CN107909161B - Traction power supply equipment maintenance method based on risk cost prediction - Google Patents

Traction power supply equipment maintenance method based on risk cost prediction Download PDF

Info

Publication number
CN107909161B
CN107909161B CN201711191099.8A CN201711191099A CN107909161B CN 107909161 B CN107909161 B CN 107909161B CN 201711191099 A CN201711191099 A CN 201711191099A CN 107909161 B CN107909161 B CN 107909161B
Authority
CN
China
Prior art keywords
maintenance
equipment
kth
value
normal
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN201711191099.8A
Other languages
Chinese (zh)
Other versions
CN107909161A (en
Inventor
林圣�
冯玎
牟大林
何正友
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Southwest Jiaotong University
Original Assignee
Southwest Jiaotong University
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Southwest Jiaotong University filed Critical Southwest Jiaotong University
Priority to CN201711191099.8A priority Critical patent/CN107909161B/en
Publication of CN107909161A publication Critical patent/CN107909161A/en
Application granted granted Critical
Publication of CN107909161B publication Critical patent/CN107909161B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

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
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/24—Classification techniques
    • G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
    • G06F18/2415—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on parametric or probabilistic models, e.g. based on likelihood ratio or false acceptance rate versus a false rejection rate
    • G06F18/24155—Bayesian classification
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • 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
    • 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
    • G06Q50/06—Energy or water supply
    • 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
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
    • Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
    • Y04S10/00—Systems supporting electrical power generation, transmission or distribution
    • Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications

Landscapes

  • Business, Economics & Management (AREA)
  • Engineering & Computer Science (AREA)
  • Human Resources & Organizations (AREA)
  • Economics (AREA)
  • Physics & Mathematics (AREA)
  • Strategic Management (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Tourism & Hospitality (AREA)
  • General Business, Economics & Management (AREA)
  • Marketing (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Health & Medical Sciences (AREA)
  • Operations Research (AREA)
  • Quality & Reliability (AREA)
  • Primary Health Care (AREA)
  • Development Economics (AREA)
  • General Health & Medical Sciences (AREA)
  • Game Theory and Decision Science (AREA)
  • Data Mining & Analysis (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • General Engineering & Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Probability & Statistics with Applications (AREA)
  • Evolutionary Computation (AREA)
  • Educational Administration (AREA)
  • Evolutionary Biology (AREA)
  • Public Health (AREA)
  • Water Supply & Treatment (AREA)
  • Artificial Intelligence (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

The invention discloses a traction power supply equipment maintenance method based on risk cost prediction. The method mainly comprises the following steps: subtracting the posterior probability of the normal state of each device from the posterior probability of the fault state, which is obtained by the Bayesian network classifier model, so as to obtain the health index of each device, and then introducing a health index correction value to obtain a corrected running state predicted value; further, the comparison operation of the predicted value of the corrected running state and the actual predicted value is carried out to obtain the predicted accuracy, and further risk cost is obtained; and obtaining the health correction value with the minimum risk cost and the optimal maintenance scheme by iterative operation of different health indexes. The maintenance scheme provided by the method reduces the error of the prediction result, improves the state prediction accuracy of each device, and obviously increases the maintenance cost.

