CN109299828A - Method for predicting fault-free running time of urban rail equipment based on survival analysis - Google Patents
Method for predicting fault-free running time of urban rail equipment based on survival analysis Download PDFInfo
- Publication number
- CN109299828A CN109299828A CN201811162487.8A CN201811162487A CN109299828A CN 109299828 A CN109299828 A CN 109299828A CN 201811162487 A CN201811162487 A CN 201811162487A CN 109299828 A CN109299828 A CN 109299828A
- Authority
- CN
- China
- Prior art keywords
- failure
- model
- fault
- operation time
- function
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 31
- 238000004458 analytical method Methods 0.000 title claims abstract description 28
- 230000004083 survival effect Effects 0.000 title claims abstract description 27
- 230000001133 acceleration Effects 0.000 claims description 8
- 238000013139 quantization Methods 0.000 claims description 8
- 230000001186 cumulative effect Effects 0.000 claims description 6
- 230000008569 process Effects 0.000 claims description 4
- 230000017105 transposition Effects 0.000 claims description 4
- 238000007476 Maximum Likelihood Methods 0.000 claims description 3
- 238000003556 assay Methods 0.000 claims description 3
- 238000006243 chemical reaction Methods 0.000 claims description 3
- 238000004140 cleaning Methods 0.000 claims description 3
- 238000005315 distribution function Methods 0.000 claims description 3
- 230000006698 induction Effects 0.000 claims description 2
- 238000012423 maintenance Methods 0.000 abstract description 29
- 230000000737 periodic effect Effects 0.000 abstract description 5
- 230000007547 defect Effects 0.000 abstract 1
- 238000007726 management method Methods 0.000 description 10
- 230000006872 improvement Effects 0.000 description 7
- 230000015556 catabolic process Effects 0.000 description 5
- 238000011156 evaluation Methods 0.000 description 4
- 238000000611 regression analysis Methods 0.000 description 3
- 238000011160 research Methods 0.000 description 3
- 238000009825 accumulation Methods 0.000 description 2
- 230000033228 biological regulation Effects 0.000 description 2
- 230000008439 repair process Effects 0.000 description 2
- 241000208340 Araliaceae Species 0.000 description 1
- 241001269238 Data Species 0.000 description 1
- 235000005035 Panax pseudoginseng ssp. pseudoginseng Nutrition 0.000 description 1
- 235000003140 Panax quinquefolius Nutrition 0.000 description 1
- 230000004888 barrier function Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 230000000052 comparative effect Effects 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 239000000428 dust Substances 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000004134 energy conservation Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 238000013277 forecasting method Methods 0.000 description 1
- 238000009472 formulation Methods 0.000 description 1
- 235000008434 ginseng Nutrition 0.000 description 1
- 238000012417 linear regression Methods 0.000 description 1
- 230000007257 malfunction Effects 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 238000012549 training Methods 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 1
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
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/18—Complex mathematical operations for evaluating statistical data, e.g. average values, frequency distributions, probability functions, regression analysis
-
- 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
- G06Q10/0635—Risk analysis of enterprise or organisation activities
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/40—Business processes related to the transportation industry
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Human Resources & Organizations (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Strategic Management (AREA)
- Theoretical Computer Science (AREA)
- Economics (AREA)
- Data Mining & Analysis (AREA)
- Entrepreneurship & Innovation (AREA)
- Marketing (AREA)
- Operations Research (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Mathematical Physics (AREA)
- Pure & Applied Mathematics (AREA)
- Development Economics (AREA)
- Quality & Reliability (AREA)
- Mathematical Optimization (AREA)
- Computational Mathematics (AREA)
- Game Theory and Decision Science (AREA)
- Mathematical Analysis (AREA)
- Databases & Information Systems (AREA)
- Evolutionary Biology (AREA)
- Probability & Statistics with Applications (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Algebra (AREA)
- Life Sciences & Earth Sciences (AREA)
- Bioinformatics & Computational Biology (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Educational Administration (AREA)
- General Engineering & Computer Science (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a method for predicting the fault-free running time of urban rail equipment based on survival analysis, which overcomes the defects of two maintenance modes of fault maintenance and periodic maintenance, can be used for analyzing the fault-free running time of the equipment, can also be used for analyzing the probability of faults after the equipment runs for a period of time, provides scientific guidance and direction for the maintenance and management of the equipment, and is a great breakthrough and reform in the technical field of urban rail transit equipment management.
