CN117744965A - Equipment abnormality control method and system - Google Patents
Equipment abnormality control method and system Download PDFInfo
- Publication number
- CN117744965A CN117744965A CN202311302053.4A CN202311302053A CN117744965A CN 117744965 A CN117744965 A CN 117744965A CN 202311302053 A CN202311302053 A CN 202311302053A CN 117744965 A CN117744965 A CN 117744965A
- Authority
- CN
- China
- Prior art keywords
- question
- answer
- index
- abnormal
- library
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 92
- 230000005856 abnormality Effects 0.000 title claims abstract description 23
- 230000002159 abnormal effect Effects 0.000 claims abstract description 222
- 238000012544 monitoring process Methods 0.000 claims abstract description 164
- 238000005516 engineering process Methods 0.000 claims abstract description 125
- 238000004458 analytical method Methods 0.000 claims abstract description 73
- 238000007726 management method Methods 0.000 claims abstract description 68
- 230000008569 process Effects 0.000 claims abstract description 52
- 238000001994 activation Methods 0.000 claims description 60
- 230000004913 activation Effects 0.000 claims description 56
- 238000011282 treatment Methods 0.000 claims description 42
- 238000011160 research Methods 0.000 claims description 25
- 230000011218 segmentation Effects 0.000 claims description 24
- 230000006870 function Effects 0.000 claims description 20
- 238000012986 modification Methods 0.000 claims description 18
- 230000004048 modification Effects 0.000 claims description 18
- 238000000605 extraction Methods 0.000 claims description 16
- 238000007689 inspection Methods 0.000 claims description 14
- 238000012550 audit Methods 0.000 claims description 12
- 230000003213 activating effect Effects 0.000 claims description 9
- 238000004590 computer program Methods 0.000 claims description 9
- 238000012549 training Methods 0.000 claims description 9
- 238000004886 process control Methods 0.000 claims description 7
- 238000004364 calculation method Methods 0.000 claims description 5
- 239000000284 extract Substances 0.000 claims description 3
- 230000000717 retained effect Effects 0.000 claims description 3
- 238000012800 visualization Methods 0.000 claims description 3
- 238000010276 construction Methods 0.000 abstract description 5
- 238000012795 verification Methods 0.000 description 32
- 230000003287 optical effect Effects 0.000 description 15
- 230000014509 gene expression Effects 0.000 description 13
- 239000013589 supplement Substances 0.000 description 12
- 238000012552 review Methods 0.000 description 6
- 230000009469 supplementation Effects 0.000 description 6
- 230000008859 change Effects 0.000 description 4
- 238000009795 derivation Methods 0.000 description 3
- 238000011161 development Methods 0.000 description 3
- 230000018109 developmental process Effects 0.000 description 3
- 230000003993 interaction Effects 0.000 description 3
- 238000012423 maintenance Methods 0.000 description 3
- 150000001875 compounds Chemical class 0.000 description 2
- 238000001514 detection method Methods 0.000 description 2
- 230000004927 fusion Effects 0.000 description 2
- 230000008520 organization Effects 0.000 description 2
- 230000002265 prevention Effects 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 239000000047 product Substances 0.000 description 2
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 230000015572 biosynthetic process Effects 0.000 description 1
- 238000013075 data extraction Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000008676 import Effects 0.000 description 1
- 230000006698 induction Effects 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 238000001556 precipitation Methods 0.000 description 1
- 230000001737 promoting effect Effects 0.000 description 1
- 238000001303 quality assessment method Methods 0.000 description 1
- 238000009877 rendering Methods 0.000 description 1
- 230000008439 repair process Effects 0.000 description 1
- 230000035945 sensitivity Effects 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
- 239000002699 waste material Substances 0.000 description 1
Landscapes
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention discloses a device abnormality control method and system. In order to solve the problems that the abnormal control of the power grid equipment in the prior art does not fully utilize huge historical abnormal resources or the utilization efficiency of the historical abnormal resources is low; according to the invention, abnormal events corresponding to the abnormal data of the historical equipment are analyzed, and a monitoring technology question-answering library, an abnormal event case library, a multiple disc analysis room and an abnormal event whole process management and control platform are respectively constructed; huge and complex historical equipment abnormal data since the power grid construction is fully utilized, and the utilization efficiency of the historical abnormal data can be greatly improved.
Description
Technical Field
The invention relates to the field of power grid equipment abnormality management and control, in particular to an equipment abnormality management and control method and system.
Background
As is well known, the power grid plays a vital role in life of common people, improving urban functions and promoting social and economic development, and the power grid development is an important technical means for optimizing configuration of energy and resources, but the power grid development is becoming a main threat of power grid safety increasingly due to large-area power failure caused by natural disasters, external damage, artificial responsibility, equipment failure in a weak system mode and the like.
Risk identification is a starting point and an important link of power grid security risk management and control. The risk identification work is to reasonably determine the risk prevention and control range, and the risk prevention and control range can be divided into an external risk and an internal risk according to the formation reasons. The external risks mainly comprise natural risks, external damage risks, artificial accident risks, electric power system reform risks and the like, and the internal risks mainly comprise power grid operation mode risks, electric load change risks, power grid equipment abnormality risks (including primary equipment risks and secondary equipment risks), power grid accident risks and the like. Many risks may also be combined from multiple reasons, complex and diverse.
The power grid distribution is wide, the historical operation data is complex and various, the expression form and the abnormality cause of the historical equipment abnormal data are complex and various, if the historical data can be fully utilized, the work experience is precipitated, the comprehensive and detailed auxiliary decision-making service is provided for commanders through an informatization technology, a convenient man-machine interaction mode is provided, the efficiency of abnormal control of the power grid equipment can be greatly improved, and resources are fully utilized. However, due to the huge amount and complexity of the historical data, it is difficult to sufficiently classify and integrate the historical data, and the historical data is not fully utilized, so that resource waste is caused. Or training the historical data by using a neural network to identify the abnormal state of the equipment in real time, but the method has the advantages of large training amount, large calculation amount and low utilization efficiency of resources.
For example, a method and apparatus for detecting an abnormality of a power grid device disclosed in chinese patent literature, which announces No. CN111242144a, the method includes: extracting features of the power grid equipment image to be detected by using a feature extraction network, and respectively inputting a plurality of feature images with different sizes output by the feature extraction network into each anchor point generation layer in the object detector; each anchor point generation layer in the object detector intercepts an input characteristic image according to a preset width and height size, and a detection head in the object detector calculates the probability value of an equipment abnormal region of each intercepted image; and the detection head sorts the calculated probability values from high to low, and the image of N before the probability value sorting is used as the image of the detected abnormal region of the power grid equipment. By the method and the device, the abnormal region in the power grid equipment image can be directly positioned.
According to the scheme, the abnormal state of the power grid equipment is detected through an image recognition technology, however, the process is large in calculation amount, huge historical equipment abnormal data of the power grid are not fully utilized, and the abnormal recognition efficiency is low.
Disclosure of Invention
The method mainly solves the problems that the abnormal control of the power grid equipment in the prior art does not fully utilize huge historical abnormal resources or the utilization efficiency of the historical abnormal resources is low; the equipment anomaly control method and system are characterized in that a monitoring technology question-answering library, an anomaly event case library, a multi-disc analysis room and an anomaly event overall process control platform are built, huge and complex historical equipment anomaly data from the construction of a power grid are fully utilized, and the utilization efficiency of the historical anomaly data can be greatly improved.
