CN-122022603-A - Influence factor ordering method for importance of subway station

CN122022603ACN 122022603 ACN122022603 ACN 122022603ACN-122022603-A

Abstract

The disclosure relates to a method for ordering influence factors of importance of subway stations. The method is applied to the technical field of data processing, acquires influence factor scoring data of subway stations, carries out consistency evaluation and cluster analysis, determines an expert consensus influence matrix of the subway stations, reduces the subjective deviation and outlier influence of a single expert, carries out influence factor association degree analysis on the subway stations based on document knowledge data of the subway stations, generates a knowledge correction matrix, can introduce document evidence constraint to enhance objectivity of a result, carries out weighted fusion on the expert consensus influence matrix and the knowledge correction matrix, generates an index weighted influence matrix of the subway stations so as to realize fusion of multi-source evidence, and finally carries out decision test, evaluation experiment and network analysis weight calculation on the index weighted influence matrix in sequence, so that on the premise of guaranteeing a logic closed loop, the influence relation among indexes is constructed more accurately, and a stable and reliable influence factor sequencing result of the importance of the subway stations is output.

Inventors

YANG XIAOXIA
ZHANG CHAOXU
LI JIHONG
CHENG SEN
TIAN YANBING
PEI XIAOJUAN
REN LING
ZHAO LINGYAN
XIN LIPING
PAN FUQUAN
QU DAYI
HUANG SHOUXIANG
DING MINGYUE
WEI JINLI
YIN YIN
WANG JIJUN
MA HAO
ZHOU MING
YANG FAZHAN
ZHANG PING
LI QIANG

Assignees

青岛理工大学
中国铁建电气化局集团有限公司
中铁建电气化局集团第三工程有限公司
中铁通信信号勘测设计院有限公司

Dates

Publication Date: 20260512
Application Date: 20260415

Claims (17)

