CN116151524A - Accurate matching method applied to nuclear power plant work order task - Google Patents

Accurate matching method applied to nuclear power plant work order task Download PDF

Info

Publication number
CN116151524A
CN116151524A CN202111349953.5A CN202111349953A CN116151524A CN 116151524 A CN116151524 A CN 116151524A CN 202111349953 A CN202111349953 A CN 202111349953A CN 116151524 A CN116151524 A CN 116151524A
Authority
CN
China
Prior art keywords
score
work order
field
weight value
equipment
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
Application number
CN202111349953.5A
Other languages
Chinese (zh)
Inventor
张廉
蔡汉坤
王晓东
刘莉
杨朦
杨逗
李贵莲
温庆邦
李武平
富会佳
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.)
China Nuclear Power Operation Technology Corp Ltd
Original Assignee
China Nuclear Power Operation Technology Corp Ltd
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 China Nuclear Power Operation Technology Corp Ltd filed Critical China Nuclear Power Operation Technology Corp Ltd
Priority to CN202111349953.5A priority Critical patent/CN116151524A/en
Publication of CN116151524A publication Critical patent/CN116151524A/en
Pending legal-status Critical Current

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
    • G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
    • G06Q10/063—Operations research, analysis or management
    • G06Q10/0631—Resource planning, allocation, distributing or scheduling for enterprises or organisations
    • G06Q10/06311—Scheduling, planning or task assignment for a person or group
    • G06Q10/063112—Skill-based matching of a person or a group to a task
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
    • G06F16/33—Querying
    • G06F16/332—Query formulation
    • 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
    • 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)
  • Physics & Mathematics (AREA)
  • Economics (AREA)
  • Theoretical Computer Science (AREA)
  • Strategic Management (AREA)
  • General Physics & Mathematics (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Marketing (AREA)
  • General Business, Economics & Management (AREA)
  • Health & Medical Sciences (AREA)
  • Tourism & Hospitality (AREA)
  • Educational Administration (AREA)
  • Game Theory and Decision Science (AREA)
  • Quality & Reliability (AREA)
  • Operations Research (AREA)
  • Development Economics (AREA)
  • Public Health (AREA)
  • Water Supply & Treatment (AREA)
  • General Health & Medical Sciences (AREA)
  • Primary Health Care (AREA)
  • Mathematical Physics (AREA)
  • Computational Linguistics (AREA)
  • Data Mining & Analysis (AREA)
  • Databases & Information Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

The invention belongs to the field of nuclear power plant work order management, and relates to an accurate matching method applied to a nuclear power plant work order task. The method comprises the following steps: if the value of the field of the 'work object' in the work order task is 'E-equipment', establishing an accurate matching rule; if the value of the field of the 'work object' in the work order task is 'S-system', establishing an accurate matching rule; if the value of the field of the 'work object' in the work order task is 'L-position', establishing an accurate matching rule; if the values of the field of the 'work object' in the work order task are 'N-name', 'U-UIC', 'M-material', a precise matching rule is established. The advantages are that: the method has the advantages that historical work order tasks can be accurately pushed to the work order task preparation stage of the work management flow, accurate pushing of the historical work order task information is achieved, labor and time for searching the work order task information are saved, burden of work order preparation personnel is reduced, quality and efficiency of work order task preparation are improved, and safe and efficient implementation of various maintenance works of the nuclear power plant is ensured.

