CN108921305A - A kind of component lifetime monitoring method - Google Patents
A kind of component lifetime monitoring method Download PDFInfo
- Publication number
- CN108921305A CN108921305A CN201810621885.5A CN201810621885A CN108921305A CN 108921305 A CN108921305 A CN 108921305A CN 201810621885 A CN201810621885 A CN 201810621885A CN 108921305 A CN108921305 A CN 108921305A
- Authority
- CN
- China
- Prior art keywords
- component
- data
- model
- lifetime
- 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
Classifications
-
- 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
- G06Q10/00—Administration; Management
- G06Q10/20—Administration of product repair or maintenance
-
- 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
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
- G06Q10/06393—Score-carding, benchmarking or key performance indicator [KPI] analysis
Landscapes
- Business, Economics & Management (AREA)
- Human Resources & Organizations (AREA)
- Engineering & Computer Science (AREA)
- Strategic Management (AREA)
- Economics (AREA)
- Entrepreneurship & Innovation (AREA)
- Physics & Mathematics (AREA)
- Quality & Reliability (AREA)
- Operations Research (AREA)
- Tourism & Hospitality (AREA)
- Marketing (AREA)
- General Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Development Economics (AREA)
- Educational Administration (AREA)
- Game Theory and Decision Science (AREA)
- Testing Resistance To Weather, Investigating Materials By Mechanical Methods (AREA)
Abstract
The present invention relates to the health controls of part of appliance and failure predication field, and in particular to a kind of component lifetime monitoring method.A kind of component lifetime monitoring method, which is characterized in that include the following steps:A)Establish the baseline model of component lifetime;B)According to the service condition of each component, baseline model is adjusted, the independent model of each component is obtained;C)The service condition of component and the independent model of component are compared, the lifetime monitoring information of component is obtained.Substantial effect of the invention is:With the service life, pressure test is combined, the quantity foot that obtains valid data sample, data acquisition comprehensively, for the lifetime than pair model foundation it is more careful, it is thus possible to lifetime of component is carried out more comprehensively and more accurately monitoring.
Description
Technical field
The present invention relates to the health controls of part of appliance and failure predication field, and in particular to a kind of component lifetime monitoring
Method.
Background technique
With the development of China's mechanotronic, the function of mechanical electronic equipment is more and more abundant and diversification, for
While the production and life of people bring very big improvement, also there is mechanical electronic equipment and tend to be complicated and insecure further
Problem.In order to make the electromechanical equipment comprising numerous components and electronic device keep the well operating status with safety, operator
The care and maintenance of electromechanical equipment must be organized regularly.The maintenance task of electromechanical equipment mainly has genuine to be responsible for and the at present
Tripartite contracts both of which, and the maintenance method almost all of use is that maintenance staff is periodically sent to carry out live maintenance operation.By
In lacking effective Supervision Measures, the maintenance of maintenance staff scene has that maintenance quality is not up to standard, leads to many electromechanical equipments
It runs in spite of illness, there are great security risks.According to counting in the industry, 8 percent tenth is that dimension the reason of electromechanical equipment failure and accident
Protect it is not in place caused by.China's economic development at present is rapid, and electromechanical equipment deployment amount and increment are in a high position, maintenance staff's number
Amount, which increases, not to be caught up with, cause maintenance staff work heavy and maintenance staff exist universal training it is insufficient with quality is uneven asks
Topic.And the growth of wage and training cost to maintenance staff, cause part maintenance contractor unit to force down and is adopted for maintenance equipment
The budget of purchase further reduced the quality and reliability of periodically live maintenance.There is also one for maintenance staff's periodically live maintenance
Fixed blindness, the unbalanced disadvantage of maintenance of " this, which is repaired, does not repair, should not repair and but repair ", is lack of pertinence, inefficiency.In maintenance
Task increasingly heavy today artificial periodically live maintenance model is only relied on, is not obviously able to satisfy electromechanical equipment safety and stability
The needs of operation.
Occur some technical solutions for improving maintenance on-site supervision effect at present, current electromechanical equipment can be alleviated and lacked
Abundant maintenance and maintenance inadequate resource and low-quality contradiction, but still comprehensive guidance cannot be provided for maintenance operation, no
Lifetime and the health condition of electromechanical equipment and component can be monitored.If the lifetime of electromechanical equipment and component can be accurately held
And health status, then it is able to carry out accurate maintenance.While greatly reducing maintenance task amount, additionally it is possible to improve maintenance quality.?
Component will damage or failure before can find, and countermeasure is taken to avoid adverse consequences.However at present for electromechanical equipment
And the faulty prediction of component and the method alarmed have the following problems:1)Modeling Theory is based on insufficient grounds, and algorithm model is simple, at present
Industry generallys use historical failure library as basis is compared, and only alarm can be being made to the major failure frequently occurred, to not having
Occurred or small probability failure still can not judge, even frequently judge by accident;2)Data sharing is difficult, effective sample number
Amount is insufficient, due to needing to accumulate fault data, only can just get sample data after equipment failure, lead to sample number
Long according to the collection period, sample data total amount is few, and lacks unified standard between operation system, and failure logging sample is independently stored,
It is difficult to be formed shared;3)There are pseudo- declining period prediction techniques, and industry directly uses a small amount of failure of electromechanical equipment and component at present
As warning information, this unilateral failure symptom judgement has also been brought into more for sign or main sensors acquisition data
Accuracy that is uncertain and reducing monitoring result;4)Rate of false alarm is high, prediction result diverging, due to lack Modeling Theory according to
According to, comparison basis is simple and crude, effective sample total amount of data is few and declining period prediction model is coarse, cause rate of false alarm high, prediction knot
Fruit poor astringency.
106952028 A of Chinese patent CN, publication date on July 14th, 2017, dynamoelectric equipment failure is examined in advance and health control
Method and system, including:Data acquisition obtains the data information of dynamoelectric equipment;Self diagnosis, to a certain dynamoelectric equipment in difference
Historical data information under operational mode and health status carries out feature extraction and model foundation, recycles the model of foundation that will work as
The data information that preceding state obtains is compared with historical data information, the current health shape of automatic identification this dynamoelectric equipment
State;Health status prediction, it is pre- according to the current health state of this dynamoelectric equipment obtained after self diagnosis and history health status
Survey the variation of this dynamoelectric equipment future health status;Cluster analysis, according to the current health state of separate unit dynamoelectric equipment to machine
Denso carries out cluster for the data information of more dynamoelectric equipments in cluster and analysis is compared, and obtains the health of more dynamoelectric equipments
State grade and risk distribution.Its technical solution still uses equipment and part history fault database benchmark as a comparison, there are
The effect problem that sample size is few, contrast model is coarse, causes the not comprehensive and accurate property of monitoring result poor.