1. The influence factor ordering method for the importance of the subway station is characterized by comprising the following steps of: obtaining influence factor scoring data of subway stations and literature knowledge data of the subway stations; carrying out consistency evaluation and cluster analysis on the influence factor scoring data to determine an expert consensus influence matrix of the subway station; According to the literature knowledge data, carrying out influence factor association degree analysis on subway stations, and determining a knowledge correction matrix; Carrying out weighted fusion on the expert consensus influence matrix and the knowledge correction matrix to generate an index weighted influence matrix of the subway station; and sequentially carrying out decision tests, evaluation experiments and network analysis weight calculation on the index weighted influence matrix to obtain influence factor sequencing results of subway station importance.
2. The method of claim 1, wherein obtaining literature knowledge data for a subway station comprises: acquiring a document record data file of a subway station; dividing the document record data file into a plurality of document records according to the record ending mark; carrying out field identification on each document record, and obtaining a field identification in each document record; according to the field identifiers, validity screening and cleaning are carried out on each document record, and a processed document record is obtained; generating a unique document identification for the processed document record; and storing the unique literature identification, the field identification and each literature record in an associated mode, generating structured literature knowledge data, and taking the structured literature knowledge data as the literature knowledge data of subway points.
3. The method of claim 1, wherein said performing a consistency assessment and cluster analysis on said influence factor scoring data to determine an expert consensus influence matrix for a subway station comprises: consistency evaluation is carried out on the influence factor scoring data, and consistency evaluation data of the influence factor scoring data are determined; Performing cluster analysis on the consistency evaluation data to obtain a cluster result of influence factor scoring data; And weighting the influence factor scoring data according to the clustering result to obtain the expert consensus influence matrix of the subway station.
4. The method of claim 3, wherein said performing a consistency assessment of said influence factor scoring data, determining consistency assessment data for influence factor scoring data, comprises: obtaining scoring data given by any expert and confidence degree given by any expert from the scoring data of the influencing factors; Constructing an expert scoring matrix according to scoring data given by a plurality of experts, and constructing a confidence coefficient matrix according to confidence coefficients given by the plurality of experts; Normalizing the expert scoring matrix to obtain a normalized expert scoring matrix; And determining intra-group correlation coefficients according to the normalized expert scoring matrix and the confidence coefficient matrix, and taking the intra-group correlation numbers as the consistency evaluation data.
5. The method of claim 3, wherein performing a cluster analysis on the consistency assessment data to obtain a cluster result of influence factor scoring data comprises: flattening expert scoring matrices contained in the consistency evaluation data into scoring vectors; Normalizing the score vector to obtain a normalized score vector; Performing dimension reduction on the standardized scoring vector by adopting a popular projection algorithm to obtain a dimension-reduced scoring vector; And clustering the scoring vectors after dimension reduction by adopting a density clustering algorithm to obtain the clustering result.
6. The method of claim 3, wherein weighting the impact factor scoring data according to the clustering result to obtain an expert consensus impact matrix for the subway station comprises: calculating the cluster weight of each cluster according to the clustering result, and determining the expert weight in each cluster; and determining the expert consensus matrix according to the expert weight in each cluster, the cluster weight of each cluster and the expert scoring matrix contained in the consistency evaluation data.
7. The method of claim 6, wherein the method of determining the expert weights within each cluster comprises any one of the following methods: acquiring a clustering confidence coefficient from the clustering result, and determining expert weights in each cluster based on the clustering confidence coefficient; Obtaining a scoring variation coefficient of any expert, and determining the expert weight in each cluster according to the scoring variation coefficient of any expert; and acquiring expert scoring vector average values in any cluster from the clustering result, and determining expert weights in each cluster according to the expert scoring vector average values in any cluster and a correlation coefficient function.
8. The method of claim 6, wherein the determining the cluster weight of each cluster comprises: acquiring expert average clustering confidence coefficient of any cluster and cohesive force of any cluster from the clustering result; and carrying out weight calculation on the cluster quality of the expert average clustering confidence coefficient of any cluster and the cohesive force of any cluster to obtain the weight of each cluster.
9. The method according to claim 1, wherein the analyzing the degree of influence factor association of the subway station according to the literature knowledge data to determine the knowledge correction matrix comprises: carrying out knowledge extraction on the literature knowledge data to generate a knowledge graph; And analyzing the influence factor association degree of the subway station based on the knowledge graph, and determining a knowledge correction matrix.
10. The method of claim 9, wherein the performing knowledge extraction on the document knowledge data to generate a knowledge graph comprises: Performing index identification on the literature knowledge data to obtain influence factor indexes in the literature knowledge data; acquiring the relation weight of every two documents in the document knowledge data; And forming the knowledge graph by taking each document in the document knowledge data as a point, taking the relation weight as a connecting line of the two documents and taking the influence factor index as attribute information of the point of the corresponding document.
11. The method of claim 10, wherein the obtaining the relationship weights for each two documents in the document knowledge data comprises: Vectorizing and encoding each document in the document knowledge data to generate a document vector representation; based on a cosine similarity calculation method, calculating semantic similarity between every two document vector representations, and generating an initial document similarity matrix; Obtaining a document vector representation with semantic similarity larger than a preset threshold value from the initial document similarity matrix, and/or obtaining a document vector representation corresponding to a preset semantic similarity from the initial document similarity matrix to generate a target document similarity matrix; and taking the semantic similarity between every two document vector representations in the target document similarity matrix as the relation weight of every two documents.
12. The method of claim 9, wherein the analyzing the degree of influence factor association for the subway station based on the knowledge graph to determine the knowledge correction matrix comprises: performing entity and relation representation learning on the knowledge graph to obtain a document entity vector and an index entity vector; and determining the knowledge correction matrix according to the document entity vector and the index entity vector.
13. The method of claim 1, wherein the weighting and fusing the expert consensus impact matrix and the knowledge modification matrix to generate an index weighted impact matrix for a subway station comprises: Respectively carrying out normalization processing on the expert consensus influence matrix and the knowledge correction matrix to obtain a normalized expert consensus influence matrix and a normalized knowledge correction matrix; And according to the fusion weight coefficient, carrying out linear weighted fusion on the normalized expert consensus influence matrix and the normalized knowledge correction matrix to obtain the index weighted influence matrix of the subway station.
14. The method of claim 1, wherein the sequentially performing decision test, evaluation test and network analysis weight calculation on the index weighted influence matrix to obtain an influence factor ranking result of the subway station importance comprises: Determining a weight vector of an influence factor index of the subway station according to the index weighted influence matrix; and sequencing the influence factor indexes according to the sequence of the weight vectors from big to small to obtain the influence factor sequencing result.
15. The method of claim 14, wherein determining the weight vector of the impact factor indicator for the subway station based on the indicator weighted impact matrix comprises: Taking the index weighted influence matrix as a direct influence matrix, and carrying out normalization processing on the direct influence matrix to obtain a normalized direct influence matrix; calculating a total influence matrix based on the normalized direct influence matrix; normalizing the total influence matrix, and taking the normalized total matrix as a super matrix; performing power iteration on the supermatrix to obtain a limit supermatrix; And determining a weight vector of the influence factor index according to the limit super matrix.
16. The method as recited in claim 15, further comprising: Acquiring expert channel source weights and knowledge channel source weights from the weight vectors, wherein the expert channel source weights are used for representing the relative contribution of expert sources in a weighted fusion process, and the knowledge channel source weights are used for representing the relative contribution of knowledge sources in the weighted fusion process; and outputting the expert channel source weight and the knowledge channel source weight.
17. The method as recited in claim 15, further comprising: calculating the influence degree of the index and the influence degree of the index according to the total influence matrix; Calculating the centrality and causality according to the index influence degree and the index affected degree; Carrying out causal classification on the influence factor indexes according to the causality to obtain causal categories; and sorting the influence factor indexes according to the centrality to obtain an importance sorting result, and generating intra-class sorting in the causal class.

