Railway network node importance evaluation method based on global and neighborhood feature fusion

Abstract

The invention discloses a global and neighborhood feature fused railway network node importance degree evaluation method. The method comprises the following steps: obtaining railway transportation network node evaluation data; constructing a railway transportation facility topology network and railway transportation service complex network model; analyzing the global importance of the railway network nodes; outputting neighborhood attributes of the railway network nodes; setting a global and neighborhood feature fusion factor, and calculating a network node comprehensive importance index; and evaluating the importance of the railway network nodes. According to the method, the node importance of the railway transportation complex network is rapidly and accurately evaluated, the node importance evaluation result is more objective and accurate by fusing the global and neighborhood characteristics of the railway network nodes, and technical service support is provided for scientific and efficient railway transportation planning and complex network node importance analysis.

Description

Technical Field

[0001] The present invention belongs to the technical field of railway networking, and in particular to a railway network node importance evaluation method integrating global and neighborhood features. Background Art

[0002] As my country's urbanization accelerates and regional coordination and linkage become increasingly close, railways, as a major livelihood project and the backbone of a comprehensive transportation system, play a crucial role in economic and social development. By the end of 2024, my country's railway operating mileage will reach 162,000 kilometers, including 48,000 kilometers of high-speed rail, forming a large-scale and densely interwoven railway network.

[0003] The railway network is an open, complex, and vast system with typical complex network characteristics. Its topology exhibits a pronounced agglomeration effect, with key nodes responsible for core passenger and freight transport or for connecting and converting traffic gradually emerging. These nodes are susceptible to interference from both internal and external factors. Therefore, studying the importance characteristics of railway network nodes, which play a prominent role in the integrated transportation system, has important theoretical and practical implications for optimizing and improving the network's collection and distribution efficiency and further developing a resilient transportation system.

[0004] Existing research on node importance assessment in complex networks is mostly based on network characteristic measurement indicators, such as degree centrality, betweenness centrality, and closeness centrality. These studies lack comprehensive consideration of factors such as node location, edge weights, and the local influence of neighboring nodes, making it difficult to fully reflect the importance characteristics of nodes in actual railway networks in both the global and neighborhood contexts.

[0005] Some importance assessment methods based on the dynamic characteristics of nodes, such as the PageRank algorithm, which iteratively calculates node weights and outputs nodes with higher PageRank values, are generally parameter-sensitive and computationally complex. They also ignore other node attributes during the iterative selection process and are not suitable for large-scale, complex transportation networks. Furthermore, the classic K-shell graph partitioning algorithm globally assigns a k value to each node in the network. This algorithm offers targeted search for key nodes at the center of the network and high computational efficiency. However, when decomposing complex networks, the K-shell algorithm suffers from the drawbacks of being unable to distinguish the importance of nodes on the same level and resulting in a coarse-grained decomposition of the network.

[0006] Through the above analysis, the problems and defects of the existing technology are as follows:

[0007] Overall, existing methods for evaluating node importance in complex networks lack a comprehensive consideration of both the global importance and local attributes of nodes. Furthermore, there is a lack of a fast, targeted node importance evaluation method for complex railway transportation networks. Therefore, a new structure or control method is needed to address at least some of these issues. Summary of the Invention

[0008] In response to the above-mentioned problems, the present invention proposes a railway network node evaluation method that integrates global and neighborhood features. The specific scheme is as follows:

[0009] A railway network node importance evaluation method based on the fusion of global and neighborhood features includes the following steps:

[0010] S1: Obtain railway transportation network node evaluation data, wherein the railway transportation network node evaluation data includes railway station and connection relationship data and railway section transportation data;

[0011] S2: Constructing a railway transportation complex network model, wherein the railway transportation complex network model includes a railway transportation facility topology network and a railway transportation service complex network;

[0012] S3: analyzing the global importance of railway network nodes, wherein the analyzing the global importance of railway network nodes includes K-shell network decomposition and network node efficiency calculation;

[0013] S4: Outputting railway network node neighborhood attributes, wherein the outputting railway network node neighborhood attributes includes network neighborhood node secondary search, network node capability strength calculation, network node adjacency capability strength calculation, and network node adjacency information entropy calculation.

[0014] S5: fusing the global and neighborhood features of the railway network nodes, wherein fusing the global and neighborhood features of the railway network nodes includes setting a fusion factor of the global and neighborhood features and calculating a comprehensive importance index of the network nodes;

[0015] S6: Evaluate the importance of railway network nodes.