Description
Fields
The present invention relates to urban rail transit equipment administrative skill fields, and in particular to a kind of Transit Equipment fault-free
Runing time prediction technique.
Background technique
Urban track traffic is as the important component in urban public tranlport system, with large conveying quantity, on time, fastly
The advantages such as speed, energy conservation, the approval with passenger to rail traffic, more and more passengers select rail traffic trip, large passenger flow
The common phenomenon of rail traffic is had become, the high-intensitive operation of rail traffic increases the frequency of use of equipment, thus tables of equipment
Failure occurs for member or the frequency of failure also increasingly increases.
For being formed for the City Rail Transit System of network, once some website or position are broken down, it will shadow
Whole even entire transportation network of route is rung, thus carries out transit equipment detection evaluation, guarantees failure-free operation, it is very heavy
It wants.The Transit Equipment Strategies of Maintenance of present is mainly breakdown maintenance and periodic maintenance.Breakdown maintenance can make equipment
Useful life reach maximization, but repair not in time, will lead to bigger loss；Meanwhile the key of periodic maintenance is to tie up
It repairs the period, excessive cycle will lead to maintenance deficiency, influence operational safety, too short to generate excessive maintenance, increase maintenance cost.Cause
And the AFC Maintenance Management System of existing masses can no longer meet the actual demand of AFC operation management.
Since low, periodic maintenance period is not easy to grasp with equipment dependability for breakdown maintenance, device service water is caused
The serious problem of adjustment, breakdown loss, it is difficult to realize Transit Equipment scientific maintenance management.From safety and utilization of resources
The considerations of, the reliability of equipment how is improved, maintenance cost is reduced, realizes equipment no shutdown operation, is urgently to be solved ask
Topic.Based on current status, the thought based on Reliability Maintenance, when proposing that one kind can be based on the equipment failure-free operation of survival analysis
Between prediction technique, obtain the reliability service time of equipment, so that it may provide reference and direction for maintenance management person.)
Summary of the invention
The present invention exactly in the prior art can not the Accurate Prediction equipment smoothly operation time aiming at the problem that, provide one
Kind Transit Equipment non-failure operation time prediction technique overcomes lacking for two kinds of maintenance modes of breakdown maintenance and periodic maintenance
Point can be used to the time of analytical equipment failure-free operation, while break down after can running a period of time with analytical equipment
Probability, to the maintenance management of equipment provide science guidance and direction, be to urban rail transit equipment administrative skill lead
The very big breakthrough and reform in domain.
To achieve the goals above, the technical solution adopted by the present invention is that: based on the Transit Equipment of survival analysis without
Failure operation time forecasting methods, include the following steps:
S1, fault signature analysis of Influential Factors: the characteristic factor that analysis impacts equipment fault；
S2, to the quantization of feature influence factor and assignment explanation；
Failure modes: fault type is divided into mechanical failure and electric information failure by S3；
Fault data statistical disposition: S4 carries out statistic of classification to the fault data of equipment, and carries out data cleansing, reject
Wrong data.
Model of failure distribution selection: S5 selects model of failure distribution by akaike information criterion；
S6 constructs the non-failure operation time prediction model based on acceleration time failure model: the failure of selection acceleration time
Model carries out the building of non-failure operation time prediction model to mechanical failure and electric information failure respectively；
S7, model prediction and analysis: the prediction of equipment non-failure operation time is carried out according to the model of building, to the pre- of model
It surveys performance to be analyzed and evaluated, and the influence according to interpretation of result characteristic factor to equipment non-failure operation time.
As an improvement of the present invention, binaryzation amount is taken to the quantization method of feature influence factor in the step S2
Change method.
As an improvement of the present invention, fault type is divided into mechanical failure and electric information failure in the step S3.
