CN111260823B - Fault diagnosis method based on I/O (input/output) measuring point fault dependency matrix - Google Patents
Fault diagnosis method based on I/O (input/output) measuring point fault dependency matrix Download PDFInfo
- Publication number
- CN111260823B CN111260823B CN202010044961.8A CN202010044961A CN111260823B CN 111260823 B CN111260823 B CN 111260823B CN 202010044961 A CN202010044961 A CN 202010044961A CN 111260823 B CN111260823 B CN 111260823B
- Authority
- CN
- China
- Prior art keywords
- test
- fault
- vehicle door
- data
- state
- 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.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G07—CHECKING-DEVICES
- G07C—TIME OR ATTENDANCE REGISTERS; REGISTERING OR INDICATING THE WORKING OF MACHINES; GENERATING RANDOM NUMBERS; VOTING OR LOTTERY APPARATUS; ARRANGEMENTS, SYSTEMS OR APPARATUS FOR CHECKING NOT PROVIDED FOR ELSEWHERE
- G07C5/00—Registering or indicating the working of vehicles
- G07C5/08—Registering or indicating performance data other than driving, working, idle, or waiting time, with or without registering driving, working, idle or waiting time
- G07C5/0808—Diagnosing performance data
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Testing Of Devices, Machine Parts, Or Other Structures Thereof (AREA)
- Testing And Monitoring For Control Systems (AREA)
Abstract
The invention discloses a fault diagnosis method based on an I/O (input/output) measuring point fault dependency matrix, which comprises the following steps of: s1: acquiring vehicle door state data; s2: structuring processing of data; s3: selecting a test point; s4: constructing a fault test dependency matrix; s5: and (4) fault diagnosis reasoning. According to the method, a mathematical modeling method based on data is adopted to perform structured processing on the control signal data of the vehicle door controller and model fault testing, so that accurate association can be established between a test result and a fault, data acquisition is convenient, and the method is wide in applicability; the test points are selected according to the structural function characteristics of the vehicle door, normal vehicle door data are used as a reference to establish a standard to select the test points and screen data, the test points are selected more reasonably, and the relative time is used as the standard to establish a fault test interval for testing in consideration of the difference of acquisition time.
Description
Technical Field
The invention relates to the field of urban rail transit, in particular to a fault diagnosis method based on an I/O (input/output) measuring point fault dependency matrix.
Background
Urban rail transit has a great deal of advantages such as the freight volume is big, efficient, the energy consumption is low, intensification, it is convenient to take, safe comfortable, is the important way that solves the urban traffic jam problem, realizes that city spatial layout adjusts and city equilibrium development, and subway realization operation automation is also one of the trade development trend, and computer control and system safety control will improve the operation degree of automation of subway. While the subway functions are more and more perfect, the structure, the component parts, the signal transmission system and the like of the vehicle system develop rapidly, so that the occurrence types of faults are more complicated, particularly, the subway door needs to be opened and closed hundreds of times every day, the degradation process of the door parts is faster, and the frequency of the door system faults is increased under the influence of factors such as environment, human factors and the like. At present, an intelligent diagnosis algorithm for vehicle door faults is not complete enough, and a large number of parameters related to vehicle doors still need to be detected on site manually, fault reasons are checked, and time and labor are consumed.
The existing fault diagnosis methods are mainly divided into three types of models, knowledge and data driving, wherein the models are mainly suitable for objects capable of establishing mathematical models, and the methods are not suitable for diagnosing faults of vehicle doors because the vehicle doors have complicated structures and are difficult to establish accurate analytical models; the diagnosis method based on knowledge is divided into an expert system and a symbolic directed graph, is suitable for objects which cannot establish a model and have insufficient sensor quantity, but has overhigh diagnosis cost for a system with large data volume; the expert system needs high professional knowledge and long-term experience accumulation, the failure type of the subway door is complex, and the expert system is not completely applicable in actual operation; the main fault diagnosis method is also based on a data-driven fault diagnosis method, and the fault diagnosis is realized by analyzing and processing online and offline data in the system operation process.
Disclosure of Invention
The purpose of the invention is as follows: aiming at the problems, the invention aims to provide a fault diagnosis method based on an I/O (input/output) measuring point fault dependency matrix, so as to realize the diagnosis of the faults of the door system of the subway vehicle.