Description

Accurate matching method applied to nuclear power plant work order task
Technical Field
The invention belongs to the field of nuclear power plant work order management, and particularly relates to an accurate matching method applied to a nuclear power plant work order task.
Background
The existing nuclear power plant mostly establishes a nuclear power safety production management system and accumulates a large amount of work order task data. The work order task is a core object in the nuclear power plant work control business process, and each work needs to create the work order task (except for simple maintenance). The preparation of the work order task is to form a complete and accurate work package, and the work package gathers files required for carrying out a certain maintenance task, so that the work package is a tool necessary for maintenance personnel of the nuclear power plant to carry out maintenance activities. The related maintenance work can be carried out only when one work order task forms a complete work package.
Currently, nuclear power plants need to develop a large number of maintenance tasks each year, about tens of thousands of maintenance works are performed on a unit for one year, and each maintenance work is a work order task and needs to be prepared by a work preparer to prepare a work package. For experienced staff, preparing a work order task may be done well, but for new staff, and newly commissioned power plants, shutdown and shutdown operation events occur when preparing work packages due to problems such as insufficient risk analysis, insufficient work package preparation, and inadequate experience feedback. Therefore, it is necessary to accurately push the history similar work order task of the current work order task to the current work order preparer, and provide references for the work order preparer when working items such as risk analysis, security measures, work licenses, work resource requirements, cooperation work and the like are manufactured, so that safe, stable and efficient operation of the nuclear power plant is ensured.
Disclosure of Invention
The invention aims to provide an accurate matching method applied to a nuclear power plant work order task, which can accurately push a historical work order task to a work order preparer in a work order task preparation stage, and can be used for accurately acquiring the historical work order task similar to the currently prepared work order task by the nuclear power plant work preparer in the work order task preparation process, so that references are provided for the work order preparer to develop work package preparation.
The technical scheme of the invention is as follows: an accurate matching method applied to a nuclear power plant work order task comprises the following steps:
step 1: if the value of the field of the 'work object' in the work order task is 'E-equipment', establishing an accurate matching rule;
step 2: if the value of the field of the 'work object' in the work order task is 'S-system', establishing an accurate matching rule;
step 3: if the value of the field of the 'work object' in the work order task is 'L-position', establishing an accurate matching rule;
step 4: if the values of the field of the 'work object' in the work order task are 'N-name', 'U-UIC', 'M-material', a precise matching rule is established.
The precise matching rule comprises the following steps:
1) Constructing business logic rules and semantic similarity rules;
2) Weight value correction score
The job order task S in 1) is paired according to the following rule Score set The result in (2) is subjected to weight value correction, and the correction of the weight influences the sorting of the result;
(1) In all matching results, weight adjustment is required according to the fields of the nuclear power plant and the machine set number of the work order task;
a) If the fields of the nuclear power plant and the machine group number are identical, the weight correction value is X 1 ；
b) If the "nuclear power plant" fields are the same and the "machine group number" fields are different, the weight correction value is X 2 ；
c) If the fields of the nuclear power plant and the unit number are different, but the pile types are the same, the weight correction value is X 3 ；
d) The weight correction value X Correction 1 ＝X 1 Or X 2 Or X 3 ；
(2) When quality defect report is generated in matched work order task, the weight correction value is X Correction 2 ；
(3) When the input work order task and the matched work order task have the large and small repair numbers in the large and small repair number field, the weight correction value is X Correction 3 ；
(4) When the input work order task and the job type field of the matched work order task belong to the periodic test and operation task, if the description of disqualification exists in the work completion detailed field in the finished report corresponding to the matched work order task, the weight correction value is X Correction 4 ；
(5) Comparing the fields of the system code and the system name of the input work order task with the fields of the system code and the system name of the push work order task, and if one of the fields of the system code and the system name is identical, the weight correction value is X Correction 5 ；
(6) If the field of the 'work order type' in the pushed work order task is the work order of the defect class, the weight correction value is X Correction 6 ；
And (3) outputting: according to the matching rule, each S of the input work order task and the matched work order task Score set And the corresponding correction coefficients are recommended according to the order of the scores from high to low.
The business logic rule construction and the semantic similarity rule construction comprise the following steps:
if the input "nuclear power plant" field of the work order task and the "nuclear power plant" field of the inquired work order task do not belong to the same nuclear power plant, the matching principle is as follows:
if the "nuclear power plant" field in the input work order task belongs to CANDU6, CNP-300, AP1000 and VVER model, matching the "work order title" and "equipment name" fields according to semantic similarity when matching other heap types, wherein the weight value is a and the score is W 1 Match total score S, s=w 1 ；
If the 'nuclear power plant' field belongs to CNP-600, CNP-1000, HPR-1000 and CPR-1000 types in the input work order task, matching the 'work order title' and 'equipment name' fields according to semantic similarity when matching the work order task of CANDU6, CNP-300, AP1000 and VVER types, wherein the weight value is b and the score is W 2 Match total score S, s=w 2 ；
If the input work order task 'nuclear power plant' field belongs to CNP-600, CNP-1000, HPR-1000 and CPR-1000 models, when matching the work order tasks of CNP-600, CNP-1000, HPR-1000 and CPR-1000 models, the following principle needs to be considered:
for CNP-600, CNP-1000, HPR-1000 and CPR-1000 models, if the field of 'equipment coding' is normal, the equipment codes are all consistent, the weight value is c, and the score is W 3 The method comprises the steps of carrying out a first treatment on the surface of the The equipment codes except the first bit set number are all the same, the weight value is c, and the score is W 4 The method comprises the steps of carrying out a first treatment on the surface of the The equipment codes are identical except that the first bit computer group number and the intermediate digital code are different, the weight value is c, and the score is W 5 The method comprises the steps of carrying out a first treatment on the surface of the The 'worksheet title' fields are matched through semantic similarity, the weight value is d, and the score is W 6 Match total score S, s=w 3 +W 6 Or s=w 4 +W 6 Or s=w 5 +W 6 ；
For CNP-600, CNP-1000, HPR-1000 and CPR-1000 models, if the field of 'equipment code' is abnormal, equipment code is extracted from the field of 'work order title'; if normal equipment codes can be extracted from the field of the 'work order title', the equipment codes are all consistent, the weight value is e, and the score is W 7 The method comprises the steps of carrying out a first treatment on the surface of the The equipment codes except the first bit set number are all the same, the weight value is e, and the score is W 8 The method comprises the steps of carrying out a first treatment on the surface of the Apparatus and method for controlling the operation of a deviceThe codes are identical except that the first bit of the code is different from the middle digital code, the weight value is e, and the score is W 9 The method comprises the steps of carrying out a first treatment on the surface of the The 'worksheet title' fields are matched through semantic similarity, the weight value is f, and the score is W 10 Match total score S, s=w 7 +W 10 Or s=w 8 +W 10 Or s=w 9 +W 10 ；
For CNP-600, CNP-1000, HPR-1000 and CPR-1000 models, if the device code and the work order title field have no normal device code, the work order title and the device name fields are matched according to semantic similarity, the weight value is g, and the score is W 11 Total score of matching s=w 11 ；
According to the rule, the output result only pushes the total score to be given the relevant score w Limiting the score And the above fraction data (i.e., S.gtoreq.w Limiting the score ) Count into set S Score set 。
The business logic rule construction and the semantic similarity rule construction comprise the following steps:
if the input "nuclear power plant" field of the work order task and the queried "nuclear power plant" field belong to the same nuclear power plant, when the "equipment code" and the "work order title" and "equipment name" fields are considered, the following principle is considered:
considering the field of 'equipment coding', if the field of 'equipment coding' is normal, the equipment coding in the field of 'work order title' is not required to be extracted;
the equipment code field is abnormal, but when the normal equipment code exists in the 'work order title', the equipment code of the 'work order title' field needs to be extracted;
when the equipment coding field is abnormal and the equipment coding is not normal in the 'work order title', only the semantic similarity of the 'work order title' and the 'equipment name' is considered.
Further comprises:
when the input work order task 'nuclear power plant' field belongs to CNP-600, CNP-1000, HPR-1000 and CPR-1000 models, the matching principle is as follows:
when the 'equipment coding' field is normal, the equipment codes are all consistent, the weight value is h, and the score is W 12 The method comprises the steps of carrying out a first treatment on the surface of the The equipment codes except the first bit set number are all the same, the weight value is h, and the score is W 13 The method comprises the steps of carrying out a first treatment on the surface of the The equipment codes are identical except that the first bit computer group number and the intermediate digital code are different, the weight value is h, and the score is W 14 The method comprises the steps of carrying out a first treatment on the surface of the The 'worksheet title' fields are matched through semantic similarity, the weight value is i, and the score is W 15 Match total score S, s=w 12 +W 10 Or s=w 13 +W 10 Or s=w 14 +W 15 ；
When the field of the 'equipment code' is abnormal and the normal equipment code exists in the 'work order title', the equipment code of the 'work order title' field needs to be extracted, the equipment codes extracted by the 'work order title' field are all consistent, the weight value is j, and the score is W 16 The method comprises the steps of carrying out a first treatment on the surface of the The equipment codes extracted from the 'work order title' field are identical except for the first bit unit number, the weight value is j, and the score is W 17 The method comprises the steps of carrying out a first treatment on the surface of the The equipment codes extracted from the 'work sheet title' field are identical except that the first bit machine group number and the intermediate digital code are different, the weight value is j, and the score is W 18 The method comprises the steps of carrying out a first treatment on the surface of the Other contents of the 'work order title' field are matched through semantic similarity, the weight value is k, and the score is W 19 Match total score S, s=w 16 +W 19 Or s=w 17 +W 19 Or s=w 18 +W 19 ；
When the field of the 'equipment code' is abnormal and the 'work order title' field is not provided with normal equipment codes, the field of the 'equipment name' is matched according to semantic similarity, the weight value is l, and the score is W 20 The 'worksheet title' fields are matched through semantic similarity, the weight value is m, and the score is W 21 Match total score S, s=w 20 +W 21 ；
When the field of the 'equipment code' is abnormal, and the 'work order title' has no normal equipment code, and the 'equipment name' field cannot be matched, the weight value is n, and the score is W 22 The total score S is matched, and the total score S,S＝W 22 ；
when the input work order task 'nuclear power plant' field belongs to the CNP-300 model, the matching principle is as follows:
the equipment codes are all consistent, the weight value is o, and the score is W 23 The method comprises the steps of carrying out a first treatment on the surface of the The device codes are not matched, the device names are matched according to the semantic similarity, the weight value is p, and the score is W 24 The method comprises the steps of carrying out a first treatment on the surface of the The 'worksheet title' fields are matched according to semantic similarity, the weight value is q, and the score is W 25 The method comprises the steps of carrying out a first treatment on the surface of the Match total score S, s=w 23 +W 25 Or s=w 24 +W 25 ；
c) When the input work order task "nuclear power plant" field belongs to the CANDU6 model, the matching principle is as follows:
the field of 'equipment codes' is normal, the equipment codes are all consistent, the weight value is r, and the score is W 26 The method comprises the steps of carrying out a first treatment on the surface of the The equipment codes except the first bit set number are all the same, the weight value is r, and the score is W 27 The method comprises the steps of carrying out a first treatment on the surface of the In the equipment coding, except for the first bit unit number, the system codes are the same, the weight value is r, and the score is W 28 The method comprises the steps of carrying out a first treatment on the surface of the The 'worksheet title' fields are matched through semantic similarity, the weight value is s, and the score is W 29 Match total score S, s=w 26 +W 29 Or s=w 27 +W 29 Or s=w 28 +W 29 ；
When the equipment code field is abnormal and the normal equipment code exists in the 'work order title', the equipment code of the 'work order title' field needs to be extracted, the equipment codes extracted by the 'work order title' field are all consistent, the weight value is t, and the score is W 30 The method comprises the steps of carrying out a first treatment on the surface of the The equipment codes extracted from the 'work sheet title' field are identical except for the first bit unit number, the weight value is t, and the score is W 31 The method comprises the steps of carrying out a first treatment on the surface of the In the equipment codes extracted from the 'work sheet title' field, if the system codes are the same except the first bit unit number, the weight value is t and the score is W 32 The method comprises the steps of carrying out a first treatment on the surface of the Other contents of the 'work order title' field are matched through semantic similarity, the weight value is u, and the score is W 33 Match total score S, s=w 30 +W 33 Or s=w 31 +W 33 Or s=w 32 +W 33 ；
When the field of the 'equipment code' is abnormal and the 'work order title' has no normal equipment code, the field of the 'equipment name' is matched according to the semantic similarity, the weight value is v, and the score is W 34 The method comprises the steps of carrying out a first treatment on the surface of the The 'worksheet title' fields are matched through semantic similarity, the weight value is W, and the score is W 35 The method comprises the steps of carrying out a first treatment on the surface of the Match total score S, s=w 34 + W 35 ；
When the field of the 'equipment code' is abnormal and the 'work order title' is not provided with normal equipment code, and the field of the 'equipment name' cannot be matched, only the semantic similarity of the field of the 'work order title' is considered, the weight value is x, and the score is W 36 Match total score S, s=w 36 ；
When the input work order task "nuclear power plant" field belongs to the VVER model, the matching principle is as follows:
the field of 'equipment coding' is normal, the equipment codes are all consistent, the weight value is y, and the score is W 37 The method comprises the steps of carrying out a first treatment on the surface of the The device codes except the first bit set number are all the same, the weight value is y, and the score is W 38 The method comprises the steps of carrying out a first treatment on the surface of the The number of the first bit set is different from the last number in the equipment coding, the weight value is y, and the score is W 39 The method comprises the steps of carrying out a first treatment on the surface of the The numbers except the first bit set number and the middle number in the equipment code are different and are the same, the weight value is y, and the score is W 40 The method comprises the steps of carrying out a first treatment on the surface of the The device code is different from the first digit group number, the middle digit and the last digit, and the weight value is y and the score is W 41 The method comprises the steps of carrying out a first treatment on the surface of the The 'worksheet title' fields are matched through semantic similarity, the weight value is z, and the score is W 42 Match total score S, s=w 37 +W 42 Or s=w 38 +W 42 Or s=w 39 +W 42 Or s=w 40 +W 42 Or s=w 41 +W 42 ；
When the field of the 'equipment code' is abnormal and the normal equipment code exists in the 'work order title', the equipment code of the 'work order title' field needs to be extracted, and the equipment codes extracted from the 'work order title' field are all consistentThe weight value is a 1 Score of W 43 The method comprises the steps of carrying out a first treatment on the surface of the The equipment codes extracted from the 'work sheet title' field are identical except for the first bit set number, and the weight value is a 1 Score of W 44 The method comprises the steps of carrying out a first treatment on the surface of the The equipment codes extracted from the 'work sheet title' field are identical except that the first bit set number and the last number are different, and the weight value is a 1 Score of W 45 The method comprises the steps of carrying out a first treatment on the surface of the The equipment codes extracted from the 'work sheet title' field are identical except that the number of the first bit set and the number in the middle are different, and the weight value is a 1 Score of W 46 The method comprises the steps of carrying out a first treatment on the surface of the The equipment codes extracted from the 'work sheet title' field are different from the first digit group number, the middle digit and the last digit, and the other digits are the same, and the weight value is a 1 Score of W 47 The method comprises the steps of carrying out a first treatment on the surface of the The word contents of the 'work order title' field are matched through semantic similarity, and the weight value is b 1 Score of W 48 Match total score S, s=w 43 +W 48 Or s=w 44 +W 48 Or s=w 45 +W 48 Or s=w 46 +W 48 Or s=w 47 +W 48 ；
When the field of the 'equipment code' is abnormal and the 'work order title' has no normal equipment code, the field of the 'equipment name' is matched according to semantic similarity, and the weight value is c 1 Score of W 49 The method comprises the steps of carrying out a first treatment on the surface of the Matching the "work order titles" through semantic similarity, wherein the weight value is d 1 Score of W 50 Match total score S, s=w 49 + W 50 ；
When the field of the 'equipment code' is abnormal, the 'work order title' has no normal equipment code, and the 'equipment name' field cannot be matched, only the semantic similarity matching of the 'work order title' is considered, and the weight value is e 1 Score of W 51 Match total score S, s=w 51 ；
e) When the input work order task 'nuclear power plant' field belongs to an AP1000 model, the matching principle is as follows:
the field of 'equipment coding' is normal, the equipment codes are all consistent, and the weight value is f 1 Score of W 52 The method comprises the steps of carrying out a first treatment on the surface of the The weight value is f except the first bit set number in the equipment coding 1 Score of W 53 The method comprises the steps of carrying out a first treatment on the surface of the The device codes are identical except that the first bit set number and the last serial number are different, and the weight value is f 1 Score of W 54 The method comprises the steps of carrying out a first treatment on the surface of the The 'worksheet title' fields are matched through semantic similarity, and the weight value is g 1 Score of W 55 Match total score S, s=w 52 +W 55 Or s=w 53 +W 55 Or s=w 54 +W 55 ；
When the equipment code field is abnormal and the normal equipment code exists in the 'work order title', the equipment code in the 'work order title' field needs to be extracted, the equipment codes extracted by the 'work order title' field are all consistent, and the weight value is h 1 Score of W 56 The method comprises the steps of carrying out a first treatment on the surface of the The equipment codes extracted from the 'worksheet title' field are identical except for the first bit set number, and the weight value is h 1 Score of W 57 The method comprises the steps of carrying out a first treatment on the surface of the The equipment code extracted from the 'work order title' field is divided into a first bit unit number and a last serial number, and the weight value is h 1 Score of W 58 The method comprises the steps of carrying out a first treatment on the surface of the Other contents of the 'work order title' field are matched through semantic similarity, and the weight value is i 1 Score of W 59 The method comprises the steps of carrying out a first treatment on the surface of the Match total score S, s=w 56 +W 59 Or s=w 57 +W 59 Or s=w 58 +W 59 ；
When the field of the 'equipment code' is abnormal and the 'work order title' has no normal equipment code, the field of the 'equipment name' is matched according to semantic similarity, and the weight value is j 1 Score of W 60 The method comprises the steps of carrying out a first treatment on the surface of the The "worksheet header" field is matched by semantic similarity, and the weight value is k 1 Score of W 61 The method comprises the steps of carrying out a first treatment on the surface of the Match total score S, s=w 60 + W 61 ；
When the field of the 'equipment code' is abnormal, the 'work order title' has no normal equipment code, and the 'equipment name' field cannot be matched, only the semantic similarity of the 'work order title' is considered, and the weight value is l 1 Score ofW 62 The method comprises the steps of carrying out a first treatment on the surface of the Match total score S, s=w 62 ；
According to the rule, the output result only pushes the total score to be given the relevant score w Limiting the score And the above fraction data (i.e., S.gtoreq.w Limiting the score ) Count into set S Score set 。
The precise matching rule in the step 2 is as follows:
if the input work order task and the system number or system name field of the matched work order task can be matched; the "System number" field is completely matched, where the weight value is m 1 Score of W 63 The method comprises the steps of carrying out a first treatment on the surface of the The "system number" field is not matched, and then the "system name" field is matched, and the "system name" field is matched according to semantic similarity, where the weight value is m 1 Score of W 64 The method comprises the steps of carrying out a first treatment on the surface of the The "work order header" field of the work order task is subjected to semantic similarity matching, and the weight value is n 1 Score of W 65 Match total score S, s=w 63 +W 65 Or s=w 64 +W 65 ；
If the input work order task cannot be matched with the system number and system name fields of the matched work order task, the semantic similarity matching is carried out on the work order title fields of the work order task, and the weight value is o 1 Score of W 66 Match total score S, s=w 66 ；
According to the rule, the output result only pushes the total score to be given the relevant score w Limiting the score And the above fraction data (i.e., S.gtoreq.w Limiting the score ) Count into set S Score set ；
And (3) outputting: according to the matching rule, each S of the input work order task and the matched work order task Score set And recommending according to the order of the scores from high to low.
The precise matching rule in the step 3 is as follows:
the precise matching rule is as follows: if the input work order task and the matched work order task are 'room number' orThe "structure" field can be matched to the "room number" field, where the weight value is p 1 Score of W 67 The method comprises the steps of carrying out a first treatment on the surface of the The "room number" field is not matched, then the "structure" field is unmatched, and the "structure" field is completely matched, where the weight value is q 1 Score of W 68 The method comprises the steps of carrying out a first treatment on the surface of the The "work order header" field of the work order task is subjected to semantic similarity matching, and the weight value is r 1 Score of W 69 Match total score S, s=w 67 +W 69 Or s=w 68 +W 69 ；
If the input work order task can not be matched with the room number and structure fields of the matched work order task, the semantic similarity matching is carried out on the work order title field of the work order task, and the weight value is s 1 Score of W 70 Match total score S, s=w 70 ；
According to the rule, the output result only pushes the total score to be given the relevant score w Limiting the score And the above fraction data (i.e., S.gtoreq.w Limiting the score ) Count into set S Score set ；
And (3) outputting: according to the matching rule, each S of the input work order task and the matched work order task Score set And recommending according to the order of the scores from high to low.
The precise matching rule in the step 4 is as follows:
the precise matching rule is as follows:
the "work order header" field of the work order task is subjected to semantic similarity matching, and the weight value is t 1 Score of W 71 Match total score S, s=w 71 ；
According to the rule, the output result only pushes the total score to be given the relevant score w Limiting the score And the above fraction data (i.e., S.gtoreq.w Limiting the score ) Count into set S Score set ；
And (3) outputting: according to the matching rule, each S of the input work order task and the matched work order task Score set According to the score fromThe recommendation is made in the order of high to low.
The invention has the beneficial effects that: at present, in the work order task preparation stage of a nuclear power plant, a work order preparer cannot timely and accurately search historical similar work order task information related to a work order task to be prepared currently, so that events occur at present due to the fact that work order preparation is insufficient, risk analysis is insufficient, work order task experience feedback with historical problems is not in place, and the like. The method can accurately push the historical work order task to the work order task preparation stage of the work management flow, realize the accurate push of the historical work order task information, save the labor and time for searching the work order task information, reduce the burden of work order preparation personnel, improve the quality and efficiency of work order task preparation, and ensure the safe and efficient implementation of various maintenance works of the nuclear power plant.