As another improvement of the invention, in the step S4 fault data cleaning, the random failure thing in 24 hours
Part reaches three times by once connection fault statistics；Same failure occurred again less than 24 hours after fault restoration, only statistics one
Secondary failure；When multiple fault modes are caused by same component failure, primary fault is only counted.
As another improvement of the invention, akaike information criterion expression formula described in the step 5 are as follows:
AIC=-2log (L)+2n
In formula, L is the maximum likelihood function value of model, and n is the number of all parameters in model.
It being improved as another kind of the invention, the step S6 constructs time prediction model by establishing risk function,
The conversion process of the risk function is as follows:
For a random time variable T, Cumulative Distribution Function can be indicated are as follows:
In formula, F (t) indicates that duration T is less than the probability of t, and S (t) makes a living store function, and f (t) is probability density function,
Expression formula are as follows:
According to the definition of risk function, the expression formula h (t) of available risk function is
H (t)=f (t)/S (t)=f (t)/(1-F (t))
It is improved as another kind of the invention, the step S6 is based on acceleration time failure model, is added in risk function
Covariant, conditional risk function and survival function embody form are as follows:
H (t | X)=ψ h0(ψt)
S (t | X)=S0(ψt)
In formula, ψ=EXP (- β ' X) indicates one group of covariant vector, β ' expression one group of estimation parameter corresponding with covariant
The transposition of vector, β ' X=β0+β1x1+…+βnxn, h0(·)、S0() is illustrated respectively in when all covariants are zero (X=0)
Baseline risk function and benchmark survival function.
As a further improvement of the present invention, in the step S7 with mean absolute percentage error come assay mould
The predictability of type, the mean absolute percentage error is less than 20%.
Compared with prior art, the present invention caused by the utility model has the advantages that
(1) non-failure operation time for passing through AFT model research equipment, each Correlative Influence Factors can more intuitively be analyzed
Influence to equipment non-failure operation time；
(2) by by the failure modes of equipment, so that the result of prediction is more accurate；
(3) this method is it can be concluded that equipment runs the probability of malfunction after a period of time.Research achievement makes maintenance personal more
Add the changing rule for understanding equipment dependability, reasonably carry out the deposit work of standby redundancy, so that maintenance management work is more
It is scientific and effective.
Detailed description of the invention
Fig. 1 is method operating process schematic diagram of the invention；
Fig. 2 is the risk function and survival function under AFT model.
Specific embodiment
Below with reference to drawings and examples, the present invention is described in detail.
Embodiment 1
A kind of urban rail equipment non-failure operation time prediction technique based on survival analysis, as shown in Figure 1, including following step
It is rapid:
S1, fault signature analysis of Influential Factors: causing the factor of equipment fault very much, respectively from people-machine-ring-pipe four
The feature influence factor of the potential characteristic factor that angle analysis impacts equipment fault, people specifically includes that maintenance personal's skill
Art is horizontal, the accumulation volume of the flow of passengers, the peak hour volume of the flow of passengers, passenger whether violation operation etc.；Environmental characteristic influence factor specifically includes that
Season, temperature, humidity, position, dust, rainfall etc.；Equipment self-condition feature influence factor includes: whether itself design is good
Good, cumulative failure number, accumulation non-failure operation time, last non-failure operation time, maintenance time etc.；Management factors master
It include: rules and regulations, risk management etc..
S2 takes the quantization of binaryzation quantization method and assignment explanation to feature influence factor.
S3, failure modes: being different type by fault type induction and conclusion, and equipment failure rate and equipment material have very high point
System, thus before carrying out non-failure operation time prediction, failure can be divided into mechanical failure and electric information failure.
Fault data statistical disposition: S4 carries out statistic of classification to the fault data of equipment, and carries out data cleansing, reject
Wrong data, in fault data cleaning, random failure event reaches three times by once connection fault statistics in 24 hours；
Same failure occurred again less than 24 hours after fault restoration, only counts primary fault；Multiple fault modes are lost by same device
When effect causes, primary fault is only counted.