103241658 B of Chinese patent CN, publication date on September 23rd, 2015, the vibrative mechanism based on Internet of Things
Health monitoring and safety pre-warning system, including:Thing network sensing layer, Internet of Things network layer and internet of things application layer, the Internet of Things
Net sensing layer, for acquiring the optical signal of vibrative mechanism health parameters, by optical signal demodulation at electric signal, and will sensing
Internet of Things network layer is transferred to after datagram number compression encapsulation;Internet of Things network layer, for receiving electrical signal data, to telecommunications
Number identifying processing is carried out, and sends internet of things application layer for the data after identifying processing;Internet of things application layer, by what is received
Data are calculated and determine whether to issue alarm signal.The present invention is based on technology of Internet of things, can be round-the-clock to goliath
Metal structure carry out Life cycle real time health monitoring and safe early warning, with no electromagnetic interference, precision height, broad quantum,
The features such as high reliablity, long service life.It is not comprehensive because data acquisition may be not present since it acquires strain information for mechanical structure
The problem of, although it can collect information needed, but lack effectively comprehensive comparison benchmark, comparison benchmark still relies on
The data of historical failure, thus still without solving the problems, such as that lifetime monitoring result is not comprehensively inaccurate.
Summary of the invention
The technical problem to be solved by the present invention is to:The skill of electromechanical equipment and the not comprehensive inaccuracy of component lifetime monitoring at present
Art problem.It proposes a kind of combination service life pressure testing data and historical failure data establishes the lifetime monitoring knot of contrast model
The more comprehensively more accurate component lifetime monitoring method of fruit.
In order to solve the above technical problems, the technical solution used in the present invention is:A kind of component lifetime monitoring method, packet
Include following steps:A)Establish the baseline model of component lifetime;B)According to the service condition of each component, baseline model is adjusted,
Obtain the independent model of each component;C)The service condition of component and the independent model of component are compared, the life of component is obtained
Phase monitoring information.Failure and lifetime monitoring use historical failure library benchmark as a comparison at present, using by accelerated life test
The component lifetime baseline model of acquisition benchmark as a comparison, can obtain more comprehensively more accurate monitoring result.Accelerating the longevity
Life test during, can unify in practical application and accelerated life test during used in sensor, make test data
It is higher with actual acquired data matching degree, improve the accuracy of lifetime monitoring.In baseline model, there can be multiple adjustable systems
Numerical value can determine the value of whole coefficients according to accelerated life test, form calibration coefficients;But in practical applications, due to
The deployed environment of each component is different, and the lifetime model of different components will lead to the model if being referred in a model and lack
Weary specific aim, thus, first according to arrangements of components environment, after calibration coefficients are corrected, each portion is established using in lifetime monitoring
The independent model of part improves the collocation degree of single component mode, and then improves lifetime monitoring precision and reliability, reduces and misses
Report rate.
Preferably, the method for building up of the baseline model includes the following steps:A1)Aging factor, the ring of definition component
Border factor and lifetime feedback;A2)Using engineering theory model and/or industry experience lifetime model as basic model;A3)It is right
Each aging factor carries out accelerated life test and/or simulation test, obtains single aging factor test data, substitutes into basic mould
Type obtains the lifetime model of single aging factor;A4)To the combination of each pair of single environment factor and single aging factor into
Row accelerated life test and/or simulation test obtain the test data of single environment factor and the combination of single aging factor, substitute into
Basic model obtains the lifetime model of single environment factor and the combination of single aging factor;A5)To multiple environmental factors and
Multiple aging factors carry out accelerated life test and/or simulation test, obtain multifactor data;A6)For single aging factor
Lifetime model and single environment factor and the lifetime model of single aging factor combination are manually provided with initial weight value
Functional relation form initial baseline model, establish the quasi- of weighted value and aging factor and environmental factor according to multifactor data
Molding type, using the model of fit and initial baseline model simultaneous as the lifetime baseline model of component.
Preferably, carrying out accelerated life test to aging factor, the data for obtaining single aging factor test data are adopted
Set method is:If aging factor is transient process from ambient condition to trystate, when aging factor is in trystate,
At interval of time t1 acquire aging factor data and lifetime feedback data and with time data correlation, conversely, then aging because
During element is from environmental condition change to trystate, aging factor data and lifetime feedback data are acquired at interval of time t2
And through conversion function and time data correlation, then when aging factor is in trystate, aging is acquired at interval of time t1
Factor data and lifetime feedback coefficient and with time data correlation;The combination of each pair of environmental factor and aging factor is carried out
Accelerated life test, the method for obtaining the test data of single environment factor and the combination of single aging factor are:If aging factor
And environmental factor is transient process from ambient condition to trystate, then is in test in aging factor and environmental factor
When state, every time t1 acquire aging factor data, environmental factor data and lifetime feedback data and with time data
Association, conversely, then in aging factor and environmental factor from ambient condition to during being in trystate, every the time
T2 acquires aging factor data, environmental factor data and lifetime feedback data and through converting function and time data correlation,
Then when aging factor and environmental factor are in trystate, aging factor data, environment are acquired at interval of time t1
Factor data and lifetime feedback coefficient and with time data correlation, wherein interval time t1 and t2 is set manually.Existing
During accelerated life test, component or equipment are placed under test mode, only take out detection part after lasting setting time
Or whether equipment fails, to pilot process there is no data acquisition is carried out, the lifetime test data of acquisition is coarse, and the present invention exists
In whole process, whole process acquisition data can obtain more careful and accurate lifetime data.In accelerated life test,
If Testing factors are unable to transition, if temperature decline is unable to transition, then a temperature fall time is needed, in existing accelerated life test
In, the data of temperature-fall period are not acquired, waste data resource, the present invention will be acquired using conversion function in temperature-fall period
Data be converted to data relevant to the lifetime, and be associated with time shaft, so that partial data effective use be got up.
Preferably, lifetime feedback includes component failure sign and failure symptom, component is established by factor data
Fail sign model and failure symptom model.Failure sign model finger produces defective mode, but still can complete it
Data Representation when function carries out failure symptom alarm at this time, assigns maintenance staff's site inspection confirmation;Fail symptom finger
It breaks down, is unable to complete Data Representation when its function, carries out fault alarm at this time, and call together and repair automatically.