Description

Traction power supply equipment maintenance method based on risk cost prediction
Technical Field
The invention relates to a traction power supply equipment maintenance method based on risk cost prediction.
Background
The maintenance work of the high-speed rail traction power supply equipment can effectively discover and process faults, and makes great contribution to guaranteeing safe and reliable operation of a traction power supply system. Currently, maintenance (including inspection and repair) of railroad traction power supply systems typically employs periodic maintenance at fixed time intervals. The maintenance mode has been implemented for many years in the traction power supply field, and can ensure the safe and reliable operation of traction power supply equipment to a large extent. However, such maintenance in a fixed period is liable to cause insufficient maintenance or excessive maintenance, and cannot ensure maintenance of high reliability of the traction power supply equipment, and at the same time, is costly.
The state maintenance method based on equipment fault prediction predicts the running state of equipment through intelligent algorithms such as a Bayesian network, a support vector machine, a neural network and the like; and (3) giving a maintenance scheme according to the predicted state, wherein the equipment with the normal predicted state is not maintained (detected and repaired), and the equipment with the fault predicted state is maintained. The maintenance scheme based on state prediction can reduce the phenomenon of insufficient maintenance or excessive maintenance. However, the existing maintenance scheme based on state prediction does not correct the state prediction result based on the historical maintenance condition, so that the error of the prediction result is large, the maintenance scheme is to be optimized, and the saved maintenance cost is limited. Meanwhile, the existing maintenance scheme does not give out the cost saved by the maintenance scheme and the traditional fixed period maintenance mode, namely, the reliability of the maintenance scheme can be improved, the maintenance cost is reduced, the maintenance scheme is difficult to personel use, and the maintenance scheme is difficult to popularize and implement.
Disclosure of Invention
The invention aims to provide a traction power supply equipment maintenance method based on risk cost prediction, which has the advantages of small prediction result error and high maintenance cost; and is easy to popularize and implement.
The technical scheme adopted for achieving the purpose of the invention is that the traction power supply equipment maintenance method based on risk cost prediction comprises the following steps:
A. input data
The actual running state value c of the equipment during each maintenance of each traction power supply equipment in the history maintenance record k Ambient temperature value X during each maintenance 1k Ice and snow value X 2k Rainfall value X 3k Lightning value X 4k Wind speed value X 5k Load value X 6k Human factor value X 7k Seven influence variable value input systems; wherein K is the number of times of maintenance, k=1, 2, …, K-1; k is the number of times of maintenance to be performed currently, c k ＝0,1；c k =0 means that the equipment is normal at the kth maintenance, c k =1 indicates that the apparatus failed at the kth maintenance;
B. probability of acquiring operational state information
Seven influencing variables are used as node evidence variables of the Bayesian network, and a Bayesian network classifier model is constructed; actual operating state value c of each equipment during each maintenance k And substituting seven influence variable values in each maintenance into a Bayesian network classifier model to obtain posterior probability P (c|X) of seven influence variables and equipment running state c 1 ,X 2 ,...,X 7 ) The method comprises the steps of carrying out a first treatment on the surface of the Further, a posterior probability P (c= 1|X) that the equipment operation state c is a fault under the combined action of seven influencing variables is obtained 1 ,X 2 ,...,X 7 ) Posterior probability P of normal device operation state c under combined action of seven influencing variables (c= 0|X) 1 ,X 2 ,...,X 7 ) The method comprises the steps of carrying out a first treatment on the surface of the Wherein X is 1 ,X 2 ,...,X 7 Respectively represent the influencing variables: ambient temperature, ice and snow, rainfall, lightning, wind speed, load and human factors;
C. calculation of posterior probability