The technical problems of the invention are mainly solved by the following technical proposal:
an equipment abnormality management and control method comprises the following steps:
analyzing abnormal events corresponding to the abnormal data of the historical equipment, and respectively constructing a monitoring technology question-answering library, an abnormal event case library, a multiple disc analysis room and an abnormal event whole process management and control platform;
the abnormal event whole process control platform controls the whole life cycle of the abnormal event, including abnormal event activation, abnormal event treatment and abnormal event completion;
the monitoring technology question-answering library modifies and updates the index according to the equipment abnormal data; dividing the search term during search, comparing in a question-answer list, and returning an index result;
the abnormal event case library stores abnormal event cases, and the question-answer records in the monitoring technology question-answer library and the whole abnormal event process of the abnormal event whole process management and control platform are updated into the abnormal event case library;
and the multi-disc analysis room carries out multi-disc analysis on the cases in the case library and updates the monitoring technology question-answer library.
The monitoring technology question-answering library, the abnormal event case library, the multi-disc analysis chamber and the abnormal event whole process management and control platform are built, huge and complex historical equipment abnormal data since the power grid is built are fully utilized, and the utilization efficiency of the historical abnormal data can be greatly improved.
Preferably, the analysis of causes, consequences, pre-studies and treatments in the index in the monitoring technical question-and-answer library are stored separately and re-associated when the updated index is modified.
The main content of the monitoring technology question and answer is divided into a history version of the monitoring technology question and answer and a using version of the monitoring technology question and answer, the question and answer content inquired out in a page is the using version, the history version comprises the using version, each modification of the monitoring technology question and answer is to create a new question and answer record, the new question and answer record is confirmed by layer-by-layer verification and no repeated record in the history version is added into the history version, the new question and answer record is in a state to be checked at the moment, the new question and answer record is visible only in a checking list, the adding of a multi-disc analysis chamber firstly enters a draft state, the end of the multi-disc enters the state to be checked, and the follow-up monitoring technology question and answer flow comprises a series of verification.
Preferably, the process of modifying the update index is:
newly adding a monitoring technology question-answer index and selecting a monitoring technology question-answer type;
filling index content according to a preset standard format, and submitting an audit;
judging whether the newly added monitoring technology question-answer index is repeated with the index in the library, if so, checking is not passed, and returning the newly added monitoring technology question-answer index again; otherwise, the monitoring technology question-answer index is stored in the monitoring technology question-answer library.
After layer-by-layer verification, determining that no repeated record in the history version is added into the history version, and editing the original existing technical question-answer items by a user to replace old content and supplement fresh content. The editors and the editing time are saved together. The revised items are issued and visible after expert approval. The expert reviews the content of the new supplement and modification of the monitoring technology questions and answers, and the expert can directly modify the questions and answers. The monitoring technical questions and answers which do not pass the verification are returned to the corresponding modules, the expert performs the verification again after resubmitting, and after the verification passes, newly added or revised items appear in the monitoring question and answer library and are given a new version. The technical question and answer content of the new revision allows the expert user to fall back even after being released, the history version can be selected for fall back, and all the history versions can be inquired in the function.
Preferably, the monitoring technique question-answer type includes a typical accident information judgment analysis treatment, a typical abnormality information judgment analysis treatment, a typical out-of-limit information treatment, and a typical shift information treatment.
When the method is added, fields such as a question name, a question and answer keyword, information paraphrasing, associated signals, reason analysis, caused results, treatment points, reporting contents after on-site inspection, uploading pictures, uploading accessories and the like are required to be filled in, so that the technical question and answer of accurate positioning and paraphrasing monitoring are facilitated. The newly added monitoring technology questions and answers need expert auditing, and can be put into storage for storage after the auditing is passed.
Preferably, the index content is combined through a knowledge graph, and an initial triplet set is formed through knowledge extraction; model training is carried out on the initial triplet set, the extracted entity is aligned by utilizing the BERT model, a standard entity is obtained, and then the standard triplet set is formed; the standard triples are stored in the Neo4j graph database, and the correlations between knowledge are visually presented.
And combining different expressions with the same meaning into the same expression through the knowledge graph, and judging whether repeated contents exist or not through the constructed knowledge graph when checking whether repeated or not, so that repeated input of abnormal data of the same equipment in different expression forms is avoided.
Preferably, the re-association procedure is:
judging whether the single index contents exist in a monitoring technology question-answer library or not; if yes, entering the next step of judgment, otherwise, calculating the similarity between the index content combination updated by modification and all index content combinations in the monitoring technology question-answering library;
judging whether the modified and updated index content combination relation exists in a monitoring technology question-answer technology library or not; if yes, ending; otherwise, calculating the similarity between the index content combination updated by modification and all index content combinations in the monitoring technology question-answering library;
Taking a plurality of index content combinations with closest similarity, carrying out expert research and judgment, facilitating all selected index content combinations, judging whether the differences between the modified updated index content and the selected index content combinations are irrelevant or not, and if yes, eliminating the relevance; otherwise, the modified updated index content combination is retained.
Preferably, the similarity calculation process is as follows:
wherein S is i Similarity to the i-th index content combination;
A n the nth index content is the ith index content combination, if the contents are the same, taking 1, otherwise taking 0;
p i taking 1 in the same area range for the similarity of the places combined with the ith index content, otherwise taking 0;
t i taking 1 in the same date range for the time similarity combined with the ith index content, otherwise taking 0;
α i for the device type coefficients, obtained by looking up a table.
Preferably, the search process is:
searching in an input frame, referring to Lucene for full-text retrieval, firstly segmenting words for creating an index in the full-text retrieval, and then executing a searching process; carrying out word segmentation storage on all keywords of the monitoring technical questions and answers;
word segmentation is carried out on the problem, and the data id of the same words are found;
and inquiring a question-answer list to be inquired according to the data id, comparing the word percentages of the list, putting the list into a return list after the percentages meet the requirements, and returning to the page.
And inquiring search words in the monitoring technical questions and answers, putting the search words in a document, comparing the data according to the input words, and comparing each phrase change, and popping up similar question and answer key phrases until the similar question and answer key words are not matched with the monitoring technical questions and answers.
Preferably, the abnormal event activation includes manual activation and automatic activation.
The event is automatically or manually activated according to the event type, and the activation of the event is completed according to a predefined automatic activation rule or manually determined the type of abnormal event activation (trip or abnormal). And generating a presentation I, clicking a sending button, and sending to the mobile phone of the user.
Preferably, the automatic activation process is as follows:
and extracting the well-distinguished key core signals according to the signal identification function, automatically activating corresponding abnormal events or tripping events, automatically filling time, a transformer substation, equipment ledgers, core light word signals and weather information in a signal activation page according to signal content, and activating corresponding abnormal event management and control.
Preferably, the manual activation process is as follows:
the user selects an event to be activated on the activation page, and manually inputs related field contents including time, transformer stations, equipment accounts, core light word signals and weather information, and manually activates abnormal event management and control. After the activation is successful, an event alarm window is popped up on the main monitoring picture (the tripping event is a red alarm and the abnormal event is a yellow alarm). After the activation event is successful, the system automatically generates a trip brief report I/an abnormal brief report I according to the set content, a user can manually modify and perfect on the basis, click to send after confirming that the error is avoided, send the brief report to a mobile phone of the user, and update the event state of the main monitoring picture after the sending is finished.
Preferably, the abnormal event treatment comprises an information research and judgment stage, an inspection treatment stage and a tracking management stage; and filling in corresponding content in each stage to generate a brief report.
A device anomaly management and control system, comprising:
monitoring a technical question-answer library, and modifying and updating an index according to the equipment abnormal data; dividing the search words during search, and carrying out the matching in a question-answer list to return an index result;
the abnormal event case library is used for storing abnormal event cases, and the question-answer records in the monitoring technology question-answer library and the whole abnormal event process of the abnormal event whole process management and control platform are updated into the abnormal event case library;
the multi-disc analysis chamber carries out multi-disc analysis on the cases in the case library and updates the technical question-answer library;
the abnormal event whole process management and control platform comprises an abnormal event activation module, an abnormal event handling module and an abnormal event ending module.