Detailed Description
The present invention will be described in further detail with reference to specific examples.
Aiming at the current situation that a work order preparer cannot timely and accurately acquire the history similar work order task reference information in the preparation process of the nuclear power plant work order, the invention realizes the accurate matching of the nuclear power plant work order task and the history similar work order task through Chinese word segmentation, semantic similarity technology, specific business rules and intelligent pushing based on the work management flow of an ASP of the nuclear power plant.
Work order tasks involved in the invention: the method is a special word in a nuclear power safety production management system, and is used for splitting a plurality of sub-tasks under a work order due to professional restrictions, safety measures restrictions and the like. The work order task is a core object in the work control business process, and each work needs to create the work order task (except for simple maintenance). The work order task information in the nuclear power safety production management system consists of task outlines, work orders, resource demands, documents, work licenses, safety measure demands, spare parts, personnel, planning time, cooperation work, risk analysis, safety measures and the like.
The invention establishes an accurate pushing method of historical similar work order tasks related to the nuclear power plant work order tasks. The invention takes the work order to be prepared (the work order to be prepared is derived from the field information brought in the work application and the work application to work order converting process) as input, takes the historical work order task information as the query object, and outputs the historical similar work order task information matched with the currently prepared work order task. The method comprises the following two key points:
(1) The invention realizes the accurate matching of the work order task to be prepared and the similar work order task by combing and establishing the business logic rule between the fields in the work order task and the similar history work order task and the Chinese word segmentation and semantic similarity technology;
(2) According to the invention, the matching results in the step (1) can be reordered by introducing the weight correction factors in the work order task to be prepared and the historical similar work order task, so that the requirements of business personnel on related information are accurately met.
Through the steps, accurate pushing of the historical work order tasks matched with the work order tasks to be prepared to the work order preparation person is finally achieved.
The business object involved in the invention is work order task information of the nuclear power plant.
The work order task involves fields such as "work order type", "nuclear power plant", "equipment code", "equipment name", "work order title", "overhaul number", "job type", "system code", "system name", "room number", "structure", "work object", etc.
According to the information characteristics of the work order task, the situations of different fields are combined for analysis in the matching process, but each situation involves business logic rule construction, semantic similarity (semantic similarity is based on Chinese word segmentation) rule construction among different fields in the work order task information, and finally weight correction factors are introduced to achieve accurate matching of the two types of information.
Because the precisely matched objects are historical work order tasks, the precisely matched objects need to be considered according to 'work objects' in the work order tasks.
An accurate matching method applied to a nuclear power plant work order task comprises the following steps:
step 1: if the value of the field of the 'work object' in the work order task is 'E-equipment', establishing an accurate matching rule:
the precise matching rule comprises the following steps:
1) Construction of business logic rules and construction of semantic similarity rules
(1) If the input "nuclear power plant" field of the work order task and the "nuclear power plant" field of the inquired work order task do not belong to the same nuclear power plant, the matching principle is as follows:
if the "nuclear power plant" field in the input work order task belongs to CANDU6, CNP-300, AP1000 and VVER model, when matching other heap types, the equipment codes corresponding to other power plants are different from the equipment codes of the input work order task, so that the "work order title" and "equipment name" fields are matched according to semantic similarity, wherein the weight value is a and the score is W 1 Match total score S, s=w 1 。
If the 'nuclear power plant' field belongs to CNP-600, CNP-1000, HPR-1000 and CPR-1000 types in the input work order task, matching the 'work order title' and 'equipment name' fields according to semantic similarity when matching the work order task of CANDU6, CNP-300, AP1000 and VVER types, wherein the weight value is b and the score is W 2 Match total score S, s=w 2 。
If the input work order task 'nuclear power plant' field belongs to CNP-600, CNP-1000, HPR-1000 and CPR-1000 models, when matching the work order tasks of CNP-600, CNP-1000, HPR-1000 and CPR-1000 models, the following principle needs to be considered:
for CNP-600, CNP-1000, HPR-1000 and CPR-1000 models, if the field of 'device coding' is normal, the device codes are all consistent (same device), the weight value is c, and the score is W 3 The method comprises the steps of carrying out a first treatment on the surface of the The equipment codes except the first bit set number are all the same, the weight value is c, and the score is W 4 The method comprises the steps of carrying out a first treatment on the surface of the The equipment codes are identical except that the first bit computer group number and the intermediate digital code are different, the weight value is c, and the score is W 5 The method comprises the steps of carrying out a first treatment on the surface of the The 'worksheet title' fields are matched through semantic similarity, the weight value is d, and the score is W 6 . Match total score S, s=w 3 +W 6 Or s=w 4 +W 6 Or s=w 5 +W 6 。
For CNP-600, CNP-1000, HPR-1000, CPR-1000 models, if the "device code" field is abnormal, the device code needs to be extracted from the "work order header" field. If the normal device codes can be extracted from the field of the 'work order title', the device codes are all consistent (same device), the weight value is e, and the score is W 7 The method comprises the steps of carrying out a first treatment on the surface of the The equipment codes except the first bit set number are all the same, the weight value is e, and the score is W 8 The method comprises the steps of carrying out a first treatment on the surface of the The equipment codes are identical except that the first bit computer group number and the intermediate digital code are different, the weight value is e, and the score is W 9 The method comprises the steps of carrying out a first treatment on the surface of the The 'worksheet title' fields are matched through semantic similarity, the weight value is f, and the score is W 10 . Match total score S, s=w 7 +W 10 Or s=w 8 +W 10 Or s=w 9 +W 10 。
For CNP-600, CNP-1000, HPR-1000 and CPR-1000 models, if the device code and the work order title field have no normal device code, the work order title and the device name fields are matched according to semantic similarity, the weight value is g, and the score is W 11 Total score of matching s=w 11 。
According to the rule, the output result only pushes the total score to be given the relevant score w Limiting the score And the above fraction data (i.e., S.gtoreq.w Limiting the score ) Count into set S Score set 。
(2) If the "nuclear power plant" field of the input work order task and the "nuclear power plant" field to be queried belong to the same nuclear power plant. When considering the "device code" and "work order title", "device name" fields. Consider the following principle:
considering the field of 'equipment code', if the field of 'equipment code' is normal (refers to the equipment code is not null and belongs to the conventional code of the factory), the equipment code in the field of 'work order header' is not required to be extracted;
The equipment code field is abnormal (the code is empty or inconsistent with the conventional code of the factory), but when the normal equipment code exists in the 'work order title', the equipment code of the 'work order title' field needs to be extracted;
the device code field is abnormal (the code is empty or inconsistent with the conventional code of the factory), and the semantic similarity of the 'work order title' and the 'device name' is only considered when the normal device code is not found in the 'work order title'.
a) When the input work order task 'nuclear power plant' field belongs to CNP-600, CNP-1000, HPR-1000 and CPR-1000 models, the matching principle is as follows:
the "device code" field is normal. The device codes are all consistent (same device), the weight value is h, and the score is W 12 The method comprises the steps of carrying out a first treatment on the surface of the The equipment codes except the first bit set number are all the same, the weight value is h, and the score is W 13 The method comprises the steps of carrying out a first treatment on the surface of the The equipment codes are identical except that the first bit computer group number and the intermediate digital code are different, the weight value is h, and the score is W 14 The method comprises the steps of carrying out a first treatment on the surface of the The 'worksheet title' fields are matched through semantic similarity, the weight value is i, and the score is W 15 . Match total score S, s=w 12 +W 10 Or s=w 13 +W 10 Or s=w 14 +W 15 。
When the field of the "equipment code" is abnormal and the normal equipment code exists in the "work order title", the equipment code of the field of the "work order title" needs to be extracted. The device codes extracted from the 'worksheet title' field are all consistent (same device), the weight value is j, and the score is W 16 The method comprises the steps of carrying out a first treatment on the surface of the The equipment codes extracted from the 'work order title' field are identical except for the first bit unit number, the weight value is j, and the score is W 17 The method comprises the steps of carrying out a first treatment on the surface of the The equipment codes extracted from the 'work sheet title' field are identical except that the first bit machine group number and the intermediate digital code are different, the weight value is j, and the score is W 18 The method comprises the steps of carrying out a first treatment on the surface of the Other contents of the 'work order title' field are matched through semantic similarity, the weight value is k, and the score is W 19 . Match total score S, s=w 16 +W 19 Or s=w 17 +W 19 Or s=w 18 +W 19 。
"device code" fieldWhen the device code is abnormal and the normal device code is not found in the 'work order title' field, the 'device name' field is matched according to the semantic similarity, the weight value is l, and the score is W 20 . The 'worksheet title' fields are matched through semantic similarity, the weight value is m, and the score is W 21 . Match total score S, s=w 20 +W 21 。
When the field of the 'equipment code' is abnormal, and the 'work order title' is not provided with normal equipment codes, and the field of the 'equipment name' cannot be matched, only the matching of the semantic similarity of the 'work order title' is considered, the weight value is n, and the score is W 22 . Match total score S, s=w 22 。
b) When the input work order task 'nuclear power plant' field belongs to the CNP-300 model, the matching principle is as follows:
The device codes are all consistent (same device), the weight value is o, and the score is W 23 The method comprises the steps of carrying out a first treatment on the surface of the The device codes are not matched, the device names are matched according to the semantic similarity, the weight value is p, and the score is W 24 The method comprises the steps of carrying out a first treatment on the surface of the The 'worksheet title' fields are matched according to semantic similarity, the weight value is q, and the score is W 25 . Match total score S, s=w 23 +W 25 Or s=w 24 +W 25 。
c) When the input work order task "nuclear power plant" field belongs to the CANDU6 model, the matching principle is as follows:
the field of 'equipment coding' is normal, the equipment codes are all consistent (same equipment), the weight value is r, and the score is W 26 The method comprises the steps of carrying out a first treatment on the surface of the The equipment codes except the first bit set number are all the same, the weight value is r, and the score is W 27 The method comprises the steps of carrying out a first treatment on the surface of the In the equipment coding, except for the first bit unit number, the system codes are the same, the weight value is r, and the score is W 28 The method comprises the steps of carrying out a first treatment on the surface of the The 'worksheet title' fields are matched through semantic similarity, the weight value is s, and the score is W 29 . Match total score S, s=w 26 +W 29 Or s=w 27 +W 29 Or s=w 28 +W 29 。
The device code field is abnormal and the work order header has positiveIn the case of a normal device code, the device code of the "work order header" field needs to be extracted. The device codes extracted from the 'work order title' field are all consistent (same device), the weight value is t, and the score is W 30 The method comprises the steps of carrying out a first treatment on the surface of the The equipment codes extracted from the 'work sheet title' field are identical except for the first bit unit number, the weight value is t, and the score is W 31 The method comprises the steps of carrying out a first treatment on the surface of the In the equipment codes extracted from the 'work sheet title' field, if the system codes are the same except the first bit unit number, the weight value is t and the score is W 32 The method comprises the steps of carrying out a first treatment on the surface of the Other contents of the 'work order title' field are matched through semantic similarity, the weight value is u, and the score is W 33 . Match total score S, s=w 30 +W 33 Or s=w 31 +W 33 Or s=w 32 +W 33 。