According to the value X of seven influencing variables at the time of the kth maintenance 1K ,X 2K ,...,X 7K Posterior probability of equipment failure P (c= 1|X) under the combined action of seven influencing variables from step B 1 ,X 2 ,...,X 7 ) Obtaining posterior probability P of equipment operation state as fault in kth maintenance k (c＝1|X 1K ,X 2K ,...,X 7K ) Wherein k=1, 2, …, K;
according to the value X of seven influencing variables at the time of the kth maintenance 1K ,X 2K ,...,X 7K Posterior probability P (c=) of device normal under combined action of seven influencing variables of step B0|X 1 ,X 2 ,...,X 7 ) Obtaining posterior probability P of normal equipment operation state in kth maintenance k (c＝0|X 1K ,X 2K ,...,X 7K ) Wherein k=1, 2, …, K;
D. running state prediction and maintenance scheme formulation
Posterior probability P for setting the operation state of the kth maintenance equipment to be normal k (c k ＝0|X 1K ,X 2K ,...,X 7K ) Subtracting the posterior probability P of the operation state of the equipment maintained for the kth time as a fault k (c k ＝1|X 1K ,X 2K ,...,X 7K ) Obtaining a device health index b of the kth maintenance; device health index b as the kth maintenance>The predicted operating state c 'of the device at the kth maintenance time' k Is normal, i.e. c' k =0; otherwise, the predicted operating state c 'of the device at the kth maintenance' k Is a fault, i.e. c' k =1; wherein k=1, 2, …, K;
and (3) making a maintenance scheme: when c' K When the device is in the condition of being 1, the device needs to be maintained when the device is currently to be maintained; when c' K When=0, the maintenance to be performed currently does not perform maintenance on the apparatus;
E. calculation of prediction accuracy
The predicted operation states of K less than or equal to K-1 are regarded as faults, namely c' k The number of devices of =1 (k.ltoreq.K-1) is added to obtain the total number of predicted faulty devices Q 1 'A'; the actual operation state of each time K is less than or equal to K-1 is regarded as failure, namely c k The number of devices of =1 (k.ltoreq.K-1) is added to obtain the total number of actual faulty devices Q 1 Further, the failure prediction accuracy F, F=Q is obtained 1 '/Q 1 ；
The predicted running state of each time of K less than or equal to K-1 is normal, namely c' k The number of devices of =0 (k.ltoreq.K-1) is added to obtain the total number of predicted normal devices Q 0 ' the actual running state of K is less than or equal to K-1 is normal, namely c k The device numbers of the (K is less than or equal to K-1) are added to obtain the total number Q of the actual normal devices 0 And then obtaining the normal prediction accuracy rate R, R=Q 0 '/Q 0 ；
F. Risk cost assessment
The risk cost S, i.e. the equipment maintenance cost that is superfluously spent from the time of repair to the time of next repair, is derived from the following formula:
S＝[Q' K0 ×(1-R)/R]×S 1 +[Q' K1 ×(1-F)/F]×S 2
wherein S is 1 To detect the fault in the maintenance of a device, S 2 Net loss, Q 'for a piece of equipment that has not been serviced but failed prior to the next service' K0 Predicting the running state as normal, i.e. c ', for the kth maintenance (maintenance to be performed at present) equipment' K Number of devices=0, Q' K1 Predicting the operating state as faulty, i.e. c ', for the kth repair (repair currently to be performed) equipment' K Number of devices=1;
G. obtaining an optimal maintenance solution
Taking different health index correction values B, repeating the operations from the step D to the step F, and calculating risk cost S values under the different health index correction values B; comparing the respective risk cost values S to a minimum risk cost value S 0 Minimum risk cost value S 0 The corresponding health index correction value B is the optimal health correction index B 0 The method comprises the steps of carrying out a first treatment on the surface of the From the optimal health correction index value B 0 The predicted operating state c 'of the respective device' k The obtained maintenance scheme for the equipment to be maintained currently is the optimal maintenance scheme with minimum risk cost;
finally, the cost difference S between the optimal maintenance scheme and the fixed period maintenance scheme p ，S p ＝Q' K0 ×S 1 -[Q' K1 ×(1-F)/F]×S 2 The method comprises the steps of carrying out a first treatment on the surface of the The cost difference is the cost S saved by the optimal maintenance scheme compared with the fixed period maintenance mode p 。