The monitoring technology question-answering library, the abnormal event case library, the multi-disc analysis chamber and the abnormal event whole process management and control platform are built, huge and complex historical equipment abnormal data since the power grid is built are fully utilized, and the utilization efficiency of the historical abnormal data can be greatly improved.
Preferably, the monitoring technology question and answer library comprises an index content storage module for respectively storing cause analysis, result, early-stage research judgment and treatment in the index.
The main content of the monitoring technology question and answer is divided into a history version of the monitoring technology question and answer and a using version of the monitoring technology question and answer, the question and answer content inquired out in a page is the using version, the history version comprises the using version, each modification of the monitoring technology question and answer is to create a new question and answer record, the new question and answer record is confirmed by layer-by-layer verification and no repeated record in the history version is added into the history version, the new question and answer record is in a state to be checked at the moment, the new question and answer record is visible only in a checking list, the adding of a multi-disc analysis chamber firstly enters a draft state, the end of the multi-disc enters the state to be checked, and the follow-up monitoring technology question and answer flow comprises a series of verification.
Preferably, the monitoring technology question-answering library comprises:
the index adding module is used for adding a monitoring technology question-answer index and selecting a monitoring technology question-answer type; filling index content according to a preset standard format, and submitting an audit;
the index auditing module judges whether the newly added monitoring technology question-answer index is repeated with the index in the library, if so, the auditing is not passed, and the newly added monitoring technology question-answer index is returned again; otherwise, the monitoring technology question-answer index is stored in the monitoring technology question-answer library.
After layer-by-layer verification, determining that no repeated record in the history version is added into the history version, and editing the original existing technical question-answer items by a user to replace old content and supplement fresh content. The editors and the editing time are saved together. The revised items are issued and visible after expert approval. The expert reviews the content of the new supplement and modification of the monitoring technology questions and answers, and the expert can directly modify the questions and answers. The monitoring technical questions and answers which do not pass the verification are returned to the corresponding modules, the expert performs the verification again after resubmitting, and after the verification passes, newly added or revised items appear in the monitoring question and answer library and are given a new version. The technical question and answer content of the new revision allows the expert user to fall back even after being released, the history version can be selected for fall back, and all the history versions can be inquired in the function.
Preferably, the monitoring technology question-answering library comprises:
the knowledge extraction module is used for extracting index contents to form an initial triplet set;
the entity alignment module is used for carrying out model training on the initial triplet set, aligning the extracted entity by using the BERT model to obtain a standard entity, and then forming the standard triplet set;
And the knowledge graph visualization module is used for storing the standard triples in a Neo4j graph database and visually presenting the correlation between the knowledge.
And combining different expressions with the same meaning into the same expression through the knowledge graph, and judging whether repeated contents exist or not through the constructed knowledge graph when checking whether repeated or not, so that repeated input of abnormal data of the same equipment in different expression forms is avoided.
Preferably, the monitoring technology question-answering library comprises:
the word segmentation module is used for storing the word segmentation of all the keywords of the monitoring technical questions; word segmentation is carried out on the problem, and the data id of the same words are found;
and the result return module is used for inquiring a question and answer list to be inquired according to the data id, comparing the word percentages of the list, putting the word percentages into a return list after the percentages meet the requirements, and returning to the page.
And inquiring search words in the monitoring technical questions and answers, putting the search words in a document, comparing the data according to the input words, and comparing each phrase change, and popping up similar question and answer key phrases until the similar question and answer key words are not matched with the monitoring technical questions and answers.
Preferably, the abnormal event activation module includes a manual activation unit and an automatic activation unit.
The event is automatically or manually activated according to the event type, and the activation of the event is completed according to a predefined automatic activation rule or manually determined the type of abnormal event activation (trip or abnormal). And generating a presentation I, clicking a sending button, and sending to the mobile phone of the user.
Preferably, the automatic activation unit extracts the well-distinguished key core signals according to the signal identification function, automatically activates corresponding abnormal events or tripping events, automatically fills time, transformer stations, equipment accounts, core light word signals and weather information in a signal activation page according to signal content, and activates corresponding abnormal event management and control.
Preferably, the manual activation unit is used for manually inputting related field contents including time, transformer substation, equipment ledger, core light word signals and weather information, wherein the user selects an event to be activated on an activation page, and manually activating abnormal event management and control.
Preferably, the abnormal event handling module includes:
the information research and judgment unit is used for enabling a user to fill corresponding content in the information research and judgment page, automatically generating an abnormal brief report II/trip brief report II according to the filled field by the system, and sending the abnormal brief report II/trip brief report II to a mobile phone of a relevant user through a short message after the user confirms that the abnormal brief report II/trip brief report II is correct;
The system automatically generates an abnormal brief report III/trip brief report III according to the filled fields, and the user confirms the error and sends the abnormal brief report III/trip brief report III to a mobile phone of a relevant user through a short message;
the tracking management and control unit is used for enabling a user to fill relevant fields in the tracking management and control page, automatically generating an abnormal briefing fourth/tripping briefing fourth according to the filled fields by the system, and sending the abnormal briefing fourth/tripping briefing fourth to a mobile phone of the relevant user through a short message after the user confirms that the abnormal briefing fourth/tripping briefing fourth is correct.
An equipment anomaly management and control electronic product, comprising:
a memory for storing a computer program;
and the processor is used for executing the computer program to realize the steps of a device abnormality management and control method.
A computer readable storage medium having stored thereon a computer program for execution by a processor to perform the steps of a method of device anomaly management.
The beneficial effects of the invention are as follows:
the monitoring technology question-answering library, the abnormal event case library, the multi-disc analysis chamber and the abnormal event whole process management and control platform are built, huge and complex historical equipment abnormal data since the power grid is built are fully utilized, and the utilization efficiency of the historical abnormal data can be greatly improved.
Drawings
FIG. 1 is a flow chart of a monitoring technique question-answer library newly added audit in the invention.
Fig. 2 is a flow chart of the new audit of the abnormal event case base of the present invention.
FIG. 3 is a flow chart of the multiple disk analysis of the present invention.
FIG. 4 is a flow chart of the device exception whole process control of the present invention.
FIG. 5 is a block diagram of a device anomaly management and control system of the present invention.
In the figure, 1, a monitoring technology question-answering library, 2, an abnormal event case library, 3, a multiple disc analysis room and 4, an abnormal event whole process management and control platform.
Detailed Description
The technical scheme of the invention is further specifically described below through examples and with reference to the accompanying drawings.
Embodiment one:
the equipment abnormality management and control method of the embodiment comprises the following steps:
and (3) analyzing abnormal events corresponding to the abnormal data of the historical equipment, and respectively constructing a monitoring technology question-answering library, an abnormal event case library, a multiple disc analysis room and an abnormal event whole process management and control platform.
1. The monitoring technology question-answering library modifies and updates the index according to the equipment abnormal data; and dividing the search term during search, comparing the search term with the question-answer list, and returning an index result.
The monitoring technology question-answering library is used for converting the working experience of precipitation into a commander through an informatization technology to provide comprehensive and detailed auxiliary decision-making service and provide a convenient man-machine interaction mode.
The monitoring technology question-answering library comprises a newly added auditing process and a retrieval process.
Specifically, the newly added auditing process is shown in fig. 1, and includes:
1) And adding a monitoring technology question-answer index, and selecting a monitoring technology question-answer type.
According to the difference of the monitoring technology question-answer types, the monitoring technology question-answer types are divided into four types, namely typical accident information judgment analysis treatment, typical abnormal information judgment analysis treatment, typical out-of-limit information treatment and typical deflection information treatment.