When the field of the 'equipment code' is abnormal and the 'work order title' has no normal equipment code, the field of the 'equipment name' is matched according to the semantic similarity, the weight value is v, and the score is W 34 The method comprises the steps of carrying out a first treatment on the surface of the The 'worksheet title' fields are matched through semantic similarity, the weight value is W, and the score is W 35 . Match total score S, s=w 34 + W 35 。
When the field of the 'equipment code' is abnormal and the 'work order title' has no normal equipment code, the field of the 'equipment name' cannot be matched, and only the semantic similarity of the field of the 'work order title' is considered. The weight value is x, and the score is W 36 . Match total score S, s=w 36 。
d) When the input work order task "nuclear power plant" field belongs to the VVER model, the matching principle is as follows:
The field of 'equipment coding' is normal, the equipment codes are all consistent (same equipment), the weight value is y, and the score is W 37 The method comprises the steps of carrying out a first treatment on the surface of the The device codes except the first bit set number are all the same, the weight value is y, and the score is W 38 The method comprises the steps of carrying out a first treatment on the surface of the The number of the first bit set is different from the last number in the equipment coding, the weight value is y, and the score is W 39 The method comprises the steps of carrying out a first treatment on the surface of the The numbers except the first bit set number and the middle number in the equipment code are different and are the same, the weight value is y, and the score is W 40 The method comprises the steps of carrying out a first treatment on the surface of the The device code is different from the first digit group number, the middle digit and the last digit, and the weight value is y and the score is W 41 The method comprises the steps of carrying out a first treatment on the surface of the The 'worksheet title' fields are matched through semantic similarity, the weight value is z, and the score is W 42 . Match total score S, s=w 37 +W 42 Or s=w 38 + W 42 Or s=w 39 +W 42 Or s=w 40 +W 42 Or s=w 41 +W 42 。
When the field of the "equipment code" is abnormal and the normal equipment code exists in the "work order title", the equipment code of the field of the "work order title" needs to be extracted. The device codes extracted from the 'work order title' field are all consistent (same device), and the weight value is a 1 Score of W 43 The method comprises the steps of carrying out a first treatment on the surface of the The equipment codes extracted from the 'work sheet title' field are identical except for the first bit set number, and the weight value is a 1 Score of W 44 The method comprises the steps of carrying out a first treatment on the surface of the The equipment codes extracted from the 'work sheet title' field are identical except that the first bit set number and the last number are different, and the weight value is a 1 Score of W 45 The method comprises the steps of carrying out a first treatment on the surface of the The equipment codes extracted from the 'work sheet title' field are identical except that the number of the first bit set and the number in the middle are different, and the weight value is a 1 Score of W 46 The method comprises the steps of carrying out a first treatment on the surface of the The equipment codes extracted from the 'work sheet title' field are different from the first digit group number, the middle digit and the last digit, and the other digits are the same, and the weight value is a 1 Score of W 47 The method comprises the steps of carrying out a first treatment on the surface of the The word content of the 'work order title' field (excluding the device code) is matched through semantic similarity, and the weight value is b 1 Score of W 48 . Match total score S, s=w 43 +W 48 Or s=w 44 +W 48 Or s=w 45 +W 48 Or s=w 46 +W 48 Or s=w 47 +W 48 。
When the field of the 'equipment code' is abnormal and the 'work order title' has no normal equipment code, the field of the 'equipment name' is matched according to semantic similarity, and the weight value is c 1 Score of W 49 ；“Matching is carried out in the work order title through semantic similarity, and the weight value is d 1 Score of W 50 . Match total score S, s=w 49 + W 50 。
When the field of the 'equipment code' is abnormal, the 'work order title' has no normal equipment code, and the 'equipment name' field cannot be matched, only the semantic similarity matching of the 'work order title' is considered, and the weight value is e 1 Score of W 51 . Match total score S, s=w 51 。
e) When the input work order task 'nuclear power plant' field belongs to an AP1000 model, the matching principle is as follows:
the field of 'device code' is normal, the device codes are all consistent (same device), and the weight value is f 1 Score of W 52 The method comprises the steps of carrying out a first treatment on the surface of the The weight value is f except the first bit set number in the equipment coding 1 Score of W 53 The method comprises the steps of carrying out a first treatment on the surface of the The device codes are identical except that the first bit set number and the last serial number are different, and the weight value is f 1 Score of W 54 The method comprises the steps of carrying out a first treatment on the surface of the The 'worksheet title' fields are matched through semantic similarity, and the weight value is g 1 Score of W 55 . Match total score S, s=w 52 +W 55 Or s=w 53 +W 55 Or s=w 54 +W 55 。
When the equipment code field is abnormal and the equipment code in the "work order title" is normal, the equipment code in the "work order title" field needs to be extracted. The device codes extracted from the 'worksheet title' field are all consistent (same device), and the weight value is h 1 Score of W 56 The method comprises the steps of carrying out a first treatment on the surface of the The equipment codes extracted from the 'worksheet title' field are identical except for the first bit set number, and the weight value is h 1 Score of W 57 The method comprises the steps of carrying out a first treatment on the surface of the The equipment code extracted from the 'work order title' field is divided into a first bit unit number and a last serial number, and the weight value is h 1 Score of W 58 The method comprises the steps of carrying out a first treatment on the surface of the Other contents of the 'work order title' field are matched through semantic similarity, and the weight value is i 1 Score of W 59 . Match total score S, s=w 56 +W 59 Or s=w 57 +W 59 Or s=w 58 +W 59 。
When the field of the 'equipment code' is abnormal and the 'work order title' has no normal equipment code, the field of the 'equipment name' is matched according to semantic similarity, and the weight value is j 1 Score of W 60 The method comprises the steps of carrying out a first treatment on the surface of the The "worksheet header" field is matched by semantic similarity, and the weight value is k 1 Score of W 61 . Match total score S, s=w 60 + W 61 。
When the field of the 'equipment code' is abnormal, the 'work order title' has no normal equipment code, and the 'equipment name' field cannot be matched, only the semantic similarity of the 'work order title' is considered, and the weight value is l 1 Score of W 62 . Match total score S, s=w 62 。
According to the rule, the output result only pushes the total score to be given the relevant score w Limiting the score And the above fraction data (i.e., S.gtoreq.w Limiting the score ) Count into set S Score set 。
2) Weight value correction score
The job order task S in 1) is paired according to the following rule Score set The result in (2) is subjected to weight value correction, and the correction of the weight influences the sorting of the results.
(1) In all matching results, weight adjustment is required according to the "nuclear power plant" and "machine group number" fields of the work order task.
a) If the fields of the nuclear power plant and the machine group number are identical, the weight correction value is X 1 ；
b) If the "nuclear power plant" fields are the same and the "machine group number" fields are different, the weight correction value is X 2 ；
c) If the fields of the nuclear power plant and the unit number are different, but the pile types are the same, the weight correction value is X 3 。
d) The weight correction value X Correction 1 ＝X 1 Or X 2 Or X 3 。
(2) When matched work order task is producedQuality defect report is generated, and the weight correction value is X Correction 2 。
(3) When the input work order task and the matched work order task have the large and small repair numbers (not necessarily the same) in the large and small repair number field, the weight correction value is X Correction 3 。
(4) When the input work order task and the "job type" field of the matched work order task belong to the periodic test and operation task. If the "work completion details" field in the finished report corresponding to the matched work order task has a description of "disqualification", the weight correction value is X Correction 4 。
(5) Comparing the fields of the system code and the system name of the input work order task with the fields of the system code and the system name of the push work order task, and if one of the fields of the system code and the system name is identical, the weight correction value is X Correction 5 。
(6) If the field of the 'work order type' in the pushed work order task is the work order of the defect class, the weight correction value is X Correction 6 。
And (3) outputting: according to the matching rule, each S of the input work order task and the matched work order task Score set And the corresponding correction coefficients are recommended according to the order of the scores from high to low.
Step 2: if the value of the field of the 'work object' in the work order task is 'S-system', establishing an accurate matching rule;
the precise matching rule is as follows:
if the input work order task can be matched with the "system number" or "system name" field of the matched work order task. The "System number" field is completely matched, where the weight value is m 1 Score of W 63 The method comprises the steps of carrying out a first treatment on the surface of the The "system number" field is not matched, and then the "system name" field is matched, and the "system name" field is matched according to semantic similarity, where the weight value is m 1 Score of W 64 The method comprises the steps of carrying out a first treatment on the surface of the Semantic similarity matching is carried out on the 'work order title' field of the work order task, and the method is rightWeight value n 1 Score of W 65 . Match total score S, s=w 63 +W 65 Or s=w 64 +W 65 。
If the "system number" and "system name" fields of the entered work order task and the matched work order task cannot be matched. The "work order header" field of the work order task performs semantic similarity matching, where the weight value is o 1 Score of W 66 . Match total score S, s=w 66 。
According to the rule, the output result only pushes the total score to be given the relevant score w Limiting the score And the above fraction data (i.e., S.gtoreq.w Limiting the score ) Count into set S Score set 。
And (3) outputting: according to the matching rule, each S of the input work order task and the matched work order task Score set And recommending according to the order of the scores from high to low.
Step 3: if the value of the field of the 'work object' in the work order task is 'L-position', establishing an accurate matching rule;
the precise matching rule is as follows: if the "room number" or "structure" field of the entered work order task and the matched work order task can be matched. The "room number" field is fully matched, where the weight value is p 1 Score of W 67 The method comprises the steps of carrying out a first treatment on the surface of the The "room number" field is not matched, then the "structure" field is unmatched, and the "structure" field is completely matched, where the weight value is q 1 Score of W 68 The method comprises the steps of carrying out a first treatment on the surface of the The "work order header" field of the work order task is subjected to semantic similarity matching, and the weight value is r 1 Score of W 69 . Match total score S, s=w 67 +W 69 Or s=w 68 +W 69 。
If the input work order task can not be matched with the room number and structure fields of the matched work order task, the semantic similarity matching is carried out on the work order title field of the work order task, and the weight value is s 1 Score of W 70 . Match total score S, s=w 70 。
According to the rule, the output result only pushes the total score to be given the relevant score w Limiting the score And the above fraction data (i.e., S.gtoreq.w Limiting the score ) Count into set S Score set 。
And (3) outputting: according to the matching rule, each S of the input work order task and the matched work order task Score set And recommending according to the order of the scores from high to low.
Step 4: if the values of the field of the 'work object' in the work order task are 'N-name', 'U-UIC', 'M-material', establishing an accurate matching rule;
the precise matching rule is as follows:
the "work order header" field of the work order task is subjected to semantic similarity matching, and the weight value is t 1 Score of W 71 . Match total score S, s=w 71 。
According to the rule, the output result only pushes the total score to be given the relevant score w Limiting the score And the above fraction data (i.e., S.gtoreq.w Limiting the score ) Count into set S Score set 。
And (3) outputting: according to the matching rule, each S of the input work order task and the matched work order task Score set And recommending according to the order of the scores from high to low.

