3, being a kind of gas turbine gascircuit fault based on Bayesian network model that the present embodiment is related to
Detection system, comprising: signal acquisition module, Bayesian network training module and Bayesian network test module, in which: signal
The data for the gas turbine gas path component that acquisition module input collects in real time, Bayesian network training module and signal acquisition
Module is connected and optimization training set and optimal inspection collection after transmitting discrete, Bayesian network test module and Bayesian network
Training module is connected and transmits the Bayesian network after initialized and parameter optimization, and the output of Bayesian network test module pushes away
The optimal Bayesian network model of disconnected accuracy rate.
As depicted in figs. 1 and 2, the data for the gas turbine gas path component that the present embodiment obtains real time signal aquisition generate
Data set obtains nominal situation parameter, anomaly parameter to be measured, training set and test set, preprocessed and cluster from data set
Analysis obtains the optimization training set and optimal inspection collection after discretization, then by optimal inspection collection to initialized and parameter
Bayesian network after optimization is tested, and the excellent of the work condition state current for real-time detection gas turbine air-channel system is obtained
Change Bayesian network model, to detect to the system failure.
The data set acquires multiple period gas turbine gas path components and operates normally data and fault data, altogether
31608 datas, it is divided to obtain nominal situation parameter, anomaly parameter to be measured, in which: the data of normal operation are divided into normally
There are three duty parameters, and fault data, which is divided into measurement parameter, eight.
The data that the training set and test set is randomly selected in nondimensionalization treated data set, are preferably instructed
Practice and integrates with the ratio of test set as 4:1.
The nominal situation parameter includes: flow factor SW, efficiency factor SE and leaving area SA.
The measurement parameter includes: three temperature T1~T3, three pressure P1~P3, low-pressure shaft revolving speed XNLP and high-pressure shaft
Revolving speed XNHP.
The training set refers to: randomly selecting from the data through nondimensionalization treated gas turbine gas path component
80% data, totally 25286.
The test set refers to: remaining 20% data after training set composition, totally 6322.
The pretreatment includes: cleaning exceptional value to delete invalid data, iterative median filter to reduce measuring instrument
Noise and combination gradient and the edge detection of Laplacian operator are to reduce wrong early warning.
The clustering refers to: being based on nominal situation parameter and anomaly parameter to be measured to instruction using K-Means algorithm
Practice collection and carry out continuous variable clustering, make training set discretization and obtain optimization training set and optimal inspection collection, specifically includes
Following steps are as follows:
A. k sample is randomly selected from data set as initial cluster centre C={ c1, c2 ..., ck }；
B. for data sets in each sample, calculate the distance for arriving k cluster centre, and it is nearest to assign it to distance
Cluster centre corresponding in class；
C. after all samples are assigned, the center of k cluster is recalculated；
D. k cluster centre being calculated for the second time is calculated gained with first time to be compared, if cluster centre is sent out
Changing then jumps to step b, and otherwise algorithm terminates.
The foundation of the bayesian network structure is combined with K2 (GA-K2) algorithm and shellfish using based on genetic algorithm
The method of this information criterion (BIC) of leaf, specifically includes the following steps:
I. initial population is generated, each individual is a kind of ordering structure of all-network node；
Ii. K2 algorithm is implemented to each of initial population individual, learns corresponding bayesian network structure out；
Iii. BIC scoring is carried out under current data set to the network structure that step ii is obtained, the expression formula of BIC scoring isWherein: first item is metrology structure model ζ and dataFitting
Degree, Section 2 are the penalty terms of structural model complexity；
Iiii. it regard BIC scoring as fitness function, is selected in original population, generate next-generation new population；
Iiiii. judge whether to reach default the number of iterations, be, carry out in next step, otherwise return step ii；
Iiiiii. the highest individual of fitness in current population is obtained final as node sequence with K2 algorithm
Bayesian network structure.
The described initialization modeling be the training set after discretization is based on to historical experience and expertise model with
Prior probability distribution is converted by expertise, specifically includes the following steps:
1) Posterior probability distribution p (θ | D) is calculated based on data set；
2) the conditional probability distribution p (D of next sample is calculatedm+1|D)。
The Bayesian network parameters optimization refers to: after generating analog sample using importance sampling method, recycling shellfish
Priori knowledge optimization network parameter study is integrated in Ye Si estimation, specifically includes the following steps:
1., according to network condition probability tables, n are obtained by way of generating random number mutually independently from root node
Sample；
2. approximate calculation Posterior Mean；
3. the conditional probability θ between network node is considered as stochastic variable, it is assumed that θ obeys known distribution p (θ), finds out survey
The sample Joint Distribution of the data set D obtained
4. Posterior distrbutionp p (θ | D) can be calculated according to the distribution of θ；
5. calculating p (x | D)=∫ p (x | θ) p (θ | D) d θ according to Bayesian Estimation formula；
6. p (x | D) is substituted into posterior probability p (wi| x, D) to obtain the optimal estimation of conditional probability θ.