The technical scheme is as follows: the invention provides a fault diagnosis method based on an I/O (input/output) measuring point fault dependency matrix, which comprises the following steps of:
s1: acquiring vehicle door state data: acquiring data capable of representing the running state of the vehicle door, and acquiring an I/O digital signal of a door controller in a vehicle door system, wherein the I/O digital signal comprises a plurality of signal variables of the door control system, including a control signal issued by the door controller to the door and a state signal fed back by the door controller;
s2: and (3) structuring processing of data: carrying out structuralization processing on the acquired vehicle door running state data according to the trigger time and the trigger state, extracting the jump of a variable, and converting the data form into sequence pair information;
s3: selecting a test point: by utilizing a fault testability analysis method of a multi-signal model, according to the variable quantity and the triggering condition of normal fault-free vehicle door data, hierarchical design is carried out on the basis of the system structure and the function of a vehicle door, and proper test points are selected among modules of a vehicle door system which are easy to have faults;
s4: constructing a fault test dependency matrix: constructing a test interval at a selected test point by using a 3 sigma criterion, testing the state of a system variable of a fault car door, and if the variable state does not trigger jump or generates error jump in the test interval, defining that the I/O (input/output) quantity test in the corresponding test point does not pass to obtain a fault dependency matrix D reflecting the correlation between the system fault type and the test signal;
s5: fault diagnosis reasoning: and according to the result of the fault dependence matrix D at each test point, eliminating normal components step by using a fault diagnosis strategy of multi-signal model reasoning, and finally determining the fault type of the component and the vehicle door, thereby establishing the correlation between the test vector result and the fault mode.
Further, the data structuring process in step S2 includes:
collecting the on-line data of the I/O parameters of the gate controller by using zjJ is variable index, j is 1,2,3, …, n, n is I/O parameter number, data structure processing is carried out according to trigger state and trigger time of I/O quantity, processed I/O data is order information, and z is used for mathematical descriptionjZ representing a structured processed I/O parameter variable of a testing point of a gate controller in a gate opening and closing recordjExpressed as:
where 0 and 1 represent the state of the I/O quantity, tj，1，tj，2，tj，3,…,
Denotes zjState transition ofTime, JjDenotes zjAll state transition times.
Further, the test points selected in step S3 are:
the basic composition of the multi-signal model is as follows:
(1)C＝{c1,c2,…,cm-m elements of the system;
(2)TP＝{tp1,tp2,…,tpmm test points;
(3) t { T1, T2, …, tn }, n available tests;
(4) a dependency matrix D between the fault types of the vehicle door system and the test points;
and the C set is a component of the vehicle door system, all the triggering moments of jump variables in the normal fault-free vehicle door opening and closing process are selected as the testing points to form a testing point set TP, the triggering states are used as a testing interval set T, tests are added at all the testing points to detect the variable states of the door controller in the door opening and closing process, and the testing results form a correlation dependency matrix D.
Further, the process of constructing the fault testing dependency matrix in step S4 is as follows:
and 3, selecting the variables of the test points to be slightly different at each triggering time, taking the signal triggering time of the door opening and closing variable as the initial 0 time, analyzing the mean value and the standard deviation sigma of the time interval between the triggering time of each variable and the initial 0 time in normal vehicle door data, constructing a fault test interval with the size of 3 sigma at the mean value of the triggering time of each test point, and obtaining a test interval set T ═ { T [ T ] T [ T ] of the vehicle door systemj，k-3σ，tj，k+3σ]}；
After obtaining the fault test interval, it needs to be determined that the fault of each module can be observed by the test point, that is, the dependency relationship between the fault and the test is expressed in the form of a correlation matrix, which is defined as follows:
whereinThe row vector is the test result of all test points under a certain fault mode; the column vector represents all faults detected by the test point, reflects the fault detection capability of the test point, and is in a test interval TjIf the variable zjIf the state of the fault dependency matrix D is not jumped or error jumped, the test is not passed, the element corresponding to the fault dependency matrix is assigned to be 1, otherwise, the element is assigned to be 0, and therefore modeling of the triggering time of the system I/O quantity and the associated rows of the triggering state and the actual running state of the system is completed, and the fault dependency matrix D is obtained.