Preferably, the aging factor of the component, environmental factor and lifetime feedback data include sensor sample
Data and artificial detection set data.If there is the aging factor required manual intervention or environmental factor, the part in component
The value of aging factor or environmental factor can partially be acquired by sensor(Such as lubricating oil liquid level), but sense in some cases
Device is difficult to acquire(As whether lubricating grease is sufficient), then need manually to set(It is sufficient that lubricating grease is set such as after adding lubricating grease).
Preferably, the sensor sample data of the aging factor, environmental factor and lifetime feedback data, by peace
The sensor with internet of things functional on component is acquired and is uploaded.
Preferably, the artificial detection of the aging factor, environmental factor and lifetime feedback data sets data, by
Setting is uploaded when artificial detection, and is remained unchanged during artificial detection twice or changed by setting function;It is described to press setting letter
Number variation aging factor and environmental factor artificial detection set data, after artificial detection, calculate detected value with by
The difference that function calculates resulting value according to last time detected value is set, the difference is uploaded as lifetime feedback data.It is described to set
Determine function to be provided by empirical function(If component is needed using lubricating grease, the lubricating grease of standard dose is added in maintenance, and is set
Lubricating grease is sufficient simultaneously to be indicated using number 1, then before maintenance next time lubricating grease as environmental factor value by 1 gradually edge set
Determine function decline, which can be by being determined after manually carrying out test of many times;In maintenance next time, artificial judgment
Lubricating grease surplus, and calculate surplus and be compared with according to setting function calculating gained surplus, difference is fed back as the lifetime
Data calculate gained surplus according to setting function as the practical lubricating grease surplus of component is lower than, and difference is more than threshold value, then judging part
Part working condition is bad, consumes lubricating grease excessive velocities).
Preferably, the method for building up of the independent model includes the following steps:B1)For each environmental factor dividing regions
Between;B2)When component dispose for the first time or component environment factor locating for section change when, N1 data generation will collecting later
Enter baseline model, update the independent model of component, wherein N1 is the positive integer set manually.In component by checking and accepting new deployment
When, component is less likely mutation damage, is trusted the most the Performance And Reliability of component at this time, collected data are recognized at this time
To be the data under component health status, by this partial data be used to correct baseline model be it is reliable, after obtaining independent model,
It then no longer modifies to independent model, collected data are used to compare with independent model at this time, and difference is as judgement
Component whether the foundation of working healthily;If detecting, environmental factor section changes, and needs to adjust the ginseng in independent model
Number is used to correct independent model, acquisition number later after detecting the variation of environmental factor section to the preceding N1 sampling of model
According to the foundation as the whether healthy operation of judgement part.The adjustment institute of N1 value whole coefficient according to involved in environmental change
It needs.As recommendation, N1 value is 2 ~ 5, goes to be worth greatly when the coefficient for involving a need to adjustment, involves a need to take when regulation coefficient is few
Small value.
Preferably, in stepb, it is described to compare the service condition of component and the independent model of component, obtain component
The method of lifetime monitoring information be:C1)It is independent that collected component aging factor and environmental factor data are substituted into component
Model, if independent model calculates gained lifetime feedback data and collected lifetime feedback data difference is more than setting threshold
Value, then give a warning and/or the maintenance project of flag member, conversely, then updating the remaining life of component;C2)When collected
When component aging factor, environmental factor and lifetime feedback data meet failure sign model, then gives a warning and/or mark
The maintenance project of component;C3)When collected component aging factor, environmental factor and lifetime feedback data meet failure disease
When shape model, then gives a warning and/or issue to call together automatically and repair.
Preferably, in stepb, by the service condition of component simultaneously with the independent model of component and with class model pair
Than obtaining the lifetime monitoring information of component;The method for building up of the same class model is:CC1)By whole components in monitoring,
Several groups are divided into according to service condition and/or ambient conditions;CC2)Collect the aging factor of each component, environment in every group
N2 data of factor and lifetime feedback reject aging factor or environmental factor data value in gathering and exceed setting range
Data, wherein N2 is the positive integer set manually;CC3)Every group of data are substituted into basic model respectively and are carried out at equalization
Reason, obtains the same class model of every group parts.Single component possesses its independent model, and the basis that independent model is established is baseline mould
The sampled data of type and only single component, due to individual data items deficient in stability and accuracy, thus present invention use will be similar
Under operating condition component grouping, to a large amount of individuals in group carry out statistics and it is regular, remove there are the data sample of failure, only retain
The data for working normally individual, substitute into baseline model after equalization, find out same class model, have with class model compared to baseline model
There is a higher accuracy, and it is established flexibly, the data that can be acquired in a short time after equipment investment based ons is modeled, number
It is short according to collection and modeling period.
Preferably, the service condition by component is simultaneously with the independent model of component and with the method for class model comparison
For:Collected component aging factor and environmental factor data are substituted into component independent model and same class model respectively, if independent
Model is more than given threshold with class model calculating gained lifetime feedback data and collected lifetime feedback data difference,
It then gives a warning and/or the maintenance project of flag member.
Preferably, carrying out step after step c:D)When the aging factor of the same model component being collected into, environment because
The quantity of element and lifetime feedback data is more than modified basis model after setting value, and updates the component lifetime in step
Baseline model.After obtaining a large amount of real data, manual analysis can be carried out, establishes empirical model or theory in new industry
Model, and baseline model is updated accordingly, more accurate reference baseline model is provided for the component of subsequent deployment, further increases life
Life phase monitoring precision.
Preferably, collecting the maintenance field data for being used for modified basis model when component carries out maintenance operation, upload
Aging factor, environmental factor and the lifetime feedback data of component.At the scene when maintenance, the voice data at maintenance scene is collected
And/or the lifetime feedback data of image data, voice data and/or image data as component, to establish new basic mould
Type provides data and supports.When field data accumulation is enough, the lifetime of component can also be monitored and sound and/or figure is added
As being used as lifetime feedback data, for assisting the accuracy of verifying lifetime monitoring result, monitoring precision is improved.
Substantial effect of the invention is:With the service life, pressure test is combined, and obtains quantity foot, the number of valid data sample
Comprehensively according to acquisition, for the lifetime than pair model foundation it is more careful, it is thus possible to lifetime of component carry out more comprehensively and
More accurately monitoring.
Detailed description of the invention
Fig. 1 is lifetime monitoring method flow diagram.
Specific embodiment
Below by specific embodiment, and in conjunction with attached drawing, a specific embodiment of the invention is further described in detail.