The test refers to: carrying out Bayesian network model test based on optimal inspection collection, output infers accuracy rate most
Excellent Bayesian network model.
The dynamic real-time update is to be updated the information newly collected obtained to network model.
It is described that carry out Bayesian network model test to test set be 6322 datas using test set to training
Bayesian model tested, since nominal situation parameter is three, infer the dimension of accuracy rate for three-dimensional, such as the following table 1 institute
Show, obtain following result:
| Infer accurate number | Infer accuracy rate | |
| At least one is correct | 6322 | 100.0% |
| At least two is correct | 6233 | 98.6% |
| At least three is correct | 6091 | 96.3% |
It is obtained by upper table, all infers correct probability up to 96.3%.
Above-mentioned specific implementation can by those skilled in the art under the premise of without departing substantially from the principle of the invention and objective with difference
Mode carry out local directed complete set to it, protection scope of the present invention is subject to claims and not by above-mentioned specific implementation institute
Limit, each implementation within its scope is by the constraint of the present invention.
Claims (10)
1. a kind of detection method of the gas turbine gascircuit fault based on Bayesian network model, which is characterized in that will believe in real time
The data of number gas turbine gas path component collected generate data set, and nominal situation parameter, to be measured is obtained from data set
Anomaly parameter, training set and test set, preprocessed and clustering obtain the optimization training set and optimal inspection after discretization
Collection, then tests the Bayesian network after initialized and parameter optimization by optimal inspection collection, obtains for real
When the current work condition state of detection gas turbine air-channel system optimization Bayesian network model, to be examined to the system failure
It surveys.
2. the detection method of the gas turbine gascircuit fault according to claim 1 based on Bayesian network model, special
Sign is the data set, operates normally data and fault data comprising multiple period gas turbine gas path components, divided to obtain
To nominal situation parameter, anomaly parameter to be measured.
3. the detection method of the gas turbine gascircuit fault according to claim 1 based on Bayesian network model, special
Sign is, the data that the training set and test set are randomly selected in nondimensionalization treated data set are preferably trained
Integrate and the ratio of test set is 4:1.
4. the detection method of the gas turbine gascircuit fault according to claim 1 based on Bayesian network model, special
Sign is that the pretreatment includes: cleaning exceptional value to delete invalid data, iterative median filter to reduce measuring instrument noise
And combine gradient and the edge detection of Laplacian operator to reduce wrong early warning.
5. the detection method of the gas turbine gascircuit fault according to claim 1 based on Bayesian network model, special
Sign is that the clustering refers to: being based on nominal situation parameter and anomaly parameter to be measured to training set using K-Means algorithm
Continuous variable clustering is carried out, training set discretization is made and obtains optimization training set and optimal inspection collection.
6. the detection method of the gas turbine gascircuit fault according to claim 1 based on Bayesian network model, special
Sign is the Bayesian network, is combined with K2 (GA-K2) algorithm and bayesian information criterion using based on genetic algorithm
(BIC) method is established to obtain.
7. the detection method of the gas turbine gascircuit fault according to claim 1 based on Bayesian network model, special
Sign is that the initialization refers to: optimization training set being based on historical experience and expertise and is modeled with by expertise
It is converted into prior probability distribution.
8. the detection method of the gas turbine gascircuit fault according to claim 1 based on Bayesian network model, special
Sign is that the parameter optimization refers to: after generating analog sample using importance sampling method, recycling Bayesian Estimation integration first
Test the study of knowledge optimization network parameter.
9. the detection method of the gas turbine gascircuit fault according to claim 1 based on Bayesian network model, special
Sign is that the test refers to: carrying out Bayesian network model test based on optimal inspection collection, output infers that accuracy rate is optimal
Bayesian network model.