Further, the fault diagnosis inference process in step S5 is:
s51, the multiple signal model method assigns four different states to each component of the system: normal, fault, suspect and unknown;
s52, initially assuming all the components are in unknown states;
s53, testing the element part, if passing, the state is updated to normal, otherwise, the state is updated to suspicion;
s54, obtaining the fault meta-parts from the meta-parts in the doubtful state by excluding the normal meta-parts;
defining C as component assembly, T as test assembly, PT as test assembly without alarm, FT as test assembly with alarm, R as dependency matrix between component assembly and test, and C (test)j) Is made by testjDetecting a failed component bond;
suppose testi∈PT，testjE is FT; definition F ═ FjIs from C (t)j) Removing the suspected component set after the normal component; g is a set of known normal components, S is a set of suspected components, and B is a set of known fault components;
the fault reasoning process of the multi-signal model is expressed as follows:
a. for testi∈PT，G←∪testi∈PTC(testi)；
b. For testj∈FT，F＝{fj}←C(testj)－G，S←S∪{fj}；
c. If | { fj} | ═ 1, then B ← B £ u { f {, f }j}，S←S－B；
Where, either ← stands for collection update, | { fjAnd | is the aggregate potential.
Has the advantages that: compared with the prior art, the invention has the following remarkable advantages:
1. the invention adopts a mathematical modeling method based on data, carries out structured processing on the control signal data of the car door controller, and models fault testing, so that accurate association can be established between a test result and a fault, the data acquisition is convenient, and the method has wide applicability;
2. the test points are selected according to the structural and functional characteristics of the vehicle door, normal vehicle door data are used as a reference to establish a standard to select the test points and screen data, the test points are selected more reasonably, and the relative time is used as the standard to establish a fault test interval for testing in consideration of the difference of the acquisition time;
3. and based on the fault reasoning of the multi-signal model, the fault component can be diagnosed according to the fault dependency matrix, a diagnosis rule is formed after the fault mode of the fault dependency matrix is determined, a rule base is established, and the diagnosis speed is increased.
Drawings
FIG. 1 is a flow chart of a fault diagnosis method based on an I/O measurement point fault dependency matrix.
Detailed Description
Example (b): the method of the invention is used for fault diagnosis of a subway sliding door system manufactured by Nanjing company, a flow chart is shown in figure 1, and the method comprises the following specific steps:
s1: obtaining sliding plug door state data
The method comprises the steps that I/O data of a door controller in each door opening and closing process are collected through a vehicle door system, the I/O data comprise 20 variables such as a door opening and closing train line, a lock in-place switch and a close in-place switch, and collected variable signals are digital values and are divided into two states of '0' and '1';
s2: structured processing of data
When a variable triggers jump in acquisition, all the variables in the system are acquiredThe state and time of the variable at that time are collected, so that the variables of the system need to be structurally processed, the trigger state and the trigger time of each variable are extracted to form sequential information, and the collected data is assumed to be in the form of zj0,0, …,1,0,0, 0 where j is 1,2,3, …, n, zjWhen the ith state of the variable is different from the state of i +1, the jump occurs, and the (i + 1) th state and the trigger time are extracted to form an order pair, which is as follows:
where 0 and 1 represent the state of the I/O quantity, tj,1，tj,2，tj,3…
Denotes zjTime of state transition ofjDenotes zjAll state transition times.
S3: selecting test points
The car door system mainly comprises an electronic door control system, a mechanical transmission and a door body, and mainly analyzes the hopping logic sequence of door controller signals in the process of opening and closing the door as follows:
and (3) door opening jump logic: the method comprises the following steps of (1) a zero-speed train line → a door permission control signal → a train line opened door → an inner side vehicle door indicator lamp → a lock in place switch) → a safety interlock loop → a close in place switch, and then the door opening is completed;
and (3) door closing jump logic: the zero-speed train line → the door permission control signal → the door closing train line → the inner side vehicle door indicator lamp and the buzzer → the closing position switch → the locking position switch → the safety interlocking loop, and the door closing is finished;
controlling the vehicle door to work according to the logic sequence, selecting triggering moments of variables such as a train line for opening the door, a train line for closing the door in place and a variable for locking the door in place at positions where the vehicle door is easy to have internal component faults and external faults in all variable jumping extracted in the step S2 as test points, adding tests at the test points to judge the states of the components and the whole door, selecting n variable test points in the door opening and closing process, and setting the variable test points as the test points T1, T2,.