2) Filling in index content according to a preset standard format, and submitting to auditing.
When the method is added, fields such as a question name, a question and answer keyword, information paraphrasing, associated signals, reason analysis, caused results, treatment points, reporting contents after on-site inspection, uploading pictures, uploading accessories and the like are required to be filled in, so that the technical question and answer of accurate positioning and paraphrasing monitoring are facilitated. The newly added monitoring technology questions and answers need expert auditing, and can be put into storage for storage after the auditing is passed.
The user repairs and compiles the original existing technical question and answer items to replace the old content and supplement the fresh content. The editors and the editing time are saved together. The revised items are issued and visible after expert approval.
The expert reviews the content of the new supplement and modification of the monitoring technology questions and answers, and the expert can directly modify the questions and answers. The monitoring technical questions and answers which do not pass the verification are returned to the corresponding modules, the expert performs the verification again after resubmitting, and after the verification passes, newly added or revised items appear in the monitoring question and answer library and are given a new version.
The technical question and answer content of the new revision allows the expert user to fall back even after being released, the history version can be selected for fall back, and all the history versions can be inquired in the function.
3) Judging whether the newly added monitoring technology question-answer index is repeated with the index in the library, if so, checking is not passed, and returning the newly added monitoring technology question-answer index again; otherwise, the monitoring technology question-answer index is stored in the monitoring technology question-answer library.
After layer-by-layer verification, determining that no repeated record in the history version is added into the history version, and editing the original existing technical question-answer items by a user to replace old content and supplement fresh content. The editors and the editing time are saved together. The revised items are issued and visible after expert approval. The expert reviews the content of the new supplement and modification of the monitoring technology questions and answers, and the expert can directly modify the questions and answers. The monitoring technical questions and answers which do not pass the verification are returned to the corresponding modules, the expert performs the verification again after resubmitting, and after the verification passes, newly added or revised items appear in the monitoring question and answer library and are given a new version. The technical question and answer content of the new revision allows the expert user to fall back even after being released, the history version can be selected for fall back, and all the history versions can be inquired in the function.
In the embodiment, index contents are combined through a knowledge graph, and an initial triplet set is formed through knowledge extraction; model training is carried out on the initial triplet set, the extracted entity is aligned by utilizing the BERT model, a standard entity is obtained, and then the standard triplet set is formed; the standard triples are stored in the Neo4j graph database, and the correlations between knowledge are visually presented.
Data source
The method comprises the steps of analyzing data organization conditions from the aspect of structural characteristics according to historical data of power grid equipment, and mainly dividing the data organization conditions into structural data, semi-structural data and unstructured data.
(2) Knowledge extraction
Converting the relational table into triples and obtaining ecological restoration knowledge by converting the database into a resource description framework data (Database to Resource Description Framework, D2R) tool for structuring; and converting into natural language description.
Semi-structured and unstructured data extraction mainly emphasizes three key processes of entity extraction, relationship extraction and attribute extraction. Entity extraction is the recognition of entities from raw text, including rule and dictionary based, statistical machine learning based, open-oriented methods. The relation extraction is carried out through semantic analysis and connection of entities, including manual construction of semantic rules, open domain information extraction frames and the like. Attribute extraction aggregates information from multi-source heterogeneous data and outlines entities, including rule-and heuristic-based methods.
(3) Knowledge fusion
Because of the problems of multiple and complex data sources, uneven knowledge quality, repeated knowledge, fuzzy relationship and the like, the ecological restoration knowledge extracted from the multi-source heterogeneous data is subject to entity alignment under unified specification, and the alignment mapping relationship pointed by the entity in the multi-source heterogeneous data is searched to achieve data fusion and form a high-quality ecological restoration knowledge base. In order to improve entity alignment efficiency, the scheme of the embodiment is based on a supervised learning entity alignment method, utilizes a BERT pre-training language model, constructs indexes, obtains candidate sets, calculates BERT semantic similarity and screens aligned entities in 4 parts, and realizes knowledge base entity alignment work.
(4) Knowledge processing
The method mainly comprises 4 links of ontology construction, knowledge reasoning, quality assessment and knowledge updating.
(5) Knowledge storage
And storing the standard triples by utilizing a Neo4j graph database to complete the construction of the ecological restoration knowledge graph.
And combining different expressions with the same meaning into the same expression through the knowledge graph, and judging whether repeated contents exist or not through the constructed knowledge graph when checking whether repeated or not, so that repeated input of abnormal data of the same equipment in different expression forms is avoided.
The reason analysis, the result, the early stage research and the treatment of the index content in the monitoring technology question-answering library are stored separately and are re-associated when the updated index is modified. The re-association process is as follows:
judging whether the single index contents exist in a monitoring technology question-answer library or not; if yes, the next step of judgment is carried out, otherwise, the similarity between the index content combination updated by modification and all index content combinations in the monitoring technology question-answering library is calculated.
Judging whether the modified and updated index content combination relation exists in a monitoring technology question-answer technology library or not; if yes, ending; otherwise, calculating the similarity between the index content combination updated by modification and all index content combinations in the monitoring technology question-answering library.
The similarity calculation process is as follows:
wherein S is i Similarity to the i-th index content combination;
A n the nth index content is the ith index content combination, if the contents are the same, taking 1, otherwise taking 0;
p i taking 1 in the same area range for the similarity of the places combined with the ith index content, otherwise taking 0;
t i taking 1 in the same date range for the time similarity combined with the ith index content, otherwise taking 0;
α i for the device type coefficients, obtained by looking up a table.
Taking a plurality of index content combinations with closest similarity, carrying out expert research and judgment, facilitating all selected index content combinations, judging whether the differences between the modified updated index content and the selected index content combinations are irrelevant or not, and if yes, eliminating the relevance; otherwise, the modified updated index content combination is retained.
The main content of the monitoring technology questions and answers is divided into a history version of the monitoring technology questions and answers and a using version of the monitoring technology questions and answers, and the questions and answers which are queried on the page are all the using versions, wherein the reasons in the monitoring technology questions and answers are analyzed, the results are caused, the early-stage research and judgment are carried out, and the disposal points are stored independently, and are re-associated when each modification is carried out. The history version comprises a used version, each time the monitoring technology question and answer is modified, a new question and answer record is created, the new question and answer record is confirmed through layer-by-layer verification, no repeated record in the history version is added to the history version, the history version is in a state to be checked, the history version is only visible in a checking list, the addition of the multi-disc analysis room firstly enters a draft state, the multi-disc end enters the state to be checked, and the follow-up monitoring technology question and answer flow comprises a series of verification.
Specifically, the retrieval process includes:
in order to adapt to different scene monitoring technology question-answer query modes, two types of question-answer query modes are adopted:
a man-machine interaction query, clicking to enter a man-machine query interface, displaying a chat frame on a page, and inputting contents to be queried by a user comprises: and the fields in the keyword, the question or the answer parsing are sent to the system.
The system searches and matches corresponding questions and answers in a question and answer library according to the content input by the user, and returns at least 5 query results of related content to the user.
The user clicks the returned result, and the page displays detailed question-answer content comprising: question and answer type, question name, question and answer keywords, information paraphrasing, associated signals, reason analysis, caused results, treatment points, reporting content after field inspection, pictures, accessories (including audio, video and files), version numbers and other fields.
The other is technical question-answer directory tree query, click technical question-answer directory tree query, classification of page display question-answer library and question-answer under each classification, and the user can manually go to query by turning pages.
The user can also query questions and answers through the query function in the page, and the queried questions and answers can be displayed under that category. The questions and answers may be queried in the form of fuzzy search question and answer questions under the entire directory or sub-directory. The user clicks the question and answer item, and the detailed content of the page display question and answer comprises fields such as question and answer type, question name, question and answer keywords, information paraphrasing, associated signals, reason analysis, caused results, disposal points, report contents after field inspection, pictures, accessories (including audio, video and files), version numbers and the like.