10. the present invention relates to a kind of systems for realizing the above method, comprising: signal acquisition module, Bayesian network training module
And Bayesian network test module, in which: the gas turbine gas path component that signal acquisition module input collects in real time
Data, Bayesian network training module is connected with signal acquisition module and optimization training set and optimal inspection after transmitting discrete
Collection, Bayesian network test module are connected with Bayesian network training module and transmit the shellfish after initialized and parameter optimization
The optimal Bayesian network model of accuracy rate is inferred in this network of leaf, the output of Bayesian network test module.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811354221.3A CN109508745A (en) | 2018-11-14 | 2018-11-14 | The detection method of gas turbine gascircuit fault based on Bayesian network model |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811354221.3A CN109508745A (en) | 2018-11-14 | 2018-11-14 | The detection method of gas turbine gascircuit fault based on Bayesian network model |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109508745A true CN109508745A (en) | 2019-03-22 |
Family
ID=65748505
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811354221.3A Pending CN109508745A (en) | 2018-11-14 | 2018-11-14 | The detection method of gas turbine gascircuit fault based on Bayesian network model |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109508745A (en) |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110147849A (en) * | 2019-05-24 | 2019-08-20 | 吉林大学 | Electric control system of diesel engine weak link recognition methods under a kind of strong electromagnetic pulse |
| CN110162016A (en) * | 2019-05-30 | 2019-08-23 | 华北电力大学 | A kind of fault modeling method of gas turbine pneumatic actuator |
| CN110348094A (en) * | 2019-07-01 | 2019-10-18 | 中国人民解放军国防科技大学 | Petroleum pipeline leakage detection method and system based on influence network |
| CN110443344A (en) * | 2019-06-20 | 2019-11-12 | 中国地质大学（武汉） | A kind of momenttum wheel method for diagnosing faults and device based on K2ABC algorithm |
| CN111091278A (en) * | 2019-12-04 | 2020-05-01 | 湃方科技(天津)有限责任公司 | Method and device for constructing edge detection model for abnormal detection of mechanical equipment |
| CN111311059A (en) * | 2020-01-16 | 2020-06-19 | 成都大汇物联科技有限公司 | Knowledge graph-based water mill room fault diagnosis method |
| CN114297255A (en) * | 2021-12-17 | 2022-04-08 | 中电信数智科技有限公司 | Network quality work order fault early warning method based on log analysis |
| CN115964649A (en) * | 2022-12-13 | 2023-04-14 | 合肥泽众城市智能科技有限公司 | Indoor gas monitoring fault analysis model |
| CN116810825A (en) * | 2023-08-30 | 2023-09-29 | 泓浒(苏州)半导体科技有限公司 | Wafer conveying mechanical arm abnormality monitoring method and system in vacuum environment |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20100113635A (en) * | 2008-11-25 | 2010-10-21 | 미츠비시 쥬고교 가부시키가이샤 | Method for managing failure of facility |
| CN103308314A (en) * | 2013-06-18 | 2013-09-18 | 上海理工大学 | Automobile engine fault maintenance method based on Bayesian network models and multi-criteria decision analysis |
-
2018
- 2018-11-14 CN CN201811354221.3A patent/CN109508745A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20100113635A (en) * | 2008-11-25 | 2010-10-21 | 미츠비시 쥬고교 가부시키가이샤 | Method for managing failure of facility |
| CN103308314A (en) * | 2013-06-18 | 2013-09-18 | 上海理工大学 | Automobile engine fault maintenance method based on Bayesian network models and multi-criteria decision analysis |
Non-Patent Citations (6)
| Title |
|---|
| CARL A. PALMER 等: "COMBINING BAYESIAN BELIEF NETWORKS WITH GAS PATH ANALYSIS", 《INTERNATIONAL GAS TURBINE & AEROENGINE CONGRESS & EXHIBITION》 * |
| YOUNG K. LEE 等: "A Fault Diagnosis Method for Industrial Gas Turbines Using Bayesian Data Analysis", 《JOURNAL OF ENGINEERING FOR GAS TURBINES AND POWER》 * |
| 侯胜利 等: "基于免疫聚类分析的特征提取及其在发动机故障诊断中的应用", 《推 进 技 术》 * |
| 朱明敏: "贝叶斯网络结构学习与推理研究", 《中国优秀博士学位论文全文数据库 信息科技辑》 * |
| 简敏: "基于GA-K2算法的贝叶斯网络研究及在个人信用评估的应用", 《中国优秀硕士学位论文全文数据库 经济与管理科学辑》 * |
| 蒲星星: "基于模型的重型燃气轮机气路故障诊断研究", 《中国优秀博士学位论文全文数据库 工程科技Ⅱ辑》 * |
Cited By (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110147849A (en) * | 2019-05-24 | 2019-08-20 | 吉林大学 | Electric control system of diesel engine weak link recognition methods under a kind of strong electromagnetic pulse |
| CN110147849B (en) * | 2019-05-24 | 2022-03-25 | 吉林大学 | Method for identifying weak link of diesel engine electric control system under strong electromagnetic pulse |
| CN110162016A (en) * | 2019-05-30 | 2019-08-23 | 华北电力大学 | A kind of fault modeling method of gas turbine pneumatic actuator |