S4: building fault test dependency matrices
In the process of opening and closing the door each time, the triggering time of each variable is not fixed, and the triggering time point of a certain variable is used as a test reference point, so that the method is inaccurate, so that an initial reference point needs to be selected, and a fault test interval is constructed near a test point by using relative time to replace a single test point for testing;
41) counting I/O data of multiple normal door opening and closing, analyzing the distribution rule of n variable triggering time selected in the step S3, taking the triggering of the door opening and closing signal variable as T1, and re-marking the triggering time as the initial T0Calculating the time interval between the trigger time of the rest n-1 variables and T1, and analyzing each test point and T by counting multiple times of normal vehicle door data0Normal distribution, mean, standard deviation of time intervals;
data points which are in accordance with normal distribution have the probability of falling in a 3 sigma interval of 97 percent, so that +3 sigma and-3 sigma intervals are constructed at the mean value of each Test point interval, the original Test points T1, T2, the original data points T2, the original data points T2, the original data points T, T2, the original data points Tn are replaced by new intervals Test1, Test2, Test.
42) In the test interval, all variable trigger states are detected, if no jump or error jump occurs, the I/O quantity in the corresponding test point is defined not to pass the test, the corresponding element of the fault dependence matrix is assigned to be 1, otherwise, the assignment is 0, and the form of the fault dependence matrix D of the system I/O quantity is obtained and is shown in table 1.
TABLE 1 Fault dependency matrix example
f0Indicating no failure mode, f1、f2、f3Representing three different faults, the row vector being all test points in a certain fault modeThe test result of (1); the column vectors represent all faults detected for the test interval.
S5: fault diagnosis reasoning
And (4) as for the fault dependence matrix obtained in the step (4), adopting a fault reasoning strategy of a multi-signal model, and endowing each component of the system with four different states by the multi-signal model method: normal, fault, suspect and unknown. Initially assuming all components are in unknown states; if the test indicates that the element passes, e.g. f in the matrix0A row, which indicates that all components pass the test, and the status is updated to normal; fail the test, e.g. f1The Test section of Test1 in the row does not pass, which indicates that the component of the structure point has an abnormality and the state is updated to be suspicious; the malfunctioning components are obtained from these in the suspect state by excluding the normal components.
Defining C as component assembly, T as test assembly, PT as test assembly without alarm, FT as test assembly with alarm, R as dependency matrix between component assembly and test, and C (test)j) Is made by testjThe faulty component binding is detected.
Suppose testi∈PT，testjE is FT; definition F ═ FjIs from C (t)j) Removing the suspected component set after the normal component; g is a set of known normal components, S is a set of suspected components, and B is a set of known fault components; the fault reasoning process of the multi-signal model is expressed as follows:
a. for testi∈PT，G←∪testi∈PTC(testi)；
b. For testj∈FT，F＝{fj}←C(testj)－G，S←S∪{fj}；
c. If | { fj} | ═ 1, then B ← B £ u { f {, f }j}，S←S－B；
Where, either ← stands for collection update, | { fjAnd | is the aggregate potential.
Claims (3)
1. The fault diagnosis method based on the I/O measuring point fault dependency matrix is characterized by comprising the following steps of:
s1: acquiring vehicle door state data: acquiring data capable of representing the running state of the vehicle door, and acquiring an I/O digital signal of a door controller in a vehicle door system;
s2: and (3) structuring processing of data: the method comprises the following steps of carrying out structuralized processing on collected vehicle door running state data according to trigger time and trigger state, extracting jump of variables, and converting a data form into sequence pair information, and specifically comprises the following steps:
collecting the on-line data of the I/O parameters of the gate controller by using zjJ is variable index, j is 1,2,3, …, n, n is I/O parameter number, data structure processing is carried out according to trigger state and trigger time of I/O quantity, processed I/O data is order information, and z is used for mathematical descriptionjZ representing a structured processed I/O parameter variable of a testing point of a gate controller in a gate opening and closing recordjExpressed as:
wherein 0 and 1 represent the state of the I/O amount,
denotes zjTime of state transition ofjDenotes zjAll state transition times of (1);
s3: selecting a test point: by utilizing a fault testability analysis method of a multi-signal model, according to the variable quantity and the triggering condition of normal fault-free vehicle door data, hierarchical design is carried out on the basis of the system structure and the function of a vehicle door, and proper test points are selected among modules of a vehicle door system which are easy to have faults:
the basic composition of the multi-signal model is as follows:
(1)C＝{c1,c2,…,cm-m elements of the system;
(2)TP＝{tp1,tp2,…,tpmm test points;
(3) t { T1, T2, …, tn }, n available tests;
(4) a dependency matrix D between the fault types of the vehicle door system and the test points;
the C set is a vehicle door system component, all triggering moments of jump variables in the normal fault-free vehicle door opening and closing process are selected as testing points to form a testing point set TP, the triggering states are used as a testing interval set T, tests are added to all the testing points to detect the variable states of the door controller in the door opening and closing process, and the testing results form a correlation dependency matrix D;
s4: constructing a fault test dependency matrix: constructing a test interval at a selected test point by using a 3 sigma criterion, testing the state of a system variable of a fault car door, and if the variable state does not trigger jump or generates error jump in the test interval, defining that the I/O (input/output) quantity test in the corresponding test point does not pass to obtain a fault dependency matrix D reflecting the correlation between the system fault type and the test signal;
s5: fault diagnosis reasoning: and (3) reasoning a fault diagnosis strategy by using a multi-signal model according to the result of the fault dependence matrix D at each test point, gradually eliminating normal components, and finally determining the fault type of the component and the vehicle door, thereby establishing the correlation between the test vector result and the fault mode.