Compared with the prior art, the invention has the beneficial effects that:
1. the method comprises the steps of obtaining a corrected running state predicted value by introducing a health index corrected value; comparing the predicted value of the corrected running state of the equipment in each maintenance with the actual state value to obtain the predicted accuracy, and further obtaining the risk cost (the cost of redundant cost in maintenance) of the maintenance scheme; and obtaining the health correction value with the minimum risk cost and the optimal maintenance scheme through iterative operation of different health indexes.
And compared with the prior art, the method directly obtains the predicted running state of each device and the maintenance scheme thereof by subtracting the posterior probability of the normal state of each device and the posterior probability of the fault state, which are obtained by the Bayesian network classifier model. According to the maintenance scheme with minimum risk cost obtained through health index iteration, the error of a prediction result is reduced, the state prediction accuracy of each device is improved, and the maintenance cost is obviously increased.
2. The invention directly and quantitatively gives the accuracy of the predicted result of the optimal maintenance scheme and saves maintenance cost compared with a fixed period maintenance mode. The best maintenance scheme is provided, the reliability is improved, the maintenance cost is reduced, the maintenance scheme can be fully persuaded for the maintenance unit to adopt, and the maintenance scheme is easy to popularize and implement.
Detailed Description
The present invention will be described in further detail with reference to the following embodiments.
Examples
The invention relates to a traction power supply equipment maintenance method based on risk cost prediction, which comprises the following steps:
A. input data
The actual running state value c of the equipment during each maintenance of each traction power supply equipment in the history maintenance record k Ambient temperature value X during each maintenance 1k Ice and snow value X 2k Rainfall value X 3k Lightning value X 4k Wind speed value X 5k Load value X 6k Human factor value X 7k Seven influence variable value input systems; wherein K is the number of times of maintenance, k=1, 2, …, K-1; k is the number of times of maintenance to be performed currently, c k ＝0,1；c k =0 means that the equipment is normal at the kth maintenance, c k =1 indicates that the apparatus failed at the kth maintenance;
B. probability of acquiring operational state information
Seven influencing variables are used as node evidence variables of the Bayesian network, and a Bayesian network classifier model is constructed; actual operating state value c of each equipment during each maintenance k And substituting seven influence variable values in each maintenance into a Bayesian network classifier model to obtain posterior probability P (c|X) of seven influence variables and equipment running state c 1 ,X 2 ,...,X 7 ) The method comprises the steps of carrying out a first treatment on the surface of the Further, a posterior probability P (c= 1|X) that the equipment operation state c is a fault under the combined action of seven influencing variables is obtained 1 ,X 2 ,...,X 7 ) Posterior probability P of normal device operation state c under combined action of seven influencing variables (c= 0|X) 1 ,X 2 ,...,X 7 ) The method comprises the steps of carrying out a first treatment on the surface of the Wherein X is 1 ,X 2 ,...,X 7 Respectively represent the influencing variables: ambient temperature, ice and snow, rainfall, lightning, wind speed, load and human factors;
C. calculation of posterior probability
According to the value X of seven influencing variables at the time of the kth maintenance 1K ,X 2K ,...,X 7K Posterior probability of equipment failure P (c= 1|X) under the combined action of seven influencing variables from step B 1 ,X 2 ,...,X 7 ) Obtaining posterior probability P of equipment operation state as fault in kth maintenance k (c＝1|X 1K ,X 2K ,...,X 7K ) Wherein k=1, 2, …, K;
according to the value X of seven influencing variables at the time of the kth maintenance 1K ,X 2K ,...,X 7K Posterior probability P of device normal under the combined action of seven influencing variables of step B (c= 0|X) 1 ,X 2 ,...,X 7 ) Obtaining posterior probability P of normal equipment operation state in kth maintenance k (c＝0|X 1K ,X 2K ,...,X 7K ) Wherein k=1, 2, …, K;
D. running state prediction and maintenance scheme formulation