Claims (10)

1. The accurate matching method applied to the work order task of the nuclear power plant is characterized by comprising the following steps of:
Step 1: if the value of the field of the 'work object' in the work order task is 'E-equipment', establishing an accurate matching rule;
step 2: if the value of the field of the 'work object' in the work order task is 'S-system', establishing an accurate matching rule;
step 3: if the value of the field of the 'work object' in the work order task is 'L-position', establishing an accurate matching rule;
step 4: if the values of the field of the 'work object' in the work order task are 'N-name', 'U-UIC', 'M-material', a precise matching rule is established.
2. The method for precisely matching the work order task of the nuclear power plant according to claim 1, wherein the precisely matching rule comprises the following steps:
1) Constructing business logic rules and semantic similarity rules;
2) Weight value correction score
The job order task S in 1) is paired according to the following rule Score set The result in (2) is subjected to weight value correction, and the correction of the weight influences the sorting of the result;
(1) In all matching results, weight adjustment is required according to the fields of the nuclear power plant and the machine set number of the work order task;
a) If the fields of the nuclear power plant and the machine group number are identical, the weight correction value is X 1 ；
b) If the "nuclear power plant" fields are the same and the "machine group number" fields are different, the weight correction value is X 2 ；
c) If the fields of the nuclear power plant and the unit number are different, but the pile types are the same, the weight correction value is X 3 ；
d) The weight correction value X Correction 1 ＝X 1 Or X 2 Or X 3 ；
(2) When quality defect report is generated in matched work order task, the weight correction value is X Correction 2 ；
(3) When the input work order task and the matched work order task have the large and small repair numbers in the large and small repair number field, the weight correction value is X Correction 3 ；
(4) When the input work order task and the job type field of the matched work order task belong to the periodic test and operation task, if the description of disqualification exists in the work completion detailed field in the finished report corresponding to the matched work order task, the weight correction value is X Correction 4 ；
(5) System coding of input work order tasksComparing the "and" system name "field with the" system code "and" system name "fields of the push work order task, if one of the" system code "and" system name "fields is identical, the weight correction value is X Correction 5 ；
(6) If the field of the 'work order type' in the pushed work order task is the work order of the defect class, the weight correction value is X Correction 6 ；
And (3) outputting: according to the matching rule, each S of the input work order task and the matched work order task Score set And the corresponding correction coefficients are recommended according to the order of the scores from high to low.
3. The method for precisely matching the work order task of the nuclear power plant according to claim 2, wherein the construction of the business logic rule and the construction of the semantic similarity rule comprise the following steps:
if the input "nuclear power plant" field of the work order task and the "nuclear power plant" field of the inquired work order task do not belong to the same nuclear power plant, the matching principle is as follows:
if the "nuclear power plant" field in the input work order task belongs to CANDU6, CNP-300, AP1000 and VVER model, matching the "work order title" and "equipment name" fields according to semantic similarity when matching other heap types, wherein the weight value is a and the score is W 1 Match total score S, s=w 1 ；
If the 'nuclear power plant' field belongs to CNP-600, CNP-1000, HPR-1000 and CPR-1000 types in the input work order task, matching the 'work order title' and 'equipment name' fields according to semantic similarity when matching the work order task of CANDU6, CNP-300, AP1000 and VVER types, wherein the weight value is b and the score is W 2 Match total score S, s=w 2 ；
If the input work order task 'nuclear power plant' field belongs to CNP-600, CNP-1000, HPR-1000 and CPR-1000 models, when matching the work order tasks of CNP-600, CNP-1000, HPR-1000 and CPR-1000 models, the following principle needs to be considered:
for CNP-600, CNP-1000, HPR-1000 and CPR-1000 models, if the field of 'equipment coding' is normal, the equipment codes are all consistent, the weight value is c, and the score is W 3 The method comprises the steps of carrying out a first treatment on the surface of the The equipment codes except the first bit set number are all the same, the weight value is c, and the score is W 4 The method comprises the steps of carrying out a first treatment on the surface of the The equipment codes are identical except that the first bit computer group number and the intermediate digital code are different, the weight value is c, and the score is W 5 The method comprises the steps of carrying out a first treatment on the surface of the The 'worksheet title' fields are matched through semantic similarity, the weight value is d, and the score is W 6 Match total score S, s=w 3 +W 6 Or s=w 4 +W 6 Or s=w 5 +W 6 ；
For CNP-600, CNP-1000, HPR-1000 and CPR-1000 models, if the field of 'equipment code' is abnormal, equipment code is extracted from the field of 'work order title'; if normal equipment codes can be extracted from the field of the 'work order title', the equipment codes are all consistent, the weight value is e, and the score is W 7 The method comprises the steps of carrying out a first treatment on the surface of the The equipment codes except the first bit set number are all the same, the weight value is e, and the score is W 8 The method comprises the steps of carrying out a first treatment on the surface of the The equipment codes are identical except that the first bit computer group number and the intermediate digital code are different, the weight value is e, and the score is W 9 The method comprises the steps of carrying out a first treatment on the surface of the The 'worksheet title' fields are matched through semantic similarity, the weight value is f, and the score is W 10 Match total score S, s=w 7 +W 10 Or s=w 8 +W 10 Or s=w 9 +W 10 ；
For CNP-600, CNP-1000, HPR-1000 and CPR-1000 models, if the device code and the work order title field have no normal device code, the work order title and the device name fields are matched according to semantic similarity, the weight value is g, and the score is W 11 Total score of matching s=w 11 ；
According to the rule, the output result only pushes the total score to be given the relevant score w Limiting the score And the above fraction data (i.e., S.gtoreq.w Limiting the score ) Count into set S Score set 。
4. The method for precisely matching the work order task of the nuclear power plant according to claim 2, wherein the construction of the business logic rule and the construction of the semantic similarity rule comprise the following steps:
if the input "nuclear power plant" field of the work order task and the queried "nuclear power plant" field belong to the same nuclear power plant, when the "equipment code" and the "work order title" and "equipment name" fields are considered, the following principle is considered:
Considering the field of 'equipment coding', if the field of 'equipment coding' is normal, the equipment coding in the field of 'work order title' is not required to be extracted;
the equipment code field is abnormal, but when the normal equipment code exists in the 'work order title', the equipment code of the 'work order title' field needs to be extracted;
when the equipment coding field is abnormal and the equipment coding is not normal in the 'work order title', only the semantic similarity of the 'work order title' and the 'equipment name' is considered.
5. The method for exact matching of a work order task in a nuclear power plant according to claim 4, further comprising:
when the input work order task 'nuclear power plant' field belongs to CNP-600, CNP-1000, HPR-1000 and CPR-1000 models, the matching principle is as follows:
when the 'equipment coding' field is normal, the equipment codes are all consistent, the weight value is h, and the score is W 12 The method comprises the steps of carrying out a first treatment on the surface of the The equipment codes except the first bit set number are all the same, the weight value is h, and the score is W 13 The method comprises the steps of carrying out a first treatment on the surface of the The equipment codes are identical except that the first bit computer group number and the intermediate digital code are different, the weight value is h, and the score is W 14 The method comprises the steps of carrying out a first treatment on the surface of the The 'worksheet title' fields are matched through semantic similarity, the weight value is i, and the score is W 15 Match total score S, s=w 12 +W 10 Or s=w 13 +W 10 Or s=w 14 +W 15 ；
The "equipment code" field is abnormal and there is a normal in the "work order headingWhen the equipment codes are used, the equipment codes for extracting the 'work order title' fields are required to be identical, the weight value is j, and the score is W 16 The method comprises the steps of carrying out a first treatment on the surface of the The equipment codes extracted from the 'work order title' field are identical except for the first bit unit number, the weight value is j, and the score is W 17 The method comprises the steps of carrying out a first treatment on the surface of the The equipment codes extracted from the 'work sheet title' field are identical except that the first bit machine group number and the intermediate digital code are different, the weight value is j, and the score is W 18 The method comprises the steps of carrying out a first treatment on the surface of the Other contents of the 'work order title' field are matched through semantic similarity, the weight value is k, and the score is W 19 Match total score S, s=w 16 +W 19 Or s=w 17 +W 19 Or s=w 18 +W 19 ；
When the field of the 'equipment code' is abnormal and the 'work order title' field is not provided with normal equipment codes, the field of the 'equipment name' is matched according to semantic similarity, the weight value is l, and the score is W 20 The 'worksheet title' fields are matched through semantic similarity, the weight value is m, and the score is W 21 Match total score S, s=w 20 +W 21 ；
When the field of the 'equipment code' is abnormal, and the 'work order title' has no normal equipment code, and the 'equipment name' field cannot be matched, the weight value is n, and the score is W 22 Match total score S, s=w 22 ；
When the input work order task 'nuclear power plant' field belongs to the CNP-300 model, the matching principle is as follows:
the equipment codes are all consistent, the weight value is o, and the score is W 23 The method comprises the steps of carrying out a first treatment on the surface of the The device codes are not matched, the device names are matched according to the semantic similarity, the weight value is p, and the score is W 24 The method comprises the steps of carrying out a first treatment on the surface of the The 'worksheet title' fields are matched according to semantic similarity, the weight value is q, and the score is W 25 The method comprises the steps of carrying out a first treatment on the surface of the Match total score S, s=w 23 +W 25 Or s=w 24 +W 25 ；
c) When the input work order task "nuclear power plant" field belongs to the CANDU6 model, the matching principle is as follows:
the field of 'equipment codes' is normal, the equipment codes are all consistent, the weight value is r, and the score is W 26 The method comprises the steps of carrying out a first treatment on the surface of the The equipment codes except the first bit set number are all the same, the weight value is r, and the score is W 27 The method comprises the steps of carrying out a first treatment on the surface of the In the equipment coding, except for the first bit unit number, the system codes are the same, the weight value is r, and the score is W 28 The method comprises the steps of carrying out a first treatment on the surface of the The 'worksheet title' fields are matched through semantic similarity, the weight value is s, and the score is W 29 Match total score S, s=w 26 +W 29 Or s=w 27 +W 29 Or s=w 28 +W 29 ；