2. The method for diagnosing the faults based on the I/O measuring point fault dependency matrix as claimed in claim 1, wherein the step S3 is to select the variables of the measuring points to be slightly different at each triggering time, use the signal triggering time of the on-off variable as an initial 0 time, analyze a mean value and a standard deviation σ of a time interval between the triggering time of each variable and the initial 0 time in the normal vehicle door data, and construct a fault measuring interval with a size of 3 σ at the mean value of the triggering time of each measuring point, so as to obtain a measuring interval set (T ═ T [ T ] of the vehicle door systemj,k-3σ，tj,k+3σ]}；
After obtaining the fault test interval, it needs to be determined that the fault of each module can be observed by the test point, that is, the dependency relationship between the fault and the test is expressed in the form of a correlation matrix, which is defined as follows:
the row vector is the test result of all test points under a certain fault mode; and the column vectors represent all faults detected by the test points, reflect the fault detection capability of the test points, and in a test interval, if the variable state does not jump or jump erroneously, the test does not pass, the element corresponding to the fault dependence matrix is assigned to be 1, otherwise, the element is assigned to be 0, so that the modeling of the triggering time of the system I/O quantity and the associated rows of the triggering state and the actual operation state of the system is completed, and the fault dependence matrix D is obtained.
3. The method for fault diagnosis based on the I/O station fault dependency matrix as claimed in claim 1, wherein the step S5 is performed by:
s51: the multi-signal model approach assigns four different states to each component of the system: normal, fault, suspect and unknown;
s52: initially assuming all components are in unknown states;
s53: the test element part is used for updating the state to be normal if the test element part passes through the test element part, and otherwise, updating to be suspicious;
s54: the failed component is obtained from the suspected components by excluding the normal components;
defining C as component assembly, T as test assembly, PT as test assembly without alarm, FT as test assembly with alarm, R as dependency matrix between component assembly and test, and C (test)j) Is made by testjDetecting a failed component bond;
suppose testi∈PT，testjE is FT; definition F ═ FjIs from C (t)j) Removing the suspected component set after the normal component; g is a set of known normal components, S is a set of suspect components, and B is a known faultA set of meta-parts;
the fault reasoning process of the multi-signal model is expressed as follows:
a. for testi∈PT，G←∪testi∈PTC(testi)；
b. For testj∈FT，F＝{fj}←C(testj)－G，S←S∪{fj}；
c. If | { fj} | ═ 1, then B ← B £ u { f {, f }j}，S←S－B；
Where, either ← stands for collection update, | { fjAnd | is the aggregate potential.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010044961.8A CN111260823B (en) | 2020-01-16 | 2020-01-16 | Fault diagnosis method based on I/O (input/output) measuring point fault dependency matrix |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010044961.8A CN111260823B (en) | 2020-01-16 | 2020-01-16 | Fault diagnosis method based on I/O (input/output) measuring point fault dependency matrix |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111260823A CN111260823A (en) | 2020-06-09 |
| CN111260823B true CN111260823B (en) | 2021-12-24 |
Family
ID=70945209
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010044961.8A Active CN111260823B (en) | 2020-01-16 | 2020-01-16 | Fault diagnosis method based on I/O (input/output) measuring point fault dependency matrix |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111260823B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112308403B (en) * | 2020-10-29 | 2023-11-21 | 北京国信会视科技有限公司 | Urban rail vehicle door opening and closing abnormality detection method |