Posterior probability P for setting the operation state of the kth maintenance equipment to be normal k (c k ＝0|X 1K ,X 2K ,...,X 7K ) Equipment operation minus kth repairPosterior probability of row state failure P k (c k ＝1|X 1K ,X 2K ,...,X 7K ) Obtaining a device health index b of the kth maintenance; device health index b as the kth maintenance>The predicted operating state c 'of the device at the kth maintenance time' k Is normal, i.e. c' k =0; otherwise, the predicted operating state c 'of the device at the kth maintenance' k Is a fault, i.e. c' k =1; wherein k=1, 2, …, K;
and (3) making a maintenance scheme: when c' K When the device is in the condition of being 1, the device needs to be maintained when the device is currently to be maintained; when c' K When=0, the maintenance to be performed currently does not perform maintenance on the apparatus;
E. calculation of prediction accuracy
The predicted operation states of K less than or equal to K-1 are regarded as faults, namely c' k The number of devices of =1 (k.ltoreq.K-1) is added to obtain the total number of predicted faulty devices Q 1 'A'; the actual operation state of each time K is less than or equal to K-1 is regarded as failure, namely c k The number of devices of =1 (k.ltoreq.K-1) is added to obtain the total number of actual faulty devices Q 1 Further, the failure prediction accuracy F, F=Q is obtained 1 '/Q 1 ；
The predicted running state of each time of K less than or equal to K-1 is normal, namely c' k The number of devices of =0 (k.ltoreq.K-1) is added to obtain the total number of predicted normal devices Q 0 ' the actual running state of K is less than or equal to K-1 is normal, namely c k The device numbers of the (K is less than or equal to K-1) are added to obtain the total number Q of the actual normal devices 0 And then obtaining the normal prediction accuracy rate R, R=Q 0 '/Q 0 ；
F. Risk cost assessment
The risk cost S, i.e. the equipment maintenance cost that is superfluously spent from the time of repair to the time of next repair, is derived from the following formula:
S＝[Q' K0 ×(1-R)/R]×S 1 +[Q' K1 ×(1-F)/F]×S 2
wherein S is 1 To detect the fault in the maintenance of a device, S 2 It is a device which is not maintained but is maintained next timeNet loss of previous failure, Q' K0 Predicting the running state as normal, i.e. c ', for the kth maintenance (maintenance to be performed at present) equipment' K Number of devices=0, Q' K1 Predicting the operating state as faulty, i.e. c ', for the kth repair (repair currently to be performed) equipment' K Number of devices=1;
G. obtaining an optimal maintenance solution
Taking different health index correction values B, repeating the operations from the step D to the step F, and calculating risk cost S values under the different health index correction values B; comparing the respective risk cost values S to a minimum risk cost value S 0 Minimum risk cost value S 0 The corresponding health index correction value B is the optimal health correction index B 0 The method comprises the steps of carrying out a first treatment on the surface of the From the optimal health correction index value B 0 The predicted operating state c 'of the respective device' k The obtained maintenance scheme for the equipment to be maintained currently is the optimal maintenance scheme with minimum risk cost; finally, the cost difference S between the optimal maintenance scheme and the fixed period maintenance scheme p ，
S p ＝(Q' K0 +Q' K1 )×S 1 -[Q' K1 S 1 +[Q' K1 ×(1-F)/F]×S 2 ]＝Q' K0 ×S 1 -[Q' K1 ×(1-F)/F]×S 2 The method comprises the steps of carrying out a first treatment on the surface of the The cost difference is the cost S saved by the optimal maintenance scheme compared with the fixed period maintenance mode p 。
Obviously, in the present invention, the actual failure at the kth (k=1, 2, …, K-1) maintenance, i.e. c k The number of devices=1 is: the sum of the number of devices predicted to fail at the kth time and actually found to fail at the time of repair and the number of devices predicted to be normal at the kth time but actually failed before the kth+1th time of repair. In the invention, c is the actual normal time in the kth maintenance k The number of devices=0 is: the sum of the number of devices predicted to be normal for the kth (k=1, 2, …, K-1) and not failed before the kth+1th repair and the number of devices predicted to be failed for the kth repair but not found (detected) at the kth repair is calculated.
X in the invention 1 ,X 2 ,...,X 7 The seven influencing variables are listed in the following table:

Claims (1)

1. a traction power supply equipment maintenance method based on risk cost prediction comprises the following steps:
A. input data
The actual running state value c of the equipment during each maintenance of each traction power supply equipment in the history maintenance record k Ambient temperature value X during each maintenance 1k Ice and snow value X 2k Rainfall value X 3k Lightning value X 4k Wind speed value X 5k Load value X 6k Human factor value X 7k Seven influence variable value input systems; wherein K is the number of times of maintenance, k=1, 2, …, K-1; k is the number of times of maintenance to be performed currently, c k ＝0,1；c k =0 means that the equipment is normal at the kth maintenance, c k =1 indicates that the apparatus failed at the kth maintenance;
B. probability of acquiring operational state information
Seven influencing variables are used as node evidence variables of the Bayesian network, and a Bayesian network classifier model is constructed; actual operating state value c of each equipment during each maintenance k And substituting seven influence variable values in each maintenance into a Bayesian network classifier model to obtain posterior probability P (c|X) of seven influence variables and equipment running state c 1 ,X 2 ,...,X 7 ) The method comprises the steps of carrying out a first treatment on the surface of the Further, a posterior probability P (c= 1|X) that the equipment operation state c is a fault under the combined action of seven influencing variables is obtained 1 ,X 2 ,...,X 7 ) Posterior probability P of normal device operation state c under combined action of seven influencing variables (c= 0|X) 1 ,X 2 ,...,X 7 ) The method comprises the steps of carrying out a first treatment on the surface of the Wherein X is 1 ,X 2 ,...,X 7 Respectively represent the influencing variables: ambient temperature, ice and snow, rainfall, lightning, wind speed, load and human factors;
C. calculation of posterior probability
According to the kth maintenanceSeven influencing variable values X 1K ,X 2K ,...,X 7K Posterior probability of equipment failure P (c= 1|X) under the combined action of seven influencing variables from step B 1 ,X 2 ,...,X 7 ) Obtaining posterior probability P of equipment operation state as fault in kth maintenance k (c＝1|X 1K ,X 2K ,...,X 7K ) Wherein k=1, 2, …, K;
according to the value X of seven influencing variables at the time of the kth maintenance 1K ,X 2K ,...,X 7K Posterior probability P of device normal under the combined action of seven influencing variables of step B (c= 0|X) 1 ,X 2 ,...,X 7 ) Obtaining posterior probability P of normal equipment operation state in kth maintenance k (c＝0|X 1K ,X 2K ,...,X 7K ) Wherein k=1, 2, …, K;
D. running state prediction and maintenance scheme formulation
Posterior probability P for setting the operation state of the kth maintenance equipment to be normal k (c k ＝0|X 1K ,X 2K ,...,X 7K ) Subtracting the posterior probability P of the operation state of the equipment maintained for the kth time as a fault k (c k ＝1|X 1K ,X 2K ,...,X 7K ) Obtaining a device health index b of the kth maintenance; device health index b as the kth maintenance>The predicted operating state c 'of the device at the kth maintenance time' k Is normal, i.e. c' k =0; otherwise, the predicted operating state c 'of the device at the kth maintenance' k Is a fault, i.e. c' k =1; wherein k=1, 2, …, K;
and (3) making a maintenance scheme: when c' K When the device is in the condition of being 1, the device needs to be maintained when the device is currently to be maintained; when c' K When=0, the maintenance to be performed currently does not perform maintenance on the apparatus;
E. calculation of prediction accuracy
The predicted operation states of K less than or equal to K-1 are regarded as faults, namely c' k The number of devices with K being less than or equal to K-1 is added to obtain the total number Q of the predicted fault devices 1 'A'; the actual operation state of each time K is less than or equal to K-1 is regarded as failure, namely c k The number of devices with K being less than or equal to K-1 is added to obtain the total number Q of actual fault devices 1 Further, the failure prediction accuracy F, F=Q is obtained 1 '/Q 1 ；
The predicted running state of each time of K less than or equal to K-1 is normal, namely c' k The number of devices with K being equal to or less than K-1 is added to obtain the total number Q of the predicted normal devices 0 ' the actual running state of K is less than or equal to K-1 is normal, namely c k The number of devices with K being less than or equal to K-1 is added to obtain the total number Q of actual normal devices 0 And then obtaining the normal prediction accuracy rate R, R=Q 0 '/Q 0 ；
F. Risk cost assessment
The risk cost S, i.e. the equipment maintenance cost that is superfluously spent from the time of repair to the time of next repair, is derived from the following formula:
S＝[Q' K0 ×(1-R)R]×S 1 +[Q' K1 ×(1-F)F]×S 2
wherein S is 1 To detect the fault in the maintenance of a device, S 2 Net loss, Q 'for a piece of equipment that has not been serviced but failed prior to the next service' K0 Predicting the running state of the Kth maintenance equipment to be normal, namely c' K Number of devices=0, Q' K1 Predicting the operation state as failure, namely c ', for the Kth maintenance equipment' K Number of devices=1, where kth repair is currently scheduled repair;
G. obtaining an optimal maintenance solution
Taking different health index correction values B, repeating the operations from the step D to the step F, and calculating risk cost S values under the different health index correction values B; comparing the respective risk cost values S to a minimum risk cost value S 0 Minimum risk cost value S 0 The corresponding health index correction value B is the optimal health correction index B 0 The method comprises the steps of carrying out a first treatment on the surface of the From the optimal health correction index value B 0 The predicted operating state c 'of the respective device' k The obtained maintenance scheme for the equipment to be maintained currently is the optimal maintenance scheme with minimum risk cost;
finally, the cost difference S between the optimal maintenance scheme and the fixed period maintenance scheme p ，S p ＝(Q' K0 +Q' K1 )×S 1 -[Q' K1 S 1 +[Q' K1 ×(1-F)F]×S 2 ]＝Q' K0 ×S 1 -[Q' K1 ×(1-F)F]×S 2 The method comprises the steps of carrying out a first treatment on the surface of the The cost difference is the cost S saved by the optimal maintenance scheme compared with the fixed period maintenance mode p 。
CN201711191099.8A 2017-11-24 2017-11-24 Traction power supply equipment maintenance method based on risk cost prediction Active CN107909161B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201711191099.8A CN107909161B (en) 2017-11-24 2017-11-24 Traction power supply equipment maintenance method based on risk cost prediction