And (3) automatically filling search words, inquiring the search words in the monitoring technology questions and answers, putting the search words in a document, comparing the data according to the input words, and popping up similar question and answer key words until the monitoring technology question and answer key words are not matched.
Monitoring technical question-answer searching, searching in an input box, and searching full text by referring to Lucene. The full text retrieval is the process of creating an index by word segmentation and then executing search.
The display route is that firstly, word segmentation is carried out on all keywords of the monitoring technical questions and answers, then word segmentation is carried out on the questions, the data id of the same words is found, the questions and answers lists to be queried are queried according to the data id, word percentage comparison is carried out on the lists, and the lists are put into a return list to return to pages after the percentage reaches the requirement.
The full text retrieval process is divided into two major parts: indexing process and searching process.
The indexing process includes gathering data, building document objects, and creating an index (writing documents to an index repository).
The search flow includes creating a query, performing a search, and rendering search results.
The Lucene full text search does not directly query the database, so the data needs to be collected first.
Lucene is a type of document used to encapsulate data, all that is required is to convert the collected data to a type of document.
Lucene is used for automatically completing word segmentation and creating indexes in the process of writing the document into an index library. Thus, an index library is created, which is formally written with documents. When searching, it is necessary to specify which domain (i.e., field) to search, and word segmentation processing is also performed on the searched keywords.
The Lucene full-text retrieval can truly realize word segmentation on keywords and then execute a search function. And the results are more accurate.
The word segmentation is the core of full text retrieval. The word segmentation is to divide a text into words according to a certain rule.
Lucene is word-segmented according to an analyzer. Different analyzers are provided for different languages. The general standard analyzer is provided for dividing a section of Chinese or English into individual keywords, when searching, the user can divide own information into words, the data in the database or the index library can be divided into words, then a matching operation is carried out, and the default Chinese word analyzer views each word as a word, for example, the Chinese word analyzer IK can be divided into "I", "love", "skill", "operation", which obviously does not meet the requirements, so that the Chinese word analyzer IK solves the problem; IK provides two word segmentation algorithms: the ik_smart and the ik_max_word are the least segmentation, and an ambiguity recognition function is added for recommendation; the ik_max_word is the finest cut, and all the cuts can be cut off; adding custom words: there are some professional data in many cases, for example: "give a life to the country" after the ik_smart word segmentation, the result is: "in", "sensitive", "as", "ancestor", "contribution", "lifetime"; the sensitive person is a name, is split and needs to be added into a dictionary as a word; a config folder is arranged under the IK catalog and is used for storing a dictionary; creating a file: mydic, add "sensitivity" to it and then write the filename to the ikAnalyzer. Cfg. Xml file.
2. The abnormal event case library stores abnormal event cases, and the question-answer records in the monitoring technology question-answer library and the whole abnormal event process of the abnormal event whole process management and control platform are updated into the abnormal event case library.
The abnormal event case library is used for converting past historical abnormal events into commanders through an informatization technology to provide comprehensive and detailed auxiliary decision-making service.
The new audit flow of the abnormal event case library is shown in fig. 2, and comprises the following steps:
1) The new abnormal event cases are filled in, and the case names, event outlines, substations, equipment types, equipment models, occurrence time, equipment names, optical word signals, equipment accounts, information research and judgment, inspection and disposal, tracking and control, accessory uploading and the like are filled in.
2) Submitting the audit, and if the audit is passed, storing the audit in a case library; if the case does not pass, returning to the abnormal case library.
In this embodiment, there are two new ways to add the abnormal event cases, one is to click on the new import of the abnormal event cases, fill in the fields of case name, event summary, time, substation, equipment name, keywords (optical word signals), monitoring and handling (operation and maintenance unit inspection and report schedule, equipment account), field inspection and handling conditions, monitoring and tracking closed loop, pictures and accessories, etc.
The other is that the user also carries out the new addition of the abnormal event case through the template importing function. The newly added abnormal event cases enter a list to be checked in an expert checking function, and the abnormal event cases can be stored in an abnormal event case library after the verification is passed.
The expert reviews the new supplement and modification content of the abnormal event case, and the expert can directly modify the abnormal event case. The abnormal event cases which do not pass the verification are returned to the corresponding modules, and the expert performs the verification again after the resubmitting, and after the verification passes, the newly added or revised abnormal event cases are stored in the abnormal event case library.
The user can inquire the abnormal event case through different inquiry conditions (transformer substation, case name, equipment name, fault occurrence time and optical character signal). The user clicks on the case entry, and the details of the page display case include fields such as case name, event summary, time, substation, equipment name, keywords (optical character signals), monitoring treatment (operation and maintenance unit inspection report schedule, equipment ledger), on-site inspection and treatment conditions, monitoring tracking closed loop, pictures and attachments, similar cases, and the like. And the similar case fields are unmatched according to the case keywords (optical character signals), the keywords with high matching degree are preferentially displayed in front (case names are displayed), and the case names are clicked to display case details.
The system of the present embodiment provides two export modes: brief information derivation and detail information derivation.
The format of the profile derivation is Excel format, the content inside contains: case name, event summary, fault occurrence time, substation, equipment name, keywords (optical character signals).
The detailed information is exported in a word format, and the content comprises: case name, event summary, time of failure occurrence, substation, equipment name, keywords (optical character signals), monitoring treatment (operation and maintenance unit inspection report schedule, equipment ledger), on-site inspection and treatment conditions, monitoring tracking closed loop and all accessories (accessories are exported in compressed package form).
Reminder list export word: and exporting by using an Apache POI and FreeMark template engine, wherein the implementation mode is that pages with the same format are drawn in a word document according to the page format, page data use placeholders for subsequent replacement of real-time data, an xml file generated by a template is stored in a project appointed directory, and the Apache POI is utilized for exporting data after page display data are acquired.
The operation process comprises the following steps:
creating a word document, setting the format and the position of data in the document, writing by using a replacement character, converting the word into an xml file which can be identified in a project after the word design is completed, dividing detailed information to be exported into fields in the project, matching the fields in a word template, and filling the fields into the word document in the same way to realize the export.
Excel derived, POI can not only operate word, his location is to operate Microsoft Office read and write, microsoft Office contains many common Office files, such as: excel, ppt, word, visio, etc. …, maintains a header, and when exported as a header, the exception case file is exported by converting the table data into a file stream according to the export hierarchy.
3. And the multi-disc analysis room carries out multi-disc analysis on the cases in the case library and updates the monitoring technology question-answer library.
The multi-disc analysis room is a supplementary function of the equipment abnormal event case library, and aims to realize the transformation from fault case experience to technical knowledge question-answering, supplement and update the technical question-answering library and ensure the vitality of the technical question-answering library. And carrying out multiple disc analysis on the cases in the case library. The complex disc analysis is to characterize the cases, and the information in the cases is subjected to expert induction in the form of a technical question-answer knowledge tree. The knowledge tree can be saved to save the result of the duplication at any time. And (3) finishing the compound disc analysis and pushing the knowledge base to push the result of the compound disc analysis to a list to be checked of the technical question-answering module, and generating an un-new technical question-answering after the checking is passed.
The multiple disc analysis is shown in fig. 3, comprising:
1) Selecting multiple analysis cases.
The user can select cases from the case library inquiry for analysis, and can also directly enter a multiple disc analysis room for case analysis.
2) Professional equipment is selected, and an optical signal is selected.
The corresponding optical word signal is selected and the optical word signal may be modified and added.
3) Selecting information definition, reason analysis, caused result, treatment key points and early stage research and judgment.