When the equipment code field is abnormal and the normal equipment code exists in the 'work order title', the equipment code of the 'work order title' field needs to be extracted, the equipment codes extracted by the 'work order title' field are all consistent, the weight value is t, and the score is W 30 The method comprises the steps of carrying out a first treatment on the surface of the The equipment codes extracted from the 'work sheet title' field are identical except for the first bit unit number, the weight value is t, and the score is W 31 The method comprises the steps of carrying out a first treatment on the surface of the In the equipment codes extracted from the 'work sheet title' field, if the system codes are the same except the first bit unit number, the weight value is t and the score is W 32 The method comprises the steps of carrying out a first treatment on the surface of the Other contents of the 'work order title' field are matched through semantic similarity, the weight value is u, and the score is W 33 Match total score S, s=w 30 +W 33 Or s=w 31 +W 33 Or s=w 32 +W 33 ；
When the field of the 'equipment code' is abnormal and the 'work order title' has no normal equipment code, the field of the 'equipment name' is matched according to the semantic similarity, the weight value is v, and the score is W 34 The method comprises the steps of carrying out a first treatment on the surface of the The 'worksheet title' fields are matched through semantic similarity, the weight value is W, and the score is W 35 The method comprises the steps of carrying out a first treatment on the surface of the Match total score S, s=w 34 +W 35 ；
When the field of the 'equipment code' is abnormal and the 'work order title' is not provided with normal equipment code, and the field of the 'equipment name' cannot be matched, only the semantic similarity of the field of the 'work order title' is considered, the weight value is x, and the score is W 36 Match total score S, s=w 36 ；
When the input work order task "nuclear power plant" field belongs to the VVER model, the matching principle is as follows:
The field of 'equipment coding' is normal, the equipment codes are all consistent, the weight value is y, and the score is W 37 The method comprises the steps of carrying out a first treatment on the surface of the The device codes except the first bit set number are all the same, the weight value is y, and the score is W 38 The method comprises the steps of carrying out a first treatment on the surface of the The number of the first bit set is different from the last number in the equipment coding, the weight value is y, and the score is W 39 The method comprises the steps of carrying out a first treatment on the surface of the The numbers except the first bit set number and the middle number in the equipment code are different and are the same, the weight value is y, and the score is W 40 The method comprises the steps of carrying out a first treatment on the surface of the The device code is different from the first digit group number, the middle digit and the last digit, and the weight value is y and the score is W 41 The method comprises the steps of carrying out a first treatment on the surface of the The 'worksheet title' fields are matched through semantic similarity, the weight value is z, and the score is W 42 Match total score S, s=w 37 +W 42 Or s=w 38 +W 42 Or s=w 39 +W 42 Or s=w 40 +W 42 Or s=w 41 +W 42 ；
When the field of the 'equipment code' is abnormal and the normal equipment code exists in the 'work order title', the equipment code of the 'work order title' field needs to be extracted, the equipment codes extracted from the 'work order title' field are all consistent, and the weight value is a 1 Score of W 43 The method comprises the steps of carrying out a first treatment on the surface of the The equipment codes extracted from the 'work sheet title' field are identical except for the first bit set number, and the weight value is a 1 Score of W 44 The method comprises the steps of carrying out a first treatment on the surface of the The equipment codes extracted from the 'work sheet title' field are identical except that the first bit set number and the last number are different, and the weight value is a 1 Score of W 45 The method comprises the steps of carrying out a first treatment on the surface of the The equipment codes extracted from the 'work sheet title' field are identical except that the number of the first bit set and the number in the middle are different, and the weight value is a 1 Score of W 46 The method comprises the steps of carrying out a first treatment on the surface of the The equipment codes extracted from the 'work sheet title' field are different from the first digit group number, the middle digit and the last digit, and the other digits are the same, and the weight value is a 1 Score of W 47 The method comprises the steps of carrying out a first treatment on the surface of the The word contents of the 'work order title' field are matched through semantic similarity, and the weight value is b 1 Score of W 48 Match total score S, s=w 43 +W 48 Or s=w 44 +W 48 Or s=w 45 +W 48 Or s=w 46 +W 48 Or s=w 47 +W 48 ；
When the field of the 'equipment code' is abnormal and the 'work order title' has no normal equipment code, the field of the 'equipment name' is matched according to semantic similarity, and the weight value is c 1 Score of W 49 The method comprises the steps of carrying out a first treatment on the surface of the Matching the "work order titles" through semantic similarity, wherein the weight value is d 1 Score of W 50 Match total score S, s=w 49 +W 50 ；
When the field of the 'equipment code' is abnormal, the 'work order title' has no normal equipment code, and the 'equipment name' field cannot be matched, only the semantic similarity matching of the 'work order title' is considered, and the weight value is e 1 Score of W 51 Match total score S, s=w 51 ；
e) When the input work order task 'nuclear power plant' field belongs to an AP1000 model, the matching principle is as follows:
the field of 'equipment coding' is normal, the equipment codes are all consistent, and the weight value is f 1 Score of W 52 The method comprises the steps of carrying out a first treatment on the surface of the The weight value is f except the first bit set number in the equipment coding 1 Score of W 53 The method comprises the steps of carrying out a first treatment on the surface of the The device codes are identical except that the first bit set number and the last serial number are different, and the weight value is f 1 Score of W 54 The method comprises the steps of carrying out a first treatment on the surface of the The 'worksheet title' fields are matched through semantic similarity, and the weight value is g 1 Score of W 55 Match total score S, s=w 52 +W 55 Or s=w 53 +W 55 Or s=w 54 +W 55 ；
When the equipment code field is abnormal and the normal equipment code exists in the 'work order title', the equipment code in the 'work order title' field needs to be extracted, and the 'work order title' field is extractedThe codes of the fetched devices are all consistent, and the weight value is h 1 Score of W 56 The method comprises the steps of carrying out a first treatment on the surface of the The equipment codes extracted from the 'worksheet title' field are identical except for the first bit set number, and the weight value is h 1 Score of W 57 The method comprises the steps of carrying out a first treatment on the surface of the The equipment code extracted from the 'work order title' field is divided into a first bit unit number and a last serial number, and the weight value is h 1 Score of W 58 The method comprises the steps of carrying out a first treatment on the surface of the Other contents of the 'work order title' field are matched through semantic similarity, and the weight value is i 1 Score of W 59 The method comprises the steps of carrying out a first treatment on the surface of the Match total score S, s=w 56 +W 59 Or s=w 57 +W 59 Or s=w 58 +W 59 ；
When the field of the 'equipment code' is abnormal and the 'work order title' has no normal equipment code, the field of the 'equipment name' is matched according to semantic similarity, and the weight value is j 1 Score of W 60 The method comprises the steps of carrying out a first treatment on the surface of the The "worksheet header" field is matched by semantic similarity, and the weight value is k 1 Score of W 61 The method comprises the steps of carrying out a first treatment on the surface of the Match total score S, s=w 60 +W 61 ；
When the field of the 'equipment code' is abnormal, the 'work order title' has no normal equipment code, and the 'equipment name' field cannot be matched, only the semantic similarity of the 'work order title' is considered, and the weight value is l 1 Score of W 62 The method comprises the steps of carrying out a first treatment on the surface of the Match total score S, s=w 62 ；
According to the rule, the output result only pushes the total score to be given the relevant score w Limiting the score And the above fraction data (i.e., S.gtoreq.w Limiting the score ) Count into set S Score set 。
6. The method for precisely matching the work order task of the nuclear power plant according to claim 1, wherein the rule for precisely matching in the step 2 is as follows:
If the input work order task and the system number or system name field of the matched work order task can be matched; the "System number" field is fully matched, where the weight value ism 1 Score of W 63 The method comprises the steps of carrying out a first treatment on the surface of the The "system number" field is not matched, and then the "system name" field is matched, and the "system name" field is matched according to semantic similarity, where the weight value is m 1 Score of W 64 The method comprises the steps of carrying out a first treatment on the surface of the The "work order header" field of the work order task is subjected to semantic similarity matching, and the weight value is n 1 Score of W 65 Match total score S, s=w 63 +W 65 Or s=w 64 +W 65 ；
If the input work order task cannot be matched with the system number and system name fields of the matched work order task, the semantic similarity matching is carried out on the work order title fields of the work order task, and the weight value is o 1 Score of W 66 Match total score S, s=w 66 ；
According to the rule, the output result only pushes the total score to be given the relevant score w Limiting the score And the above fraction data (i.e., S.gtoreq.w Limiting the score ) Count into set S Score set ；
And (3) outputting: according to the matching rule, each S of the input work order task and the matched work order task Score set And recommending according to the order of the scores from high to low.
7. The method for precisely matching the work order task of the nuclear power plant according to claim 1, wherein the rule of precisely matching in the step 3 is as follows:
the precise matching rule is as follows: if the input work order task can be matched with the room number or structure field of the matched work order task, the room number field is completely matched, and the weight value is p 1 Score of W 67 The method comprises the steps of carrying out a first treatment on the surface of the The "room number" field is not matched, then the "structure" field is unmatched, and the "structure" field is completely matched, where the weight value is q 1 Score of W 68 The method comprises the steps of carrying out a first treatment on the surface of the The "work order header" field of the work order task is subjected to semantic similarity matching, and the weight value is r 1 Score of W 69 Match total score S, s=W 67 +W 69 Or s=w 68 +W 69 ；
If the input work order task can not be matched with the room number and structure fields of the matched work order task, the semantic similarity matching is carried out on the work order title field of the work order task, and the weight value is s 1 Score of W 70 Match total score S, s=w 70 ；
According to the rule, the output result only pushes the total score to be given the relevant score w Limiting the score And the above fraction data (i.e., S.gtoreq.w Limiting the score ) Count into set S Score set ；
And (3) outputting: according to the matching rule, each S of the input work order task and the matched work order task Score set And recommending according to the order of the scores from high to low.
8. The method for precisely matching the work order task of the nuclear power plant according to claim 1, wherein the rule of precisely matching in the step 4 is as follows:
the precise matching rule is as follows:
the "work order header" field of the work order task is subjected to semantic similarity matching, and the weight value is t 1 Score of W 71 Match total score S, s=w 71 。
9. The method of claim 8, wherein the step 4 comprises
According to the rule, the output result only pushes the total score to be given relative score w Limiting the score And the above fraction data (i.e., S.gtoreq.w Limiting the score ) Count into set S Score set 。
10. The method of claim 8, wherein in step 4,
and (3) outputting: according toThe matching rule above, each S of the input work order task and the matched work order task Score set And recommending according to the order of the scores from high to low.
CN202111349953.5A 2021-11-15 2021-11-15 Accurate matching method applied to nuclear power plant work order task Pending CN116151524A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202111349953.5A CN116151524A (en) 2021-11-15 2021-11-15 Accurate matching method applied to nuclear power plant work order task