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201711191099.8A CN107909161B (en) 2017-11-24 2017-11-24 Traction power supply equipment maintenance method based on risk cost prediction

Publications (2)

Publication Number Publication Date
CN107909161A CN107909161A (en) 2018-04-13
CN107909161B true CN107909161B (en) 2023-07-25

Family

ID=61847740

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201711191099.8A Active CN107909161B (en) 2017-11-24 2017-11-24 Traction power supply equipment maintenance method based on risk cost prediction

Country Status (1)

Country Link
CN (1) CN107909161B (en)

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105096053A (en) * 2015-08-14 2015-11-25 哈尔滨工业大学 Health management decision-making method suitable for complex process system
CN106503813A (en) * 2016-10-27 2017-03-15 清华大学 Prospective maintenance decision-making technique and system based on hoisting equipment working condition
CN106599541A (en) * 2016-11-23 2017-04-26 华南理工大学 Online structure and parameter identification method for dynamic power load model

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6748341B2 (en) * 2002-04-12 2004-06-08 George E. Crowder, Jr. Method and device for machinery diagnostics and prognostics

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105096053A (en) * 2015-08-14 2015-11-25 哈尔滨工业大学 Health management decision-making method suitable for complex process system
CN106503813A (en) * 2016-10-27 2017-03-15 清华大学 Prospective maintenance decision-making technique and system based on hoisting equipment working condition
CN106599541A (en) * 2016-11-23 2017-04-26 华南理工大学 Online structure and parameter identification method for dynamic power load model

Non-Patent Citations (5)

* Cited by examiner, † Cited by third party
Title
Maintenance Decision-making Model Based on POMDP for Traction Power Supply Equipment and Its Application;Sheng Lin 等;《2016 Prognostics and System Health Management Conference (PHM-Chengdu)》;1-6 *
Vulnerability Analysis of Power Distribution Systems for Cost-Effective Resource Allocation;Carl Johan Wallnerstrom 等;《IEEE Transactions on Power Systems》;第27卷(第1期);224-232 *
基于贝叶斯网络的可靠性概率风险评价方法研究;熊耀刚;《中国优秀硕士学位论文全文数据库 信息科技辑》(第05期);I140-82 *
电站设备短期维修风险决策模型研究;董玉亮 等;《华北电力大学学报》;第34卷(第03期);45-49 *
考虑状态演变过程的高速铁路牵引供电设备维修策略;张奥;《中国优秀硕士学位论文全文数据库 工程科技Ⅱ辑》(第07期);C033-155 *

Also Published As

Publication number Publication date
CN107909161A (en) 2018-04-13

Similar Documents

Publication Publication Date Title
CN111555716B (en) Method, device, equipment and storage medium for determining working state of photovoltaic array
CN111812427A (en) Health state assessment method and system for power electronic device
CN111582542B (en) Power load prediction method and system based on anomaly repair
CN104390657A (en) Generator set operating parameter measuring sensor fault diagnosis method and system
CN106096058B (en) Charging network fault analysis quantitative method and quantitative device based on AHP
CN116308304B (en) New energy intelligent operation and maintenance method and system based on meta learning concept drift detection
CN111488896A (en) Distribution line time-varying fault probability calculation method based on multi-source data mining
CN111339661B (en) Automatic planning method for high-voltage cable inspection cycle
CN117764422B (en) Intelligent energy-saving operation and maintenance management cloud platform
CN109359742B (en) Method for generating preventive maintenance period of subway subsystem
CN110443481B (en) Power distribution automation terminal state evaluation system and method based on hybrid K-nearest neighbor algorithm
CN107909161B (en) Traction power supply equipment maintenance method based on risk cost prediction
CN114297265A (en) Intelligent operation and maintenance method based on Internet of things technology
CN110690703B (en) Fault defense online policy table generation method, online safety and stability emergency control method and system
CN112489841A (en) Water level fault-tolerant control method for steam generator of nuclear power unit
CN109033569B (en) Method for optimizing strength and times of preventive maintenance threshold of shipboard aircraft sensor system
CN114400776B (en) Digital mirror image-based substation automation equipment state diagnosis method and system
CN114708718A (en) Wind generating set temperature cluster control method, device, equipment and medium
CN112966785B (en) Intelligent constellation state identification method and system
CN112748663B (en) Wind power torque fault-tolerant control method based on data-driven output feedback
CN114386837A (en) Historical data-based power system bad deviation identification method and system
CN112001073A (en) Reliability analysis research method of traction power supply system
CN113052233A (en) Thermal power station equipment fault early warning system and method based on big data and neural network
CN117318024B (en) CNN neural network-based photovoltaic power generation power prediction management method and system
CN115434878B (en) Wind generating set temperature cluster control method, device, equipment and medium

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