CN-122021853-A - Highway electromechanical operation and maintenance knowledge base construction method based on big data

CN122021853ACN 122021853 ACN122021853 ACN 122021853ACN-122021853-A

Abstract

The invention discloses a construction method of an electromechanical operation and maintenance knowledge base of a highway based on big data, which relates to the technical field of traffic and transportation informatization and comprises the steps of generating a relation candidate through co-occurrence relation, syntactic dependency and cross-modal similarity according to a fragment set, an entity set and an attribute table, carrying out causal direction inspection on the relation candidate and a causal trigger word by combining a time sequence, generating a causal triplet with confidence, splicing the causal triplet into a causal chain set, identifying a typical evolution mode, binding tracing information for each causal chain node and each edge, generating an evidence set, merging causal chain facts with semantic equivalence based on the evidence set, forming a candidate fact set, establishing an evidence binding table, carrying out comprehensive reliability calculation on the candidate fact set, and generating a fusion fact set. The invention enables the electromechanical operation and maintenance knowledge base to have dynamic evolution and high reliability, thereby effectively supporting preventive maintenance and intelligent decision of the electromechanical equipment of the expressway.

Inventors

FAN JIFEI
HE GUOTAO
LI JUNFENG
FENG HAO
HAN YU
GU YUE

Assignees

陕西高速电子工程有限公司

Dates

Publication Date: 20260512
Application Date: 20260130

Claims (10)

1. A highway electromechanical operation and maintenance knowledge base construction method based on big data is characterized by comprising the following steps of, Collecting multi-source heterogeneous data of the electromechanical operation and maintenance of the expressway, establishing a source index, and preprocessing the multi-source heterogeneous data to obtain a standardized event table, a standardized text table and a standardized multimedia transcription table; dividing a standardized event table, a standardized text table and a standardized multimedia transcription table into fragment sets through a sliding window, generating semantic representation vectors, and extracting entity sets and attribute tables; Generating a relation candidate through co-occurrence relation, syntax dependence and cross-modal similarity according to the fragment set, the entity set and the attribute table, and carrying out causal direction detection by combining a time sequence and a causal trigger word to generate a causal triplet with confidence; Splicing the causal triplets into a causal chain set, identifying a typical evolution mode, binding traceability information for each causal chain node and each causal chain edge, and generating an evidence set; combining the causal chain facts with semantic equivalence based on the evidence set to form a candidate fact set, establishing an evidence binding table, and performing comprehensive credibility calculation on the candidate fact set to generate a fusion fact set; And when the main conclusion in the fusion fact set meets the warehousing reliability threshold, writing the main conclusion and the traceability index into the electromechanical operation and maintenance knowledge base.
2. The method for constructing a knowledge base of electromechanical operation and maintenance of a highway based on big data according to claim 1, wherein said steps of collecting multi-source heterogeneous data of electromechanical operation and maintenance of a highway and establishing a source index are as follows, Collecting multisource heterogeneous data of electromechanical operation and maintenance of the expressway; And establishing a source index according to the source number, the acquisition mode, the acquisition time and the acquisition place of the multi-source heterogeneous data.
3. The method for constructing the highway electromechanical operation and maintenance knowledge base based on big data according to claim 2, wherein the standardized event table, text table and multimedia transcription table are obtained after preprocessing the multi-source heterogeneous data, Performing equipment unique coding, time alignment and space position mapping on the multi-source heterogeneous data to generate preprocessed multi-source heterogeneous data; dividing the preprocessed multi-source heterogeneous data according to data categories and unifying formats to generate a standardized event table, a standardized text table and a standardized multimedia transcription table.
4. The method for constructing the high-data-based electromechanical operation and maintenance knowledge base of the highway according to claim 3, wherein said dividing the standardized event table, the standardized text table and the standardized multimedia transcription table into segment sets through the sliding window, generating the semantic representation vector comprises the following steps, Dividing time intervals by a sliding window through a standardized event table, a standardized text table and a standardized multimedia transcription table to generate a fragment set; And carrying out semantic coding on the fragment set to generate a fragment semantic primary representation, and obtaining a semantic representation vector through vector fusion.
5. The method for constructing a knowledge base of electromechanical operation and maintenance of highway based on big data as set forth in claim 4, wherein the steps of extracting entity sets and attribute tables are as follows, Extracting an entity set from the fragment set according to the semantic representation vector; And identifying, unifying and normalizing the attributes of the entity set to generate an attribute table.
6. The method for constructing a knowledge base of electromechanical operation and maintenance of highway based on big data according to claim 5, wherein said generating relationship candidates through co-occurrence relationship, syntax dependency and cross-modal similarity based on fragment set, entity set and attribute table comprises the following steps, Counting the co-occurrence condition of the entity sets in the fragment sets to generate a co-occurrence relation; identifying a dependency structure between entity sets by a syntax dependency analysis method, and generating a syntax dependency relationship; Calculating semantic similarity of different sources in the fragment set, and generating a cross-modal similarity relation; and merging the co-occurrence relationship, the syntax dependency relationship and the cross-modal similarity relationship into a relationship candidate.
7. The method for constructing an electromechanical operation and maintenance knowledge base of a highway based on big data according to claim 6, wherein said combining time sequence and causal trigger words performs causal direction check to generate a causal triplet with confidence, specifically comprising the following steps, Carrying out causal direction detection according to the time sequence information in the relation candidate and fragment set and the causal trigger word to obtain a relation set with causality; And weighting and calculating causality relations in the relation set based on the source credibility, freshness and consistency to generate a causal triplet with credibility.
8. The method for constructing an electromechanical operation and maintenance knowledge base of a highway based on big data according to claim 7, wherein said splicing causal triplets into causal chain sets, and identifying a typical evolution mode, binding traceability information for each causal chain node and edge, generating evidence sets, specifically comprising the following steps, Splicing the causal triples with the confidence according to the equipment identification and the time sequence to generate a causal chain set; Counting the occurrence frequency of a causal chain set, carrying out pattern mining, and identifying a typical evolution pattern; binding tracing information for nodes and edges in the causal chain set based on the source index; and combining according to the matching relation of the typical evolution mode and the causal chain set after the tracing information is bound, and generating an evidence set.
9. The method for constructing an electromechanical operation and maintenance knowledge base of a highway based on big data as set forth in claim 8, wherein said combining the causal link facts with semantic equivalence based on the evidence set forms a candidate fact set, and establishing an evidence binding table, performing comprehensive credibility calculation on the candidate fact set to generate a fusion fact set, specifically comprising the following steps, Based on the evidence set, identifying causal chain facts with semantic equivalence and combining the causal chain facts to form a candidate fact set; Establishing a corresponding relation between the candidate fact set and the evidence set, and generating an evidence binding table; and carrying out weighted operation on the source credibility, freshness and consistency based on the evidence binding table to obtain a candidate fact set with the credibility, and obtaining a fusion fact set.
10. The method for constructing an electromechanical operation and maintenance knowledge base of a highway based on big data as set forth in claim 9, wherein when the main conclusion in the fusion fact set meets the threshold of the warehousing reliability, the main conclusion and the traceability index are written into the electromechanical operation and maintenance knowledge base, specifically as follows, Extracting a main conclusion from the fusion fact set, and screening out the main conclusion meeting the warehousing reliability threshold value; And extracting a tracing index of the main conclusion meeting the credibility threshold according to the tracing information, and writing the tracing index into an electromechanical operation and maintenance knowledge base.