Corresponding information definition, reason analysis, result generation, treatment key points and early stage research and judgment are selected, and the information definition, the reason analysis, the result generation, the treatment key points and the early stage research and judgment content can be modified and newly increased.
The method can be used for modifying and adding equipment profession, optical word signals, information definition, reason analysis, caused results, treatment key points and early-stage research and judgment.
4) And (3) finishing the analysis of the multiple discs, and checking the newly added analysis multiple discs.
If the modification or the addition is performed, the method is directly finished, otherwise, the auditing is performed.
4. The abnormal event whole process control platform controls the whole life cycle of the abnormal event, and comprises the functions of whole process treatment monitoring pages, activation events, abnormal event treatment, abnormal event tracking and the like.
The whole process is handled and monitored the page:
the device has the functions of tripping event type display, device defect management and control display, abnormal positioning display, technical question answering, case display and the like.
The equipment abnormality whole process control flow is shown in fig. 4, and includes:
1) The system has two activation modes for activating an event:
1. automatic activation:
and extracting the well-distinguished key core signals according to the signal identification function, automatically activating corresponding abnormal events or tripping events, automatically filling time, transformer substation, equipment account, core light word signals, weather and other information in a signal activation page according to the signal content, and activating corresponding abnormal event management and control.
2. Manual activation:
the user selects an event to be activated on the activation page, and manually inputs related field contents including information such as time, transformer substation, equipment ledger, core light word signals, weather and the like, and manually activates abnormal event management and control. After the activation is successful, an event alarm window is popped up on the main monitoring picture (the tripping event is a red alarm and the abnormal event is a yellow alarm).
After the activation event is successful, the system automatically generates a trip brief report I/an abnormal brief report I according to the set content, a user can manually modify and perfect on the basis, click to send after confirming that the error is avoided, send the brief report to a mobile phone of the user, and update the event state of the main monitoring picture after the sending is finished.
2) Event handling, abnormal event handling is divided into three phases.
The first information research and judgment stage:
the user fills in corresponding content in the information research and judgment page, the system automatically generates an abnormal brief report II/trip brief report II according to the filled-in field, and the user confirms the error and sends the error to the mobile phone of the relevant user through a short message.
Second examination treatment phase:
the user fills in the record in the treatment process in the treatment check page, the system automatically generates an abnormal brief report three/a trip brief report three according to the filled-in field, and the user confirms the error and sends the error to the mobile phone of the relevant user through a short message.
The third tracking and controlling stage:
the user fills in relevant fields in the tracking management and control page, the system automatically generates an abnormal brief report four/trip brief report four according to the filled-in fields, and the user confirms that the error is absent and then sends the abnormal brief report four/trip brief report four to the mobile phone of the relevant user through a short message.
Abnormal event tracking:
the user can monitor in real time which link the abnormal event treatment is carried out to and what is filled in by each link. The user can export the content in each link in Excel form. The user can manually close the tracking control of the abnormal event on the abnormal event tracking page.
3) And (5) uploading the analysis report after the event is completed, and generating an abnormal event case.
After the treatment of the abnormal event is completed, an event ending link is entered, and the user can upload an analysis report or generate a case.
The scheme of the embodiment establishes a monitoring technology question-answering library, an abnormal event case library, a multi-disc analysis room and an abnormal event whole process management and control platform, makes full use of huge and complex historical equipment abnormal data since the power grid is built, and can greatly improve the utilization efficiency of the historical abnormal data.
Embodiment two:
an equipment abnormality management and control system of the present embodiment adopts an equipment abnormality management and control method as shown in fig. 5, and includes a monitoring technology question-answering library 1, an abnormal event case library 2, a multiple disc analysis chamber 3 and an abnormal event whole process management and control platform 4.
The monitoring technology question-answering library 1 modifies and updates the index according to the equipment abnormal data; and dividing the search term during search, and carrying out matching in a question-answer list to return an index result.
The monitoring technology question and answer library 1 comprises an index content storage module for respectively storing cause analysis, cause result, early-stage research judgment and treatment in the index.
The main content of the monitoring technology question and answer is divided into a history version of the monitoring technology question and answer and a using version of the monitoring technology question and answer, the question and answer content inquired out in a page is the using version, the history version comprises the using version, each modification of the monitoring technology question and answer is to create a new question and answer record, the new question and answer record is confirmed by layer-by-layer verification and no repeated record in the history version is added into the history version, the new question and answer record is in a state to be checked at the moment, the new question and answer record is visible only in a checking list, the adding of a multi-disc analysis chamber firstly enters a draft state, the end of the multi-disc enters the state to be checked, and the follow-up monitoring technology question and answer flow comprises a series of verification.
Further, the monitoring technology question and answer library 1 further includes:
the index adding module is used for adding a monitoring technology question-answer index and selecting a monitoring technology question-answer type; filling in index content according to a preset standard format, and submitting to auditing.
The index auditing module judges whether the newly added monitoring technology question-answer index is repeated with the index in the library, if so, the auditing is not passed, and the newly added monitoring technology question-answer index is returned again; otherwise, the monitoring technology question-answer index is stored in the monitoring technology question-answer library.
After layer-by-layer verification, determining that no repeated record in the history version is added into the history version, and editing the original existing technical question-answer items by a user to replace old content and supplement fresh content. The editors and the editing time are saved together. The revised items are issued and visible after expert approval. The expert reviews the content of the new supplement and modification of the monitoring technology questions and answers, and the expert can directly modify the questions and answers. The monitoring technical questions and answers which do not pass the verification are returned to the corresponding modules, the expert performs the verification again after resubmitting, and after the verification passes, newly added or revised items appear in the monitoring question and answer library and are given a new version. The technical question and answer content of the new revision allows the expert user to fall back even after being released, the history version can be selected for fall back, and all the history versions can be inquired in the function.
Further, the monitoring technology question and answer library 1 further includes:
and the knowledge extraction module is used for extracting index contents to form an initial triplet set.
And the entity alignment module is used for carrying out model training on the initial triplet set, aligning the extracted entity by using the BERT model to obtain a standard entity, and then forming the standard triplet set.
And the knowledge graph visualization module is used for storing the standard triples in a Neo4j graph database and visually presenting the correlation between the knowledge.
And combining different expressions with the same meaning into the same expression through the knowledge graph, and judging whether repeated contents exist or not through the constructed knowledge graph when checking whether repeated or not, so that repeated input of abnormal data of the same equipment in different expression forms is avoided.
Further, the monitoring technology question and answer library 1 further includes:
the word segmentation module is used for storing the word segmentation of all the keywords of the monitoring technical questions; word segmentation is carried out on the problem, and the data id of the same words are found;
and the result return module is used for inquiring a question and answer list to be inquired according to the data id, comparing the word percentages of the list, putting the word percentages into a return list after the percentages meet the requirements, and returning to the page.
And inquiring search words in the monitoring technical questions and answers, putting the search words in a document, comparing the data according to the input words, and comparing each phrase change, and popping up similar question and answer key phrases until the similar question and answer key words are not matched with the monitoring technical questions and answers.
The abnormal event case library 2 stores abnormal event cases, and updates the question-answer records in the monitoring technology question-answer library and the abnormal event whole process of the abnormal event whole process management and control platform into the abnormal event case library.
The abnormal event case library 2 includes:
the abnormal event case adding module is used for filling in case names, event outlines, substations, equipment types, equipment models, occurrence time, equipment names, optical word signals, equipment accounts, information research and judgment, inspection and disposal, tracking and control, uploading accessories and the like.
The case auditing module audits the newly added abnormal event cases, and the audits pass and are stored in a case library; and if the verification is not passed, the abnormal event case is newly added.
And the multi-disc analysis room 3 carries out multi-disc analysis on the cases in the case library and updates the technical question-answer library.