As shown in Figure 1, lifetime monitoring method flow diagram, includes the following steps:A)Establish the baseline of component lifetime
Model;B)According to the service condition of each component, baseline model is adjusted, the independent model of each component is obtained;C)By component
Service condition and the independent model of component compare, and obtain the lifetime monitoring information of component.
As explanation:Using uitraviolet intensity and access times as the independent variable of baseline model, the two is respectively provided with one and is
Number, can determine one group of calibration coefficients according to accelerated life test result data.In practical applications, ambient ultraviolet line changes model
It is with limit, thus calibration coefficients can meet the rule that the variation of ambient ultraviolet line changes the lifetime;But accelerated life test is not
The possible whole ambient conditions of exhaustion, when arrangements of components is when uitraviolet intensity is much higher than accelerated life test state, such as in space
Middle use, intensity when uitraviolet intensity is used much higher than ground, coefficient of correspondence are also no longer complies with ultraviolet light variation to life
The rule of phase variation, thus needs to be turned up the coefficient of ultraviolet light, improves ultraviolet light to the specific gravity of lifetime aging, after coefficient adjustment
The independent model that is monitored as component of baseline model.
The method for building up of baseline model includes the following steps:A1)Aging factor, environmental factor and the lifetime of definition component
Feedback;A2)Using engineering theory model and/or industry experience lifetime model as basic model;A3)To each aging factor into
Row accelerated life test and/or simulation test obtain single aging factor test data, substitute into basic model and obtain single aging
The lifetime model of factor;A4)Accelerated life test is carried out to the combination of each pair of single environment factor and single aging factor
And/or simulation test, the test data of single environment factor and the combination of single aging factor is obtained, basic model is substituted into and obtains list
One environmental factor and the lifetime model of single aging factor combination;A5)To multiple environmental factors and multiple aging factors into
Row accelerated life test and/or simulation test obtain multifactor data;A6)For the lifetime model and list of single aging factor
One environmental factor and the lifetime model of single aging factor combination are manually provided with the functional relation composition of initial weight value
Initial baseline model is established the model of fit of weighted value and aging factor and environmental factor according to multifactor data, will be fitted
Model and lifetime baseline model of the initial baseline model simultaneous as component.
As explanation:If A, B and C is aging factor, single aging factor is established respectively in step A3 and step A4
Lifetime model W (A), W (B) and W (C) and single environment factor and the lifetime model W of single aging factor combination
(AD), W (BE) and W (CF), the functional relation composition initial baseline model for being manually provided with initial weight value is Wbase=aW
(A)+bW (B)+cW (C)+dW (AD)+eW (BE)+fW (CF), wherein a, b, c, d, e and f are weighted value, are obtained in step A5
Obtain the data group of largely [W, A, B, C, D, E, F] format, i.e., multifactor data, by every group [W, A, B, C, D, E, F] substitution Wbase=
AW (A)+bW (B)+cW (C)+dW (AD)+eW (BE)+fW (CF), can obtain aging factor and environmental factor data [A, B,
C, D, E, F] with the mapping relations of model coefficient [a, b, c, d, e, f], which is fitted, weighted value is obtained and declines
The model of fit Q ([A, B, C, D, E, F], [a, b, c, d, e, f]) of old factor and environmental factor, in application, declining acquisition
Old factor and environmental factor data [A, B, C, D, E, F] substitute into model of fit Q ([A, B, C, D, E, F], [a, b, c, d, e, f])
It obtains corresponding [a, b, c, d, e, f], [a, b, c, d, e, the f] that obtains is substituted into Wbase=aW(A)+bW(B)+cW(C)+dW
(AD)+eW (BE)+fW (CF) can find out baseline model solution value Wbase, lifetime feedback data W that actual acquisition is arrived with
WbaseComparison is judged as if difference is more than threshold value and does not meet baseline model, conversely, judgement meets baseline model.In this example
Middle WbaseModel in W (A), W (B), W (C), W (AD), W (BE) be set as the calculated relationship being added with W (CF), which closes
System can also be other types, and such as multiplication relationship, multiplication is combined with addition or exponential relationship, the setting of the calculated relationship only influence
The calculation amount of model foundation and required data volume do not necessarily affect model accuracy, can will add up relationship and close as initial
System establishes model, can adjust calculated relationship according to data and experience accumulation in actual application.
Accelerated life test is carried out to aging factor, the collecting method for obtaining single aging factor test data is:
If aging factor from ambient condition to trystate be transient process, when aging factor is in trystate, at interval of when
Between t1 acquisition aging factor data and lifetime feedback data and with time data correlation, conversely, then in aging factor from environment
During state change to trystate, aging factor data and lifetime feedback data are acquired at interval of time t2 and through converting
Function and time data correlation acquire aging factor data at interval of time t1 then when aging factor is in trystate
And lifetime feedback coefficient and with time data correlation;Accelerated aging is carried out to the combination of each pair of environmental factor and aging factor
Test, the method for obtaining the test data of single environment factor and the combination of single aging factor are:If aging factor and environment
Factor is transient process from ambient condition to trystate, then when aging factor and environmental factor are in trystate,
Every time t1 acquire aging factor data, environmental factor data and lifetime feedback data and with time data correlation, instead
It, then acquire from ambient condition to during being in trystate every time t2 in aging factor and environmental factor
Aging factor data, environmental factor data and lifetime feedback data and through conversion function and time data correlation, then exist
When aging factor and environmental factor are in trystate, aging factor data, environmental factor number are acquired at interval of time t1
Accordingly and lifetime feedback coefficient and with time data correlation, wherein interval time t1 and t2 is set manually.
Lifetime feedback includes component failure sign and failure symptom, and the failure sign mould of component is established by factor data
Type and failure symptom model.Failure sign model finger produces defective mode or has notable difference compared with normal condition,
But still Data Representation when can complete its function, failure symptom alarm is carried out at this time, assigns maintenance staff's site inspection true
Recognize;The symptom finger that fails breaks down, and is unable to complete Data Representation when its function, carries out fault alarm at this time, and automatic
It calls together and repairs.
Aging factor, environmental factor and the lifetime feedback data of component include sensor sample data and artificial detection
Set data.The sensor sample data of aging factor, environmental factor and lifetime feedback data, by being mounted on component
Sensor with internet of things functional is acquired and is uploaded.If component exist the aging factor that requires manual intervention or environment because
Element, then the value of the part aging factor or environmental factor, can partially be acquired, such as lubricating oil liquid level, but have by sensor
A little situation lower sensors are difficult to acquire, such as whether lubricating grease is sufficient, due to lubricating grease paste sensor inspection inconvenient to use
It surveys, so carrying out data setting after needing to manually check or add, such as setting lubricating grease is sufficient after adding lubricating grease.