Description

Influence factor ordering method for importance of subway station Technical Field The disclosure relates to the technical field of data processing, in particular to a method for ordering influence factors of importance of subway stations. Background Urban rail transit is taken as an important component of an urban public transport system, and is responsible for a great deal of commuting and traveling demands. The subway station is used as a key node of a rail transit network, and the importance of the subway station is not only reflected in the capability of passenger flow distribution and transfer organization, but also closely related to factors such as urban space structure, peripheral land utilization function, operation guarantee level and the like. The subway station importance influencing factors are identified, quantitatively modeled and a sequencing result is formed, and decision basis can be provided for station planning and site selection, operation resource allocation, passenger flow organization optimization, facility transformation priority formulation, risk toughness improvement and the like. Therefore, it is desirable to provide a method for sorting influence factors of subway station importance, so as to determine stable and reliable influence factor sorting results. Disclosure of Invention In order to solve the technical problems, the present disclosure provides a method for sorting influence factors of importance of subway stations. In a first aspect, the present disclosure provides a method for sorting influence factors of importance of subway stations, including: obtaining influence factor scoring data of subway stations and literature knowledge data of the subway stations; carrying out consistency evaluation and cluster analysis on the influence factor scoring data to determine an expert consensus influence matrix of the subway station; According to the literature knowledge data, carrying out influence factor association degree analysis on subway stations, and determining a knowledge correction matrix; Carrying out weighted fusion on the expert consensus influence matrix and the knowledge correction matrix to generate an index weighted influence matrix of the subway station; and sequentially carrying out decision tests, evaluation experiments and network analysis weight calculation on the index weighted influence matrix to obtain influence factor sequencing results of subway station importance. In a second aspect, the present disclosure provides an influence factor ordering apparatus for subway station importance, including: the first acquisition module is used for acquiring influence factor scoring data of the subway station and literature knowledge data of the subway station; The first determining module is used for carrying out consistency evaluation and cluster analysis on the influence factor scoring data to determine an expert consensus influence matrix of the subway station; The second determining module is used for analyzing the influence factor association degree of the subway station according to the literature knowledge data and determining a knowledge correction matrix; The generation module is used for carrying out weighted fusion on the expert consensus influence matrix and the knowledge correction matrix to generate an index weighted influence matrix of the subway station; And the second acquisition module is used for sequentially carrying out decision tests, evaluation experiments and network analysis weight calculation on the index weighted influence matrix to obtain influence factor sequencing results of subway station importance. In a third aspect, embodiments of the present disclosure further provide an electronic device, including: one or more processors; storage means for storing one or more programs, The one or more programs, when executed by the one or more processors, cause the one or more processors to implement the method provided by the first aspect. In a fourth aspect, embodiments of the present disclosure also provide a computer-readable storage medium having stored thereon a computer program which, when executed by a processor, implements the method provided by the first aspect. Compared with the prior art, the technical scheme provided by the embodiment of the disclosure has the following advantages: According to the influence factor sequencing method for the importance of the subway station, the influence factor scoring data of the subway station are obtained by experts, consistency evaluation and cluster analysis are carried out, the expert consensus influence matrix of the subway station is determined, so that the subjective deviation and the outlier influence of a single expert are reduced, the influence factor association degree analysis is carried out on the subway station based on the document knowledge data of the subway station to generate a knowledge correction matrix, document evidence constraint can be introduced to enhance the objectivity of