The multiplex disc analysis chamber 3 includes:
and the multiple-disc analysis case selection module can enable a user to select a case from case library inquiry for analysis, and can also directly enter a multiple-disc analysis room for case analysis.
And the optical word signal selection module is used for selecting a corresponding optical word signal and modifying and adding the optical word signal.
The content selection module selects corresponding information definition, reason analysis, caused result, treatment key point and early stage research and judgment, and can modify and add the information definition, reason analysis, caused result, treatment key point and early stage research and judgment content.
The abnormal event whole process control platform 4 comprises an abnormal event activation module, an abnormal event handling module and an abnormal event ending module.
The abnormal event activation module comprises a manual activation unit and an automatic activation unit.
The event is automatically or manually activated according to the event type, and the activation of the event is completed according to a predefined automatic activation rule or manually determined the type of abnormal event activation (trip or abnormal). And generating a presentation I, clicking a sending button, and sending to the mobile phone of the user.
The automatic activation unit extracts the well-distinguished key core signals according to the signal identification function, automatically activates corresponding abnormal events or tripping events, automatically fills time, transformer stations, equipment accounts, core light word signals and weather information in a signal activation page according to signal content, and activates corresponding abnormal event management and control.
And the manual activation unit is used for enabling a user to select an event to be activated on the activation page, manually inputting related field contents including time, a transformer substation, equipment ledgers, core light word signals and weather information, and manually activating abnormal event management and control.
The abnormal event handling module includes:
the information research and judgment unit is used for enabling a user to fill corresponding content in the information research and judgment page, automatically generating an abnormal brief report II/trip brief report II according to the filled field by the system, and sending the abnormal brief report II/trip brief report II to a mobile phone of a relevant user through a short message after the user confirms that the abnormal brief report II/trip brief report II is correct;
the system automatically generates an abnormal brief report III/trip brief report III according to the filled fields, and the user confirms the error and sends the abnormal brief report III/trip brief report III to a mobile phone of a relevant user through a short message;
the tracking management and control unit is used for enabling a user to fill relevant fields in the tracking management and control page, automatically generating an abnormal briefing fourth/tripping briefing fourth according to the filled fields by the system, and sending the abnormal briefing fourth/tripping briefing fourth to a mobile phone of the relevant user through a short message after the user confirms that the abnormal briefing fourth/tripping briefing fourth is correct.
The abnormal event finalization module includes:
and the analysis report uploading unit enters an event ending link after the treatment of the abnormal event is completed, and a user can upload the analysis report.
And the case generation unit enters an event finishing link after the treatment of the abnormal event is completed, and generates an abnormal event case.
The scheme of the embodiment establishes a monitoring technology question-answering library, an abnormal event case library, a multi-disc analysis room and an abnormal event whole process management and control platform, makes full use of huge and complex historical equipment abnormal data since the power grid is built, and can greatly improve the utilization efficiency of the historical abnormal data.
Embodiment III:
an apparatus abnormality management and control electronic product of the present embodiment, and an apparatus abnormality management and control method of the first embodiment is operated, including a memory and a processor.
The memory is used for storing a computer program.
The processor is configured to execute the computer program to implement the steps of an apparatus anomaly management method according to the first embodiment.
Embodiment four:
a computer-readable storage medium of the present embodiment stores a computer program that is executed by a processor to implement the steps of an apparatus abnormality management method of the first embodiment.
It should be understood that the examples are only for illustrating the present invention and are not intended to limit the scope of the present invention. Further, it is understood that various changes and modifications may be made by those skilled in the art after reading the teachings of the present invention, and such equivalents are intended to fall within the scope of the claims appended hereto.
Claims (23)
1. The equipment abnormality control method is characterized by comprising the following steps of:
analyzing abnormal events corresponding to the abnormal data of the historical equipment, and respectively constructing a monitoring technology question-answering library, an abnormal event case library, a multiple disc analysis room and an abnormal event whole process management and control platform;
the abnormal event whole process control platform controls the whole life cycle of the abnormal event, including abnormal event activation, abnormal event treatment and abnormal event completion;
the monitoring technology question-answering library modifies and updates the index according to the equipment abnormal data; dividing the search term during search, comparing in a question-answer list, and returning an index result;
the abnormal event case library stores abnormal event cases, and the question-answer records in the monitoring technology question-answer library and the whole abnormal event process of the abnormal event whole process management and control platform are updated into the abnormal event case library;
and the multi-disc analysis room carries out multi-disc analysis on the cases in the case library and updates the monitoring technology question-answer library.
2. The method of claim 1, wherein the analysis of the cause, the consequences, the prior study and the disposition of the index content in the monitoring technical question-and-answer library are stored separately and re-associated when the updated index is modified.
3. The device anomaly management method according to claim 1 or 2, wherein the process of modifying the update index is: newly adding a monitoring technology question-answer index and selecting a monitoring technology question-answer type;
filling index content according to a preset standard format, and submitting an audit;
judging whether the newly added monitoring technology question-answer index is repeated with the index in the library, if so, checking is not passed, and returning the newly added monitoring technology question-answer index again; otherwise, the monitoring technology question-answer index is stored in the monitoring technology question-answer library.
4. A method of device anomaly management as claimed in claim 3 wherein the monitoring technique question-answer types include a typical accident information judgment analysis treatment, a typical anomaly information judgment analysis treatment, a typical out-of-limit information treatment and a typical shift information treatment.
5. The method for managing and controlling equipment anomalies according to claim 3, wherein the index contents are combined through a knowledge graph, and an initial triplet set is formed through knowledge extraction; model training is carried out on the initial triplet set, the extracted entity is aligned by utilizing the BERT model, a standard entity is obtained, and then the standard triplet set is formed; the standard triples are stored in the Neo4j graph database, and the correlations between knowledge are visually presented.
6. The device anomaly management method according to claim 2 or 5, wherein the re-association process is:
judging whether the single index contents exist in a monitoring technology question-answer library or not; if yes, entering the next step of judgment, otherwise, calculating the similarity between the index content combination updated by modification and all index content combinations in the monitoring technology question-answering library;
judging whether the modified and updated index content combination relation exists in a monitoring technology question-answer technology library or not; if yes, ending; otherwise, calculating the similarity between the index content combination updated by modification and all index content combinations in the monitoring technology question-answering library;
taking a plurality of index content combinations with closest similarity, carrying out expert research and judgment, facilitating all selected index content combinations, judging whether the differences between the modified updated index content and the selected index content combinations are irrelevant or not, and if yes, eliminating the relevance; otherwise, the modified updated index content combination is retained.
7. The method for managing and controlling equipment anomalies according to claim 6, wherein the similarity calculation process is as follows:
wherein S is i Similarity to the i-th index content combination;
A n The nth index content is the ith index content combination, if the contents are the same, taking 1, otherwise taking 0;
p i taking 1 in the same area range for the similarity of the places combined with the ith index content, otherwise taking 0;
t i taking 1 in the same date range for the time similarity combined with the ith index content, otherwise taking 0;
α i for the device type coefficients, obtained by looking up a table.
8. A method for device anomaly management as claimed in claim 3, wherein the search process is:
searching in an input frame, referring to Lucene for full-text retrieval, firstly segmenting words for creating an index in the full-text retrieval, and then executing a searching process; carrying out word segmentation storage on all keywords of the monitoring technical questions and answers;
word segmentation is carried out on the problem, and the data id of the same words are found;
and inquiring a question-answer list to be inquired according to the data id, comparing the word percentages of the list, putting the list into a return list after the percentages meet the requirements, and returning to the page.
9. The device anomaly management method of claim 1, wherein the anomaly event activation comprises manual activation and automatic activation.