The artificial detection of aging factor, environmental factor and lifetime feedback data sets data, on when artificial detection
Setting is passed, and remains unchanged during artificial detection twice or changes by setting function;By the aging factor of setting function variation
And the artificial detection of environmental factor sets data, after artificial detection, calculates detected value and by setting function according to last time
Detected value calculates the difference of resulting value, and difference is uploaded as lifetime feedback data.Wherein, setting function is given by empirical function
Out.Concrete example such as component is needed using lubricating grease, and the lubricating grease of standard dose is added in maintenance, and sets lubricating grease abundance
And using number 1 indicate, then before maintenance next time lubricating grease as environmental factor value by 1 gradually along set function under
Drop, which can be by being determined after manually carrying out test of many times;In maintenance next time, more than artificial judgment lubricating grease
Amount, and calculate surplus and be compared with according to setting function calculating gained surplus, difference is as lifetime feedback data, such as portion
The practical lubricating grease surplus of part calculates gained surplus lower than according to setting function, and difference is more than threshold value, then judgement part work shape
State is bad, consumes lubricating grease excessive velocities.
The method for building up of independent model includes the following steps:B1)For each environmental factor demarcation interval;B2)As component head
When section locating for secondary deployment or component environment factor changes, the N1 data collected later are substituted into baseline model, more
The independent model of new component, wherein N1 is the positive integer set manually.When component is by checking and accepting new deployment, component less may be used
It can be mutated damage, the Performance And Reliability of component is trusted the most at this time, collected data are considered component health at this time
Data under state, being used to correct baseline model for this partial data is reliably, after obtaining independent model, then no longer to independence
Model is modified, and collected data are used to compare with independent model at this time, and whether difference is healthy as judgement part
The foundation of work;If detecting, environmental factor section changes, and needs to adjust the parameter in independent model, is detecting ring
The preceding N1 sampling of model is used to correct independent model after the variation of border factor section, acquisition data later are as judgement part
Whether health operation foundation.N1 value should be needed for the adjustment of whole coefficient according to involved in environmental change.As recommendation, N1
Value is 2 ~ 5, goes to be worth greatly when the coefficient for involving a need to adjustment, involves a need to get the small value when regulation coefficient is few.
In stepb, the service condition of component and the independent model of component are compared, obtains the lifetime monitoring letter of component
The method of breath is:C1)Collected component aging factor and environmental factor data are substituted into component independent model, if independent model
Calculate gained the lifetime feedback data and collected lifetime feedback data difference be more than given threshold, then give a warning and/
Or the maintenance project of flag member, conversely, then updating the remaining life of component;C2)When collected component aging factor, environment
When factor and lifetime feedback data meet failure sign model, then give a warning and/or the maintenance project of flag member;
C3)When collected component aging factor, environmental factor and lifetime feedback data meet failure symptom model, then issue
Warning and/or automatic sending are called together and are repaired.
Step is carried out after stepb:C)When aging factor, environmental factor and the life of the same model component being collected into
The quantity of phase feedback data is more than after setting value, and modified basis model forms new experience lifetime model, and in step more
New component lifetime baseline model.After obtaining a large amount of real data, manual analysis can be carried out, the warp in new industry is established
Model is tested, and updates baseline model accordingly, more accurate reference baseline model is provided for the component of subsequent deployment, further increases
Lifetime monitoring precision.
When component carries out maintenance operation, collect maintenance field data, upload the aging factor of component, environmental factor and
Lifetime feedback data.At the scene when maintenance, the voice data and/or image data at maintenance scene are collected, when field data product
When tiring out enough, addition sound and/or image will be monitored for the lifetime of component as monitoring foundation, data basis is provided.
Embodiment one:It is monitored, is included the following steps using lifetime of the method for the present invention to button switch:
F1)Function, failure mode, failure symptom and physical attribute, the button that analysis button switchs are with automatic resilience function
Trigger device, when a button is pressed, pole plate connect, issue electric signal, then removing external force button should spring back rapidly, pole
Plate is restored to off-state.Button includes two pole plates and is arranged in the rubber layer that two pole plates directly play rebound effect, when
When rubber layer aging lost its bounce, button pole plate springs back slack-off after removing external force, causes pole plate turn-on time to extend, reduction presses
Button switch performance causes pole plate to always remain on, i.e. button useless, needs replacing when rubber layer continuation aging;In conjunction with physics
The main reason for switch button failure is the fatigue aging and deterioration by oxidation of rubber layer known to gaining knowledge, and is influenced in practical applications
The main cause of rubber layer fatigue aging is access times, and the main cause for influencing rubber layer deterioration by oxidation is temperature and uitraviolet intensity,(Oxygen
Gas concentration is relatively stable, thus its influence be it is more constant, do not have monitoring), using human factor access times as declining
Old factor, using temperature and uitraviolet intensity as environmental factor, establishing button switch setting theoretical model is Wk-b=60000-a*
n-b*T*ec*x, wherein n is access times, and x is uitraviolet intensity index, and T is temperature, and a, b, c are model coefficient;
F2)Accelerated life test is carried out to button switch, only carries out the button switch lifetime results under the influence of access times, temperature
It is remained unchanged with uitraviolet intensity, obtains enough [Wk-b, n] format data group, by [Wk-b, n] format data group substitute into
Wk-b=60000-a*n-b*T*ec*xThe value for calculating a, b, c identical with data group quantity, after equalization, obtains button switch
To the aging model W (n) of the single aging factor of access times;
F3)Accelerated life test is carried out to button switch, carries out being higher than ambient ultraviolet line intensity, being lower than ambient ultraviolet line respectively
Intensity, temperature are higher than environment temperature or temperature lower than in the case where environment temperature, and access times are to button switch lifetime results
Influence, obtain [Wk-b, n, x] and [Wk-b, n, T] format data group, similar step F2 obtains button switch about using
The aging model W (n, T) of the aging model W (n, x) and access times and temperature of number and uitraviolet intensity;
F4)Accelerated life test is carried out to button switch, carries out being higher than ambient ultraviolet line intensity or lower than ambient ultraviolet line intensity
And in the case that temperature is lower than environment temperature higher than environment temperature or temperature, shadow of the access times to button switch lifetime results
It rings, obtains [Wk-b, n, x, T] format data group;
F5)Establish initial baseline model:Wbase=qW (n)+rW (n, x)+sW (n, T), wherein q, r, s are initial coefficients, will
[Wk-b, n, x, T] format data substitute into initial baseline model can in the hope of a class value [q, r, s], establish [n, x, T] with [q, r,
S] mapping relations, and be fitted to function F ([n, x, T], [q, r, s]), Modulus Model is associated with by function F as Modulus Model
Baseline model after initial baseline model, as button switch;
F6)High, normal, basic third gear is divided into uitraviolet intensity, temperature is equally divided into high temperature, room temperature and low temperature third gear;
F7)In practical applications, after button switch deployment, one natural gift of interval 3 times to button switch progress data acquisition, acquisition number
According to including uitraviolet intensity, access times and environment temperature, [n, x, T] data are surveyed, the function F of Modulus Model is passed through
([n, x, T], [q, r, s]), is calculated [q, r, s], and [q, r, s] is substituted into baseline model Wbase=qW(n)+rW(n,x)+sW
(n, T) obtains the independent model W of tested button switchd;
F8)In subsequent monitoring, acquisition data include uitraviolet intensity, access times and environment temperature, surveyed [n, x,
T] data, substitute into independent model WdMiddle calculating, calculated result is as button switch remaining life result;
F9)When detecting that uitraviolet intensity or temperature shelves do not change, it is re-execute the steps F7, then circulation executes step
Rapid F8-F9, until the completing button switch lifetime monitors task;
F10)The total data of all monitored button switches is collected as effective sample data, it is new by manually establishing
Industry experience model then executes step F1-F10 for the subsequent button switch to come into operation.