S5, model of failure distribution selection: common model of failure distribution has exponential distribution, normal distribution, log series model, prestige
Boolean's distribution etc. is chosen akaike information criterion (Akaike ' sInformationCriterion, AIC) and is selected as distributed model
The criterion selected, AIC are that the concept based on entropy proposes, are a kind of methods of more efficiently tradeoff models fitting data Optimality,
Expression formula are as follows:
AIC=-2log (L)+2n
In formula, L is the maximum likelihood function value of model, and n is the number of all parameters in model, and the value of AIC is smaller, shows
Model is more excellent.
S6, construct the non-failure operation time prediction model based on acceleration time failure model: according to theory of survival analysis,
The distributed model of failure selects acceleration time failure model to carry out failure-free operation to mechanical failure and electric information failure respectively
Time prediction model construction；
Step S6 constructs time prediction model by establishing risk function, and survival analysis is ground by establishing risk function
Study carefully the regularity of distribution of equipment non-failure operation time, the conversion process of risk function is as follows.
For a random time variable T, Cumulative Distribution Function can be indicated are as follows:
In formula, F (t) indicates that duration T is less than the probability of t, and S (t) makes a living store function, when indicating equipment failure-free operation
Between exceed the probability of t, also referred to as survival rate；F (t) is probability density function, expression formula are as follows:
Probability density function gives the instant probability that equipment terminates within the time [t, t+ Δ t].According to risk function
Definition, the expression formula h (t) of available risk function are
H (t)=f (t)/S (t)=f (t)/(1-F (t)) (3)
To explain influence of the characteristic factor to equipment non-failure operation time, AFT model is chosen, is added in risk function
Covariant, covariant are to cause the variation of non-failure operation time by the product with time variable, conditional risk function and
Survival function embodies form are as follows:
H (t | X)=ψ h0(ψt) (4)
S (t | X)=S0(ψt) (5)
In formula, ψ=EXP (- β ' X) indicates one group of covariant vector, β ' expression one group of estimation parameter corresponding with covariant
The transposition of vector, β ' X=β0+β1x1+…+βnxn, h0(·)、S0() is illustrated respectively in when all covariants are zero (X=0)
Baseline risk function and benchmark survival function.
S7, model prediction and analysis: the prediction of equipment non-failure operation time is carried out according to the model of building, to the pre- of model
It surveys performance to be analyzed and evaluated, and the influence according to interpretation of result characteristic factor to equipment non-failure operation time, choose average
Absolute percent error (MAPE) is evaluation index, is evaluated the performance of model, when MAPE value is less than 20%, then it is assumed that
This method is feasible.
Selection automatic ticket checker for rail transportation equipment is research object, takes the event in 1 year of a certain station automatic fare collection machine equipment
Hindering data, the present invention is further illustrated.
Step1, automatic fare collection machine equipment incipient fault analysis of Influential Factors, in terms of people-machine-ring-pipe four, the side of people
It is analyzed from maintenance personal's technical level, cultural quality, passenger flow cumulant, peak hour flow etc. in face；Equipment aspect is set from itself
Meter, cumulative failure number, non-failure operation time etc. analysis；It is analyzed in terms of environment from temperature, humidity, rainfall etc.；
Management aspect is analyzed from rules and regulations, risk management etc..
The quantization of Step2, potential feature influence factor extract not for the potential feature influence factor of above-mentioned analysis
The corresponding variable with factor, and assignment explanation is carried out, such as table 1:
The potential covariant of table 1 and assignment explanation
Step3, equipment fault classification, are divided into mechanical failure and electric information failure for automatic ticket checker equipment fault.With
For mechanical failure.
Step4, fault data statistical disposition, in data handling, in 24 hours random failure event reach three times by
Once connection fault statistics；Same failure occurred again less than 24 hours after fault restoration, only counts primary fault；Multiple events
When barrier mode is caused by same component failure, primary fault is only counted；Using the fault data of the first eight months as the training of model
Collection, rear four months fault datas are as test set.