| CN110443344B (en) * | 2019-06-20 | 2022-05-06 | 中国地质大学（武汉） | Momentum wheel fault diagnosis method and device based on K2ABC algorithm |
| CN110443344A (en) * | 2019-06-20 | 2019-11-12 | 中国地质大学（武汉） | A kind of momenttum wheel method for diagnosing faults and device based on K2ABC algorithm |
| CN110348094A (en) * | 2019-07-01 | 2019-10-18 | 中国人民解放军国防科技大学 | Petroleum pipeline leakage detection method and system based on influence network |
| CN110348094B (en) * | 2019-07-01 | 2023-01-06 | 中国人民解放军国防科技大学 | Oil pipeline leak detection method and system based on influence network |
| CN111091278A (en) * | 2019-12-04 | 2020-05-01 | 湃方科技(天津)有限责任公司 | Method and device for constructing edge detection model for abnormal detection of mechanical equipment |
| CN111091278B (en) * | 2019-12-04 | 2023-09-08 | 湃方科技(天津)有限责任公司 | Edge detection model construction method and device for mechanical equipment anomaly detection |
| CN111311059A (en) * | 2020-01-16 | 2020-06-19 | 成都大汇物联科技有限公司 | Knowledge graph-based water mill room fault diagnosis method |
| CN111311059B (en) * | 2020-01-16 | 2023-08-29 | 成都大汇物联科技有限公司 | Waterwheel house fault diagnosis method based on knowledge graph |
| CN114297255A (en) * | 2021-12-17 | 2022-04-08 | 中电信数智科技有限公司 | Network quality work order fault early warning method based on log analysis |
| CN114297255B (en) * | 2021-12-17 | 2024-04-19 | 中电信数智科技有限公司 | Network quality work order fault early warning method based on log analysis |
| CN115964649A (en) * | 2022-12-13 | 2023-04-14 | 合肥泽众城市智能科技有限公司 | Indoor gas monitoring fault analysis model |
| CN116810825A (en) * | 2023-08-30 | 2023-09-29 | 泓浒(苏州)半导体科技有限公司 | Wafer conveying mechanical arm abnormality monitoring method and system in vacuum environment |
| CN116810825B (en) * | 2023-08-30 | 2023-11-03 | 泓浒(苏州)半导体科技有限公司 | Wafer conveying mechanical arm abnormality monitoring method and system in vacuum environment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109508745A (en) | The detection method of gas turbine gascircuit fault based on Bayesian network model | |
| CN109522600B (en) | Complex equipment residual service life prediction method based on combined deep neural network | |
| CN105046352B (en) | Water supply network leakage loss calculation method based on blood vessel bionic principle | |
| JP5292477B2 (en) | Diagnostic device and diagnostic method | |
| Wachla et al. | A method of leakage location in water distribution networks using artificial neuro-fuzzy system | |
| US11415469B2 (en) | Edge-cloud collaboration platform for intelligent coking monitoring of cracking furnace tubes and working method thereof | |
| CN114943179B (en) | Reliability assessment and remaining life prediction method based on multi-source degradation data fusion | |
| CN104077445B (en) | Accelerated life test statistical analysis technique based on fuzzy theory | |
| CN108508863A (en) | A kind of electromechanical equipment method for diagnosing faults based on gray model | |
| JP2006063986A (en) | Apparatus and method for assessing health of device | |
| CN104504509B (en) | A kind of thermal power plant's Consumption Difference Analysing System and its method using dynamic benchmark value | |
| KR20100083765A (en) | System and method for empirical ensemble-based virtual sensing of gas emission | |
| CN105221933A (en) | A kind of pipeline network leak detecting method in conjunction with resistance identification | |
| Xu et al. | New RUL prediction method for rotating machinery via data feature distribution and spatial attention residual network | |
| CN105512483A (en) | Remaining Life Prediction of Aeroengine Based on Damage Propagation Model and Data Analysis | |
| CN104295373B (en) | Diesel engine fault Forecasting Methodology based on three point models | |
| CN110826587B (en) | Turboshaft engine fault detection method based on improved weighted one-class support vector machine | |
| CN101237357B (en) | Online failure detection method for industrial wireless sensor network | |
| CN104713730A (en) | Method for determining degeneration rate of aircraft engine according to vibration signal | |
| CN115130731A (en) | Method, equipment and medium for detecting fault of wind driven generator | |
| Osigwe et al. | Integrated gas turbine system diagnostics: components and sensor faults quantification using artificial neural networks | |
| CN105203327A (en) | Gas channel measurement parameter selecting method applied to engine gas channel analysis | |
| CN117454145B (en) | A method for aircraft engine fault diagnosis based on multimodal sensors | |
| JP2015230576A (en) | Data creation system for plant diagnosis | |
| CN115687940A (en) | A method and device for evaluating equipment health based on bulldozer distance measurement |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190322 |
|
| RJ01 | Rejection of invention patent application after publication |