10. The device anomaly management method of claim 9, wherein the automatic activation process is: and extracting the well-distinguished key core signals according to the signal identification function, automatically activating corresponding abnormal events or tripping events, automatically filling time, a transformer substation, equipment ledgers, core light word signals and weather information in a signal activation page according to signal content, and activating corresponding abnormal event management and control.
11. The method for controlling equipment abnormality according to claim 9, wherein the manual activation process is: the user selects an event to be activated on the activation page, and manually inputs related field contents including time, a transformer substation, a device ledger, a core light word signal and weather information, and manually activates abnormal event management and control.
12. A method of device exception handling according to claim 1 or 9 or 10 or 11, wherein said exception handling comprises an information research phase, an inspection handling phase and a trace handling phase; and filling in corresponding content in each stage to generate a brief report.
13. An equipment abnormality management system employing an equipment abnormality management method according to any one of claims 1 to 12, characterized by comprising:
monitoring a technical question-answer library, and modifying and updating an index according to the equipment abnormal data; dividing the search words during search, and carrying out the matching in a question-answer list to return an index result;
the abnormal event case library is used for storing abnormal event cases, and the question-answer records in the monitoring technology question-answer library and the whole abnormal event process of the abnormal event whole process management and control platform are updated into the abnormal event case library;
The multi-disc analysis chamber carries out multi-disc analysis on the cases in the case library and updates the technical question-answer library;
the abnormal event whole process management and control platform comprises an abnormal event activation module, an abnormal event handling module and an abnormal event ending module.
14. The system of claim 13, wherein the monitoring technology question-answering library includes an index content storage module for storing cause analysis, cause results, pre-study and disposition in the index, respectively.
15. The system for managing and controlling equipment abnormality according to claim 13 or 14, characterized in that said monitoring technique question-and-answer library comprises:
the index adding module is used for adding a monitoring technology question-answer index and selecting a monitoring technology question-answer type; filling index content according to a preset standard format, and submitting an audit;
the index auditing module judges whether the newly added monitoring technology question-answer index is repeated with the index in the library, if so, the auditing is not passed, and the newly added monitoring technology question-answer index is returned again; otherwise, the monitoring technology question-answer index is stored in the monitoring technology question-answer library.
16. The system for managing and controlling equipment anomalies according to claim 13, wherein the monitoring technique question-and-answer library includes:
The knowledge extraction module is used for extracting index contents to form an initial triplet set;
the entity alignment module is used for carrying out model training on the initial triplet set, aligning the extracted entity by using the BERT model to obtain a standard entity, and then forming the standard triplet set;
and the knowledge graph visualization module is used for storing the standard triples in a Neo4j graph database and visually presenting the correlation between the knowledge.
17. The system for managing and controlling equipment anomalies according to claim 13, wherein the monitoring technique question-and-answer library includes:
the word segmentation module is used for storing the word segmentation of all the keywords of the monitoring technical questions; word segmentation is carried out on the problem, and the data id of the same words are found;
and the result return module is used for inquiring a question and answer list to be inquired according to the data id, comparing the word percentages of the list, putting the word percentages into a return list after the percentages meet the requirements, and returning to the page.
18. The system of claim 13, wherein the abnormal event activation module comprises a manual activation unit and an automatic activation unit.
19. The system according to claim 18, wherein the automatic activation unit extracts the well-differentiated key core signals according to the signal recognition function, automatically activates the corresponding abnormal event or trip event, automatically fills time, transformer substation, equipment ledger, core light word signals and weather information in the signal activation page according to the signal content, and activates the corresponding abnormal event management.
20. The system of claim 18, wherein the manual activation unit, the user selects an event to be activated on the activation page, and the manual input of the related field contents includes time, transformer station, equipment ledger, core light word signal and weather information, and the manual activation of the abnormal event management.
21. The device anomaly management system of claim 13, wherein the anomaly event handling module comprises:
the information research and judgment unit is used for enabling a user to fill corresponding content in the information research and judgment page, automatically generating an abnormal brief report II/trip brief report II according to the filled field by the system, and sending the abnormal brief report II/trip brief report II to a mobile phone of a relevant user through a short message after the user confirms that the abnormal brief report II/trip brief report II is correct;
the system automatically generates an abnormal brief report III/trip brief report III according to the filled fields, and the user confirms the error and sends the abnormal brief report III/trip brief report III to a mobile phone of a relevant user through a short message;
the tracking management and control unit is used for enabling a user to fill relevant fields in the tracking management and control page, automatically generating an abnormal briefing fourth/tripping briefing fourth according to the filled fields by the system, and sending the abnormal briefing fourth/tripping briefing fourth to a mobile phone of the relevant user through a short message after the user confirms that the abnormal briefing fourth/tripping briefing fourth is correct.
22. An equipment anomaly management and control electronic product, comprising:
a memory for storing a computer program;
a processor for executing the computer program to implement the steps of a device anomaly management method as claimed in any one of claims 1 to 12.
23. A computer readable storage medium having stored thereon a computer program for execution by a processor to perform the steps of a device anomaly management method according to any one of claims 1 to 12.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311302053.4A CN117744965A (en) | 2023-10-09 | 2023-10-09 | Equipment abnormality control method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311302053.4A CN117744965A (en) | 2023-10-09 | 2023-10-09 | Equipment abnormality control method and system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117744965A true CN117744965A (en) | 2024-03-22 |
Family
ID=90278249
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311302053.4A Pending CN117744965A (en) | 2023-10-09 | 2023-10-09 | Equipment abnormality control method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117744965A (en) |
-
2023
- 2023-10-09 CN CN202311302053.4A patent/CN117744965A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN114579875B (en) | Equipment fault diagnosis and maintenance knowledge recommendation system based on knowledge graph | |
| CN112612902A (en) | Knowledge graph construction method and device for power grid main device | |
| CN115619383B (en) | Fault diagnosis method and device based on knowledge graph and computing equipment | |
| CN111967761A (en) | Monitoring and early warning method and device based on knowledge graph and electronic equipment | |
| US20110060712A1 (en) | Method and system for design check knowledge construction | |
| CN116703337B (en) | Project document examination system and method based on artificial intelligence technology | |
| CN115757689A (en) | Information query system, method and equipment | |
| CN112036129A (en) | Power distribution network emergency drilling scheme digital aided design system and method | |
| CN113487211A (en) | Nuclear power equipment quality tracing method and system, computer equipment and medium | |
| CN105260300A (en) | Service test method based on CAS (General Classification Standards of China Accounting Standards) application platform | |
| CN115547466A (en) | Medical institution registration and review system and method based on big data | |
| CN115983571A (en) | Construction project auditing method and system based on artificial intelligence for construction industry | |
| CN115619090A (en) | Safety assessment method based on model and data driving | |
| JPH09212353A (en) | Method and device for supporting reused design | |
| CN113657854A (en) | Man-machine cooperation intelligent audit analysis method for incomplete data | |
| CN109636303A (en) | A kind of storage method and system of semi-automatic extraction and structured document information | |
| US8239362B1 (en) | Using metadata fragments as authoritative manufacturing work instructions | |
| CN117744965A (en) | Equipment abnormality control method and system | |
| Liu et al. | Construction of intelligent query system for metro electromechanical equipment faults based on the knowledge graph | |
| CN114862233A (en) | Intelligent decision method and intelligent decision system | |
| CN114997154A (en) | Automatic construction method and system for speaker-to-speaker robot corpus | |
| CN113901034A (en) | Method for automatically identifying administrative non-complaint execution case source | |
| CN116501757B (en) | ER diagram-based simulation data construction method and device | |
| CN117787920B (en) | Intelligent data processing system and method for informationized project | |
| Liu et al. | Application of Knowledge Graph Technology in the Field of Power Grid Infrastructure |
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 |