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202111349953.5A CN116151524A (en) 2021-11-15 2021-11-15 Accurate matching method applied to nuclear power plant work order task

Publications (1)

Publication Number Publication Date
CN116151524A true CN116151524A (en) 2023-05-23

Family

ID=86339352

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202111349953.5A Pending CN116151524A (en) 2021-11-15 2021-11-15 Accurate matching method applied to nuclear power plant work order task

Country Status (1)

Country Link
CN (1) CN116151524A (en)

Similar Documents

Publication Publication Date Title
US20230109993A1 (en) Industrial internet of things system for controlling production line parameter and control methods thereof
CN111461668A (en) Digital auditing system and method based on process automation technology
CN104572895B (en) MPP databases and Hadoop company-datas interoperability methods, instrument and implementation method
CN104778540A (en) BOM (bill of material) management method and management system for building material equipment manufacturing
CN105405069B (en) Electricity purchase operation decision analysis and data processing method
CN107067170A (en) A kind of cost accounting and quotation system based on assembling decomposition texture
CN104281525B (en) A kind of defect data analysis method and the method utilizing its reduction Software Testing Project
CN102999524A (en) Method and system for searching document association
CN104063519A (en) BPA power grid data analyzing and managing method and system based on EXCEL
CN114462737A (en) Accurate matching method applied to nuclear power plant work order task and operation event report
CN114462736A (en) Experience feedback intelligent recommendation method for nuclear power plant radiation work license application
CN113535818A (en) Method and equipment for constructing audit comprehensive knowledge base
CN116151524A (en) Accurate matching method applied to nuclear power plant work order task
CN112508381A (en) Computing system for project cost of mechanical industry and implementation method thereof
CN109408788A (en) A kind of text marking method towards judgement document
CN109033452A (en) A kind of data warehouse is intelligent to construct stowage and system
CN113409001A (en) Method for controlling automatic pricing of construction engineering quantity list
CN115759562A (en) Logic relation and construction period calculation algorithm of preventive maintenance work order for overhaul of nuclear power plant
CN111538760A (en) Method for establishing distribution load line loss correlation analysis model based on Apriori algorithm
CN111178699A (en) Method for constructing intelligent checking system of scheduling operation tickets
CN113064924A (en) Nuclear power big data experience retrieval and pushing method
CN112801537B (en) Non-stop accident analysis method, system, medium and equipment for power generation enterprise
CN113204864A (en) Method and system suitable for transmission network simulation data version control
CN115080687A (en) Accurate matching method for nuclear power plant work order tasks and quality defect report information
CN116957369A (en) Nuclear power plant quality defect report and production non-conforming item accurate matching method

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