As extension embodiment of the invention, in stepb, by the service condition of component independent mould with component simultaneously
Type and with class model compare, obtain the lifetime monitoring information of component;Method for building up with class model is:CC1)It will be in monitoring
Whole components are divided into several groups according to service condition and/or ambient conditions;CC2)Collect the aging of each component in every group
N2 data of factor, environmental factor and lifetime feedback are rejected aging factor or environmental factor data value in gathering and are exceeded
The data of setting range, wherein N2 is the positive integer set manually;CC3)Every group of data are substituted into basic model respectively to go forward side by side
Row equalization processing, obtains the same class model of every group parts.Single component possesses its independent model, the base that independent model is established
Plinth is baseline model and the only sampled data of single component, due to individual data items deficient in stability and accuracy, thus the present invention
Using by under similar operating condition component be grouped, to a large amount of individuals in group carry out statistics and it is regular, removal there are the data of failure
Sample only retains the data for working normally individual, substitutes into baseline model after equalization, same class model is found out, compared with class model
There is higher accuracy in baseline model, and it is established flexibly, the data that can be acquired in a short time after equipment investment are base
Plinth is modeled, and data collection and modeling period are short.
It is with the independent model of component and with the method for class model comparison simultaneously by the service condition of component:It will be collected
Component aging factor and environmental factor data substitute into component independent model and same class model respectively, if independent model or same class model
Calculate gained the lifetime feedback data and collected lifetime feedback data difference be more than given threshold, then give a warning and/
Or the maintenance project of flag member, conversely, then updating the remaining life of component.
Embodiment two:Using with the lifetime monitoring method with class model, the monitoring of rolling bearing remaining life is carried out
Method includes the following steps:H1)Analysis obtain rolling bearing aging factor be access times n, environmental factor include load N,
Impact load N ', dust D and lubricating status E, wherein load N is included in environmental factor, class as the working environment of rolling bearing
It is similar to embodiment one, establishes rolling bearing about access times n, load N, impact load N ', dust D and lubricating status E
Baseline model WbaseWith independent model Wd, details are not described herein;H2)Setting time length is reached to sufficient amount of come into operation
Rolling bearing carry out data collection, be that underloading and heavily loaded two shelves are other by load partition, come into operation according to rolling bearing
Actual conditions, according to load shelves not, whether there is or not impact load, whether there is or not whether dust and lubricating status good, as differentiation, by work
Make state rolling bearing all the same and be divided into same group, as rolling bearings whole in first group be underloading, without impact load,
No dust and lubrication is good;H3)Collect the rolling bearing access times n with group, load N, impact load N ', dust D and profit
Sliding state E and lifetime feedback data, there are the data of faulty bearing for rejecting, only retain the data for working normally individual,
These data are substituted into basic model, obtain the coefficient [k1, k2 ... ks] of different rolling bearings(Ki is baseline model coefficient, i
∈ [1, s], s are baseline model number of parameters)It, will with the mapping relations of aging factor and environmental factor data [n, N, N ', D, E]
After the coefficient [k1, k2 ... kn] of different rolling bearings and [n, N, N ', D, E] data are to equalization processing is carried out, then it is fitted to and is
Number function:
F'([n, N, N ', D, E], [k1, k2 ... kn]), after being then associated with coefficient function F ' with baseline model, as same group of mould
Type Wg;H4)In practical applications, rolling bearing access times n, load N, impact load N ', dust D and lubricating status are acquired
E substitutes into independent model WdSolve and obtains independent model as a result, then substituting into service condition divides identical same group model Wg
It is solved, is obtained with group model as a result, the two takes smaller value as the lifetime monitoring result of tested rolling bearing, remaining step
Suddenly it is similar to embodiment one, can refer to the implementation of embodiment one.
Above-mentioned embodiment is only a preferred solution of the present invention, not the present invention is made in any form
Limitation, there are also other variations and modifications on the premise of not exceeding the technical scheme recorded in the claims.
Claims (12)
1. a kind of component lifetime monitoring method, which is characterized in that
Include the following steps:
A)Establish the baseline model of component lifetime;
B)According to the service condition of each component, baseline model is adjusted, the independent model of each component is obtained;
C)The service condition of component and the independent model of component are compared, the lifetime monitoring information of component is obtained.
2. a kind of component lifetime monitoring method according to claim 1, which is characterized in that
The method for building up of the baseline model includes the following steps:
A1)Aging factor, environmental factor and the lifetime feedback of definition component;
A2)Using engineering theory model and/or industry experience lifetime model as basic model;
A3)Accelerated life test and/or simulation test are carried out to each aging factor, obtain single aging factor test data,
Substitute into the lifetime model that basic model obtains single aging factor;
A4)Accelerated life test and/or simulation test are carried out to the combination of each pair of single environment factor and single aging factor,
The test data for obtaining single environment factor and the combination of single aging factor substitutes into basic model and obtains single environment factor and list
The lifetime model of one aging factor combination;
A5)Accelerated life test and/or simulation test are carried out to multiple environmental factors and multiple aging factors, obtained multifactor
Data;
A6)The lifetime model combined for the lifetime model and single environment factor of single aging factor and single aging factor
It is manually provided with the functional relation composition initial baseline model of initial weight value, weighted value is established according to multifactor data and declines
The model of fit of old factor and environmental factor, using the model of fit and initial baseline model simultaneous as the lifetime of component
Baseline model.