Description

Highway electromechanical operation and maintenance knowledge base construction method based on big data Technical Field The invention relates to the technical field of traffic transportation informatization, in particular to a method for constructing an electromechanical operation and maintenance knowledge base of a highway based on big data. Background Along with the continuous expansion of the traffic scale of the expressway and the continuous increase of the types and the quantity of the electromechanical equipment, the electromechanical operation and maintenance of the expressway gradually becomes an important link for guaranteeing the safe operation of the road and improving the service level. In recent years, with the rapid development of technologies such as big data, artificial intelligence and knowledge graph, the academic world and industry have attempted to apply these emerging technologies to the electromechanical operation and maintenance scene of the expressway. For example, equipment operation data are collected through deployment sensors, maintenance events are recorded through a work order system, text information is extracted through combination with a natural language processing method, and automatic management of partial operation and maintenance knowledge is achieved. Meanwhile, with the application of multi-mode technologies such as voice recognition, video analysis and the like, the sources of electromechanical operation and maintenance knowledge tend to be diversified, and the characteristics of large data volume, complex structure and high real-time requirement are presented. However, the related art still has certain disadvantages. On the one hand, the existing research stays at the level of a single data source or a static knowledge base, systematic fusion and causal modeling for the electromechanical operation and maintenance multisource heterogeneous data of the expressway are lacked, and the dynamic relationship among equipment faults, reasons and disposal measures is difficult to accurately reveal, so that the knowledge base is insufficient in updating lag and evolution capability. On the other hand, the current knowledge updating mostly adopts a simple rule matching or statistical method, lacks a tracing mechanism and reliability calculation, and is difficult to ensure the reliability and the interpretability of a knowledge conclusion, so that uncertainty exists in knowledge base when supporting decision making and guiding actual operation and maintenance. Disclosure of Invention The present invention has been made in view of the above-described problems occurring in the prior art. Therefore, the invention provides a highway electromechanical operation and maintenance knowledge base construction method based on big data, which solves the problems of insufficient knowledge dynamic evolution capability and insufficient knowledge reliability guarantee in the existing highway electromechanical operation and maintenance knowledge construction technology. In order to solve the technical problems, the invention provides the following technical scheme: The invention provides a method for constructing a highway electromechanical operation and maintenance knowledge base based on big data, which comprises the steps of collecting multisource heterogeneous data of the highway electromechanical operation and maintenance, establishing a source index, Collecting multisource heterogeneous data of electromechanical operation and maintenance of the expressway; And establishing a source index according to the source number, the acquisition mode, the acquisition time and the acquisition place of the multi-source heterogeneous data. The invention provides a construction method of an electromechanical operation and maintenance knowledge base of a highway based on big data, which comprises the following steps of preprocessing multi-source heterogeneous data to obtain a standardized event table, a standardized text table and a standardized multimedia transcription table, Performing equipment unique coding, time alignment and space position mapping on the multi-source heterogeneous data to generate preprocessed multi-source heterogeneous data; dividing the preprocessed multi-source heterogeneous data according to data categories and unifying formats to generate a standardized event table, a standardized text table and a standardized multimedia transcription table. The invention provides a construction method of an electromechanical operation and maintenance knowledge base of a highway based on big data, which comprises the steps of dividing a standardized event table, a standardized text table and a standardized multimedia transcription table into fragment sets through a sliding window to generate semantic expression vectors, Dividing time intervals by a sliding window through a standardized event table, a standardized text table and a standardized multimedia transcription table to generate a fragmen