Step5, fault model distribution selection, according to the fault data of statistics, respectively with Loglogistic, Weibull,
Five kinds of Logistic, Lognormal, Normal common survival analysis functions successively carry out regression analysis, obtain each AIC's
Value, such as table 2:
The AIC comparison that table 2 is respectively distributed
| AIC | Mechanical failure |
| Loglogistic | 539.592 |
| Weibull | 435.479 |
| Logistic | 870.111 |
| Lognormal | 539.639 |
| Normal | 1479.622 |
According to AIC's as a result, choosing distribution building on the basis of Weibull accelerates failure model.
Step6, non-failure operation time prediction model of the building based on AFT, the cumulative distribution letter of non-failure operation time t
Number can indicate are as follows:
In formula, F (t) indicates that the duration is less than the probability of t, and S (t) makes a living store function, when indicating equipment failure-free operation
Between exceed the probability of t, also referred to as survival rate；F (t) is probability density function.The expression formula h (t) of risk function is
H (t)=f (t)/S (t)=f (t)/(1-F (t)) (3)
Influence of the AFT model explanation characteristic factor to equipment non-failure operation time is chosen, association is added in this in risk function
Variable, covariant are to cause the variation of non-failure operation time by the product with time variable, conditional risk function and life
Store function embodies form are as follows:
H (t | X)=ψ h0(ψt) (4)
S (t | X)=S0(ψt) (5)
In formula, ψ=EXP (- β ' X) indicates one group of covariant vector, β ' expression one group of estimation parameter corresponding with covariant
The transposition of vector, β ' X=β0+β1x1+…+βnxn.Since mechanical failure meets Weibull distribution, according to statistical knowledge
The probability density function of Weibull distribution are as follows:
F (t)=λ p (λ t)p-1EXP(-(λT)P) (8)
In formula, λ, p are respectively the dimensional parameters and form parameter of Weibull distribution, can derive Weibull distribution
Risk function and survival function expression formula are respectively
H (t)=(λ p) (λ t)p-1 (9)
S (t)=exp (- λ tp) (10)
Using 0.05 as significance critical value, selected using input variable of the method for gradual regression to model,
In view of model will be able to reflect more fully information, it is believed that be not more than 0.1 when be still effective.It is picked by regression analysis
Except the variable of interference model, available covariant selection and parameter estimation result such as table 3:
3 variables choice of table and parameter Estimation table
According to the estimated result of 3 model parameter of table, the risk of automatic fare collection machine equipment non-failure operation time model is obtained
Function, survival function are as follows.
Step7, the performance evaluation to model are referred to using mean absolute percentage error (MAPE) as the evaluation of model
Mark, its expression formula of MAPE are as follows:
Wherein, AiIt is the actual value of i-th of observed quantity, PiIt is the predicted value of i-th of observed quantity.
Very accurate, good and inaccurate three ranks are represented using the section of 10%, 20% and 50% MAPE value,
For the performance of further comparative illustration this method prediction, same fault data is chosen, establishes multiple linear regression model, together
Sample predicts the non-failure operation time of equipment, compares the prediction result of two kinds of models.Prediction result is shown, by being based on
The MAPE value of the non-failure operation time prediction model of AFT is 9.65%, the failure free time prediction model based on regression analysis
MAPE value be 16.43%, it can be seen that the non-failure operation time prediction model proposed in this paper based on AFT have it is very quasi-
True estimated performance, can be good at the fault observer for reflecting equipment, and the formulation maintenance policy for science provides good ginseng
It examines.
The basic principles, main features and advantages of the present invention have been shown and described above.The technology of the industry
Personnel only illustrate the present invention it should be appreciated that the present invention is not limited by examples detailed above described in examples detailed above and specification
Principle, various changes and improvements may be made to the invention without departing from the spirit and scope of the present invention, these variation and
Improvement all fall within the protetion scope of the claimed invention.The claimed scope of the invention is by appended claims and its is equal
Object defines.
Claims (8)
1. a kind of urban rail equipment non-failure operation time prediction technique based on survival analysis, it is characterised in that: including walking as follows
It is rapid:
S1, fault signature analysis of Influential Factors: the characteristic factor that analysis impacts equipment fault；
S2, to the quantization of feature influence factor and assignment explanation；
S3, failure modes: being different type by fault type induction and conclusion；
Fault data statistical disposition: S4 carries out statistic of classification to the fault data of equipment, and carries out data cleansing, reject mistake
Data.