3. a kind of component lifetime monitoring method according to claim 2, which is characterized in that
Accelerated life test is carried out to aging factor, the collecting method for obtaining single aging factor test data is:If declining
Old factor is transient process from ambient condition to trystate, then when aging factor is in trystate, at interval of time t1
Acquire aging factor data and lifetime feedback data and with time data correlation, conversely, then in aging factor from ambient condition
It changes to during trystate, acquires aging factor data and lifetime feedback data at interval of time t2 and through converting function
With time data correlation, then when aging factor is in trystate, at interval of time t1 acquire aging factor data and
Lifetime feedback coefficient and with time data correlation;
Accelerated life test is carried out to the combination of each pair of environmental factor and aging factor, obtain single environment factor and single is declined
The method of the test data of old factor combination is:If aging factor and environmental factor are wink from ambient condition to trystate
Change process acquires aging factor data, environment every time t1 then when aging factor and environmental factor are in trystate
Factor data and lifetime feedback data and with time data correlation, conversely, then in aging factor and environmental factor from ring
Border state acquires aging factor data, environmental factor data and life every time t2 to during being in trystate
Life phase feedback data and through conversion function and time data correlation, is then in test shape in aging factor and environmental factor
When state, close at interval of time t1 acquisition aging factor data, environmental factor data and lifetime feedback coefficient and with time data
Connection, wherein interval time t1 and t2 is set manually.
4. a kind of component lifetime monitoring method according to claim 2 or 3, which is characterized in that
Lifetime feedback include component failure sign and failure symptom, by factor data establish component failure sign model and
Fail symptom model.
5. a kind of component lifetime monitoring method according to claim 2 or 3, which is characterized in that
Aging factor, environmental factor and the lifetime feedback data of the component include sensor sample data and artificial detection
Set data.
6. a kind of component lifetime monitoring method according to claim 5, which is characterized in that
The artificial detection of the aging factor, environmental factor and lifetime feedback data sets data, on when artificial detection
Setting is passed, and remains unchanged during artificial detection twice or changes by setting function;
It is described to set data by the aging factor of setting function variation and the artificial detection of environmental factor, in artificial detection
Afterwards, it calculates detected value and calculates the difference of resulting value according to last time detected value by setting function, the difference is anti-as the lifetime
Data are presented to upload.
7. a kind of component lifetime monitoring method according to claim 2 or 3, which is characterized in that
The method for building up of the independent model includes the following steps:
B1)For each environmental factor demarcation interval;
B2)When component dispose for the first time or component environment factor locating for section change when, N1 data will collecting later
Substitute into baseline model, adjust baseline model weight parameter value make collected N1 data substitute into baseline model calculated result and
Testing result difference is respectively less than given threshold, and independent model of the baseline model as component after adjustment, wherein N1 is set manually.
8. a kind of component lifetime monitoring method according to claim 2 or 3, which is characterized in that
In stepb, described to compare the service condition of component and the independent model of component, obtain the lifetime monitoring letter of component
The method of breath is:
C1)Collected component aging factor and environmental factor data are substituted into component independent model, if independent model calculates institute
Obtaining lifetime feedback data and collected lifetime feedback data difference is more than given threshold, then gives a warning and/or mark
The maintenance project of component;
C2)When collected component aging factor, environmental factor and lifetime feedback data meet failure sign model, then
It gives a warning and/or the maintenance project of flag member;
C3)When collected component aging factor, environmental factor and lifetime feedback data meet failure symptom model, then
It gives a warning and/or issues to call together automatically and repair.
9. a kind of component lifetime monitoring method according to claim 2 or 3, which is characterized in that
In stepb, the service condition of component is compared with the independent model of component and with class model simultaneously, obtains the life of component
Life phase monitoring information;
The method for building up of the same class model is:
CC1)By whole components in monitoring, several groups are divided into according to service condition and/or ambient conditions;
CC2)N2 data for collecting the aging factor of each component in every group, environmental factor and lifetime feedback, reject collection
Aging factor or environmental factor data value exceed the data of setting range in closing, and wherein N2 is set manually;
CC3)The weight parameter value of adjustment baseline model make every group collected by N2 data substitute into baseline model calculated result and
Testing result difference is respectively less than given threshold, same class model of the baseline model as every group parts after adjustment.
10. a kind of component lifetime monitoring method according to claim 9, which is characterized in that
The service condition by component is with the independent model of component and with the method for class model comparison simultaneously:
Collected component aging factor and environmental factor data are substituted into component independent model and same class model respectively, if independent
Model is more than given threshold with class model calculating gained lifetime feedback data and collected lifetime feedback data difference,
It then gives a warning and/or the maintenance project of flag member.
11. a kind of component lifetime monitoring method according to claim 2 or 3, which is characterized in that
Step is carried out after step c:D)It is anti-when the aging factor for the same model component being collected into, environmental factor and lifetime
The quantity for presenting data is more than modified basis model after setting value, and updates component lifetime baseline model in step.