| CN113221496B (en) * | 2021-05-06 | 2022-06-14 | 电子科技大学 | Fault diagnosis method based on three-dimensional testability analysis model |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101980225A (en) * | 2010-11-16 | 2011-02-23 | 中国人民解放军63908部队 | Method for implementing testability analysis and diagnosis decision system for electronic products |
| CN106406295A (en) * | 2016-12-02 | 2017-02-15 | 南京康尼机电股份有限公司 | Rail transit vehicle door system fault diagnosis and early warning method based on multiple conditions |
-
2020
- 2020-01-16 CN CN202010044961.8A patent/CN111260823B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101980225A (en) * | 2010-11-16 | 2011-02-23 | 中国人民解放军63908部队 | Method for implementing testability analysis and diagnosis decision system for electronic products |
| CN106406295A (en) * | 2016-12-02 | 2017-02-15 | 南京康尼机电股份有限公司 | Rail transit vehicle door system fault diagnosis and early warning method based on multiple conditions |
Non-Patent Citations (4)
| Title |
|---|
| 冷轧液压AGC系统的建模与故障诊断方法研究;孟宪锋;《中国优秀硕士学位论文全文数据库 工程科技Ⅰ辑》;20170315;全文 * |
| 城轨列车客室车门系统故障诊断方法研究——基于改进的TOPSIS法与贝叶斯网络;徐霖;《中国优秀硕士学位论文全文数据库 工程科技Ⅱ辑》;20151015;第17-20页 * |
| 基于多信号模型的故障诊断策略优化设计及在液压AGC系统中的应用;黄丹丹;《中国优秀硕士学位论文全文数据库 信息科技辑》;20190215;参见第17-20页 * |
| 数据驱动的地铁车门微小故障智能诊断方法;施文;《仪器仪表学报》;20190731;全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111260823A (en) | 2020-06-09 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Coble et al. | Applying the general path model to estimation of remaining useful life | |
| CN102208028B (en) | Fault predicting and diagnosing method suitable for dynamic complex system | |
| CN111639467B (en) | Aero-engine service life prediction method based on long-term and short-term memory network | |
| CN111274737A (en) | Method and system for predicting remaining service life of mechanical equipment | |
| CN112131212A (en) | Hybrid cloud scene-oriented time sequence data anomaly prediction method based on ensemble learning technology | |
| CN110807257A (en) | Method for predicting residual life of aircraft engine | |
| Chen et al. | A hybrid prognostic method based on gated recurrent unit network and an adaptive Wiener process model considering measurement errors | |
| CN105955241B (en) | A kind of quality fault localization method based on joint data-driven production process | |
| CN108490923B (en) | System design method for detecting and positioning tiny faults of electric traction system | |
| CN112101431A (en) | Electronic equipment fault diagnosis system | |
| CN102945311A (en) | Method for diagnosing fault by functional fault directed graph | |
| CN104035431B (en) | The acquisition methods of kernel functional parameter and system for non-linear process monitoring | |
| CN111260823B (en) | Fault diagnosis method based on I/O (input/output) measuring point fault dependency matrix | |
| CN108304661A (en) | Diagnosis prediction method based on TDP models | |
| CN103927343A (en) | Comprehensive diagnosis and prediction ability verifying method of PHM (prognostics and health management) system | |
| CN102608519B (en) | Circuit failure diagnosis method based on node information | |
| CN111860839A (en) | Shore bridge fault monitoring method based on multi-signal fusion and Adam optimization algorithm | |
| Son et al. | Deep learning-based anomaly detection to classify inaccurate data and damaged condition of a cable-stayed bridge | |
| CN105137354A (en) | Motor fault detection method based on nerve network | |
| CN112395684A (en) | Intelligent fault diagnosis method for high-speed train running part system | |
| Morgan et al. | Detection and diagnosis of incipient faults in heavy-duty diesel engines | |
| CN114281054A (en) | LSTM-CNN-based airplane remote fault analysis method and system | |
| Xiao et al. | Integrated system-level prognosis for hybrid systems subjected to multiple intermittent faults | |
| CN113919207A (en) | Top-level open type electrical intelligent health monitoring and management system | |
| CN113627358A (en) | Multi-feature fusion turnout intelligent fault diagnosis method, system and equipment |
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 |