Model of failure distribution selection: S5 selects model of failure distribution by akaike information criterion；
S6 constructs the non-failure operation time prediction model based on acceleration time failure model: selection acceleration time failure model
The building of non-failure operation time prediction model is carried out to different types of faults respectively；
S7, model prediction and analysis: the prediction of equipment non-failure operation time is carried out according to the model of building, to the predictability of model
It can be carried out assay, and the influence according to interpretation of result characteristic factor to equipment non-failure operation time.
2. a kind of urban rail equipment non-failure operation time prediction technique based on survival analysis according to claim 1,
It is characterized in that: binaryzation quantization method being taken to the quantization method of feature influence factor in the step S2.
3. a kind of urban rail equipment non-failure operation time prediction technique according to claim 1, it is characterised in that: the step
Fault type is divided into mechanical failure and electric information failure in rapid S3.
4. a kind of urban rail equipment non-failure operation time prediction technique based on survival analysis according to claim 2 or 3,
It is characterized by: random failure event reaches three times by once connection in 24 hours in the step S4 fault data cleaning
Property fault statistics；Same failure occurred again less than 24 hours after fault restoration, only counts primary fault；Multiple fault modes by
When same component failure causes, primary fault is only counted.
5. a kind of urban rail equipment non-failure operation time prediction technique based on survival analysis according to claim 4,
It is characterized in that: akaike information criterion expression formula described in the step 5 are as follows:
AIC=-2log (L)+2n
In formula, L is the maximum likelihood function value of model, and n is the number of all parameters in model.
6. a kind of urban rail equipment non-failure operation time prediction technique based on survival analysis according to claim 4,
Be characterized in that: the step S6 constructs time prediction model, the conversion process of the risk function by establishing risk function
It is as follows:
For a random time variable T, Cumulative Distribution Function can be indicated are as follows:
In formula, F (t) indicates that duration T is less than the probability of t, and S (t) makes a living store function, and f (t) is probability density function, expression
Formula are as follows:
According to the definition of risk function, the expression formula h (t) of available risk function is
H (t)=f (t)/S (t)=f (t)/(1-F (t))
7. a kind of urban rail equipment non-failure operation time prediction technique based on survival analysis according to claim 6,
Be characterized in that: the step S6 be based on acceleration time failure model, risk function be added covariant, conditional risk function and
Survival function embodies form are as follows:
H (t | X)=ψ h0(ψt)
S (t | X)=S0(ψt)
In formula, ψ=EXP (- β ' X) indicates one group of covariant vector, β ' expression one group of estimation parameter vector corresponding with covariant
Transposition, β ' X=β0+β1x1+…+βnxn, h0(·)、S0() is illustrated respectively in baseline risk letter when all covariants are zero
Several and benchmark survival function.
8. a kind of urban rail equipment non-failure operation time prediction side based on survival analysis according to claim 5 or 6 or 7
Method, it is characterised in that: described average with mean absolute percentage error come the predictability of assay model in the step S7
Absolute percent error is less than 20%.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811162487.8A CN109299828A (en) | 2018-09-30 | 2018-09-30 | Method for predicting fault-free running time of urban rail equipment based on survival analysis |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811162487.8A CN109299828A (en) | 2018-09-30 | 2018-09-30 | Method for predicting fault-free running time of urban rail equipment based on survival analysis |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109299828A true CN109299828A (en) | 2019-02-01 |
Family
ID=65161579
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811162487.8A Pending CN109299828A (en) | 2018-09-30 | 2018-09-30 | Method for predicting fault-free running time of urban rail equipment based on survival analysis |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109299828A (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110400013A (en) * | 2019-07-22 | 2019-11-01 | 西北工业大学 | A kind of mobile application extinction prediction technique based on multi-task learning mechanism |