12. a kind of component lifetime monitoring method according to claim 11, which is characterized in that
When component carries out maintenance operation, collect the maintenance field data for being used for modified basis model, upload the aging of component because
Element, environmental factor and lifetime feedback data.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810621885.5A CN108921305B (en) | 2018-06-15 | 2018-06-15 | Component life period monitoring method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810621885.5A CN108921305B (en) | 2018-06-15 | 2018-06-15 | Component life period monitoring method |
Publications (2)
Publication Number | Publication Date |
---|---|
CN108921305A true CN108921305A (en) | 2018-11-30 |
CN108921305B CN108921305B (en) | 2021-07-02 |
Family
ID=64420551
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810621885.5A Active CN108921305B (en) | 2018-06-15 | 2018-06-15 | Component life period monitoring method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108921305B (en) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111311883A (en) * | 2020-05-12 | 2020-06-19 | 成都千嘉科技有限公司 | Gas alarm replacement early warning method and device and electronic equipment |
CN111967778A (en) * | 2020-08-19 | 2020-11-20 | 杭州铂钰信息科技有限公司 | Data security detection method and system based on security baseline model |
CN113551716A (en) * | 2021-09-22 | 2021-10-26 | 江西省公路工程检测中心 | Stress-strain detection method for mechanical property safety monitoring cloud system |
CN116953417A (en) * | 2023-09-20 | 2023-10-27 | 国网湖北省电力有限公司经济技术研究院 | Power transformer service life assessment device and method |
Citations (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2006146763A (en) * | 2004-11-24 | 2006-06-08 | Shimizu Corp | Care-need certificated person number prediction system |
US20090228230A1 (en) * | 2008-03-06 | 2009-09-10 | General Electric Company | System and method for real-time detection of gas turbine or aircraft engine blade problems |
KR101091754B1 (en) * | 2011-06-23 | 2011-12-08 | (주)티앤테크 | Equipment maintanance management system and method thereof |
CN102508957A (en) * | 2011-10-28 | 2012-06-20 | 北京航空航天大学 | Assessment method of electronic complete machine acceleration service life |
CN102567560A (en) * | 2010-12-27 | 2012-07-11 | 北京圣涛平试验工程技术研究院有限责任公司 | Method and system for estimating service life of MOS (Metal Oxide Semiconductor) device |
CN103778280A (en) * | 2014-01-08 | 2014-05-07 | 上海交通大学 | Method for remain useful life prognostic of lithium ion battery with model active updating strategy |
CN103926536A (en) * | 2014-03-07 | 2014-07-16 | 南京航空航天大学 | Method for predicting residual service life of lithium ion battery on basis of DST and BMC technologies |
CN105468850A (en) * | 2015-11-26 | 2016-04-06 | 电子科技大学 | Multi-residual error regression prediction algorithm based electronic product degradation trend prediction method |
CN105653851A (en) * | 2015-12-27 | 2016-06-08 | 北京化工大学 | Residual life prediction method of antifriction bearing on the basis of staged physical model and particle filter |
-
2018
- 2018-06-15 CN CN201810621885.5A patent/CN108921305B/en active Active
Patent Citations (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2006146763A (en) * | 2004-11-24 | 2006-06-08 | Shimizu Corp | Care-need certificated person number prediction system |
US20090228230A1 (en) * | 2008-03-06 | 2009-09-10 | General Electric Company | System and method for real-time detection of gas turbine or aircraft engine blade problems |
CN102567560A (en) * | 2010-12-27 | 2012-07-11 | 北京圣涛平试验工程技术研究院有限责任公司 | Method and system for estimating service life of MOS (Metal Oxide Semiconductor) device |
KR101091754B1 (en) * | 2011-06-23 | 2011-12-08 | (주)티앤테크 | Equipment maintanance management system and method thereof |
CN102508957A (en) * | 2011-10-28 | 2012-06-20 | 北京航空航天大学 | Assessment method of electronic complete machine acceleration service life |
CN103778280A (en) * | 2014-01-08 | 2014-05-07 | 上海交通大学 | Method for remain useful life prognostic of lithium ion battery with model active updating strategy |
CN103926536A (en) * | 2014-03-07 | 2014-07-16 | 南京航空航天大学 | Method for predicting residual service life of lithium ion battery on basis of DST and BMC technologies |
CN105468850A (en) * | 2015-11-26 | 2016-04-06 | 电子科技大学 | Multi-residual error regression prediction algorithm based electronic product degradation trend prediction method |
CN105653851A (en) * | 2015-12-27 | 2016-06-08 | 北京化工大学 | Residual life prediction method of antifriction bearing on the basis of staged physical model and particle filter |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111311883A (en) * | 2020-05-12 | 2020-06-19 | 成都千嘉科技有限公司 | Gas alarm replacement early warning method and device and electronic equipment |
WO2021227409A1 (en) * | 2020-05-12 | 2021-11-18 | Chengdu Qianjia Technology Co. Ltd | Gas alarm replacement warning method and device, and electronic apparatus including the device |
CN111967778A (en) * | 2020-08-19 | 2020-11-20 | 杭州铂钰信息科技有限公司 | Data security detection method and system based on security baseline model |
CN113551716A (en) * | 2021-09-22 | 2021-10-26 | 江西省公路工程检测中心 | Stress-strain detection method for mechanical property safety monitoring cloud system |
CN116953417A (en) * | 2023-09-20 | 2023-10-27 | 国网湖北省电力有限公司经济技术研究院 | Power transformer service life assessment device and method |
CN116953417B (en) * | 2023-09-20 | 2023-12-15 | 国网湖北省电力有限公司经济技术研究院 | Power transformer service life assessment device and method |
Also Published As
Publication number | Publication date |
---|---|
CN108921305B (en) | 2021-07-02 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108921305A (en) | A kind of component lifetime monitoring method | |
CN111047082B (en) | Early warning method and device of equipment, storage medium and electronic device | |
US11067973B2 (en) | Data collection system, abnormality detection method, and gateway device | |
CN103903408B (en) | Method for early warning and system are investigated in equipment fault | |
JP5388060B2 (en) | Elevator parts improvement planning system and parts improvement planning method thereof | |
JP4832609B1 (en) | Abnormal sign diagnosis device and abnormality sign diagnosis method | |
US20160342392A1 (en) | Monitoring system and monitoring method | |
CN109416531A (en) | The different degree decision maker of abnormal data and the different degree determination method of abnormal data | |
CN109948860A (en) | A kind of mechanical system method for predicting residual useful life and system | |
CN109071162B (en) | Elevator remote maintenance support system and elevator remote maintenance support method | |
KR101214765B1 (en) | System and method for monitoring electrical power equipment | |
CN117078017A (en) | Intelligent decision analysis system for monitoring power grid equipment | |
KR101843365B1 (en) | Integrated Diagnostic System and Database based on rules and cases | |
CN106093838A (en) | Gateway electric energy meter remote online verification method and system | |
CN105243180B (en) | Electromechanical equipment fault early warning method and system | |
CN110865924B (en) | Health degree diagnosis method and health diagnosis framework for internal server of power information system | |
JP6880560B2 (en) | Failure prediction device, failure prediction method and failure prediction program | |
CN110703743A (en) | Equipment failure prediction and detection system and method | |
CN107656156A (en) | A kind of equipment fault diagnosis and running status appraisal procedure and system based on cloud platform | |
CN110287347A (en) | Using the method for big data detection electric power computer room failure | |
JP2009086896A (en) | Failure prediction system and failure prediction method for computer | |
CN118070203B (en) | Equipment portrait construction method and system based on big data | |
CN116861503A (en) | Method for constructing digital twin model of power transformer based on big data | |
JP2009155043A (en) | Elevator component improvement schedule system and elevator component improvement schedule method | |
US20240018940A1 (en) | Method and system for performing condition monitoring of wind turbine components |
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 |