| CN110516848A (en) * | 2019-07-31 | 2019-11-29 | 国网江苏省电力有限公司 | A kind of electric equipment maintenance Cost Optimization Approach based on survival analysis model |
| CN110609233A (en) * | 2019-10-25 | 2019-12-24 | 沃特威（广州）电子科技有限公司 | Method for predicting SOH of energy storage battery based on big data |
| CN111144639A (en) * | 2019-12-24 | 2020-05-12 | 国电南京自动化股份有限公司 | Subway equipment fault prediction method and system based on ALLN algorithm |
| CN116523349A (en) * | 2023-05-19 | 2023-08-01 | 北京协合运维风电技术有限公司 | Wind power station reliability analysis method and system |
-
2018
- 2018-09-30 CN CN201811162487.8A patent/CN109299828A/en active Pending
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110400013A (en) * | 2019-07-22 | 2019-11-01 | 西北工业大学 | A kind of mobile application extinction prediction technique based on multi-task learning mechanism |
| CN110516848A (en) * | 2019-07-31 | 2019-11-29 | 国网江苏省电力有限公司 | A kind of electric equipment maintenance Cost Optimization Approach based on survival analysis model |
| CN110516848B (en) * | 2019-07-31 | 2022-08-16 | 国网江苏省电力有限公司 | Power equipment maintenance cost optimization method based on survival analysis model |
| CN110609233A (en) * | 2019-10-25 | 2019-12-24 | 沃特威（广州）电子科技有限公司 | Method for predicting SOH of energy storage battery based on big data |
| CN111144639A (en) * | 2019-12-24 | 2020-05-12 | 国电南京自动化股份有限公司 | Subway equipment fault prediction method and system based on ALLN algorithm |
| CN116523349A (en) * | 2023-05-19 | 2023-08-01 | 北京协合运维风电技术有限公司 | Wind power station reliability analysis method and system |
| CN116523349B (en) * | 2023-05-19 | 2024-01-23 | 北京协合运维风电技术有限公司 | Wind power station reliability analysis method and system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109299828A (en) | Method for predicting fault-free running time of urban rail equipment based on survival analysis | |
| CN107843718B (en) | Method for evaluating aging state of transformer insulating oil | |
| Balijepalli et al. | Modeling and analysis of distribution reliability indices | |
| CN105894133B (en) | Wind turbine component maintenance and spare part demand forecasting method | |
| CN107831415B (en) | Interval value fuzzy set method for transformer insulation paper aging state evaluation | |
| CN107831300B (en) | Transformer insulating oil degradation evaluation method based on three-dimensional trapezoidal probability fuzzy set | |
| CN103400310B (en) | Method for evaluating power distribution network electrical equipment state based on historical data trend prediction | |
| CN109299827A (en) | Urban rail equipment fault prediction method based on long-time and short-time memory cyclic neural network | |
| MX2013000577A (en) | Machine learning for power grids. | |
| CN105844435A (en) | Subway vehicle fault information management system based on FMECA | |
| CN104392391A (en) | Power grid running safety risk quantification method | |
| CN108898311A (en) | A kind of data quality checking method towards intelligent distribution network repairing dispatching platform | |
| CN105035902A (en) | Elevator safety condition evaluation method | |
| CN111999692B (en) | Multi-table error calibration method and device | |
| CN107194149A (en) | A kind of Expressway Electromechanical System health index appraisal procedure | |
| CN109740772A (en) | Railroad train Measuring error analysis method based on big data | |
| Luo et al. | A new accident analysis method based on complex network and cascading failure | |
| Khabibullin et al. | The study and management of reliability parameters for automotive equipment using simulation modeling | |
| CN117689214A (en) | Dynamic safety assessment method for energy router of flexible direct-current traction power supply system | |
| CN118211034A (en) | Multi-dimensional civil aviation passenger flow prediction method based on KNN regression model | |
| Li et al. | Study on resource scheduling method of predictive maintenance for equipment based on knowledge | |
| Du et al. | Risk management of weather-related failures in distribution systems based on interpretable extra-trees | |
| As et al. | Estimation of travel time variability using bus probe data | |
| Dai et al. | Smart Electricity Meter Reliability Analysis Based on In-service Data | |
| CN205643673U (en) | Metering device scraps alarm device based on measurement instrument follow -up of quality evaluation system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190201 |