[0016] Preferably, step S1 includes the following processing steps:

[0017] S101: The railway station and connection relationship data are the railway stations in the area to be evaluated and their connection relationships, where:

[0018] V={v1,v2,v3,…,v m}, represents the railway station data set in the area to be evaluated, and the railway stations are the nodes in the corresponding network;

[0019] S102: The railway section transportation data is the passenger volume or freight volume between railway stations in the area to be evaluated. The data is obtained according to the railway passenger network or railway freight network to be evaluated, and the passenger volume or freight volume between railway stations in the area to be evaluated is organized into a matrix form.

[0020] Recorded as section transportation data matrix

[0021] Where: w ij Represents node v i To node v j The section passenger volume or freight volume, the railway section transportation data is used to determine the edge weights of the railway transportation complex network.

[0022] Preferably, in step S101, the connection relationship between the railway stations is represented by the adjacency matrix A m×m describe;

[0023] If the node v i To node v j If there is a connection between them, then the element in the i-th row and j-th column of the adjacency matrix is a ij =1;

[0024] If the node v i To node v j If there is no connection between them, then the element in row i and column j of the adjacency matrix is a ij =0, recorded as Where m = |V| is the number of railway stations in the area to be evaluated.

[0025] Preferably, in step S2, constructing a railway transportation complex network model includes the following steps:

[0026] S201: Based on the topological mapping method of the transportation network structure, using the Space L spatial modeling rule, based on the railway station data set V of the area to be evaluated obtained in step S1 and the adjacency matrix A m×m , establish the railway transportation facilities topology network RCN T =(V,E), where E={e1, e2, e3, ..., e n}, indicating that according to the adjacency matrix A m×m The generated network edge set, n = |E| is the number of edges in the railway transportation facility topology network.

[0027] S202: Railway Transport Facilities Topology Network (RCN) T As a carrier, combined with the railway section transportation data matrix W m×m , giving the actual passenger or freight flow direction to the railway transportation facility topology network, and superimposing weights on the corresponding edges to establish the railway transportation service complex network RCN S=(V,E,W);

[0028] If site v i To site v j There is a connection relationship (a ij =1), node v i To node v j The edge weight between them is assigned as w ij .

[0029] Further preferably, the established railway transportation facility topology network RCN T is an undirected unweighted graph, and the adjacency matrix A m×m is a symmetric square matrix; Railway transportation service complex network RCN S is a directed weighted graph, and the section transportation data matrix W m×m is an asymmetric matrix.

[0030] Preferably, in step S3, the specific steps of analyzing the global importance of railway network nodes are:

[0031] S301: The railway transportation facility topology network RCN established in step S201 T Perform K-shell algorithm decomposition and get each node v after decomposition i ks i The K-shell algorithm searches for key nodes in the network center from a global perspective, and obtains the ks value of the node to represent the node in the railway transportation facility topology network RCN. T levels within the whole;

[0032] S302: For the railway transportation facility topology network RCN established in step S201 T All nodes v in i , calculate its node efficiency where d ij For node v i To node v j The shortest path length is m, and m is the number of railway stations in the area to be evaluated.

[0033] Preferably, in step S4, outputting the neighborhood attributes of the railway network node includes the following steps:

[0034] S401: For the railway transportation service complex network RCN established in step S202 S Perform secondary search of neighboring nodes based on RCN S The topological connection relationship and the directionality of the connection between nodes, node v i The set of neighboring nodes connected to it during the first search is recorded as θ i ;

[0035] When vj ∈θ i When node v j The set of neighboring nodes connected to the second search is recorded as θ j , the railway transport service complex network RCN S All nodes in the node complete the secondary search of their neighboring nodes;

[0036] S402: Calculate the directed weighted network RCN S Node capability strength, node v i Node capability in-degree intensity and node capability out-degree intensity Using the in-degree strength of directed networks The influence coefficient κ is calculated after comprehensive calculation of node v i Node capability strength

[0037] where v i Node capability in-degree strength v i Node capability out-degree strength

[0038] S403: Calculate the directed weighted network RCN S Node adjacency strength, combined with the secondary search results of the network neighborhood nodes in step S401, node v i The adjacent node v j (v j ∈θ i ) Corresponding adjacency strength

[0039] S404: Calculate the directed weighted network RCN S Node adjacency information entropy uses a probability function to characterize the probability of an evaluation node being selected among its adjacent nodes, i.e., p ij For node v i In the network, the neighboring node v j The probability of selection;

[0040] According to step S402 and step S403, the directed weighted network RCN S Midpoint v i Adjacency Information Entropy in Calculating the Railway Transport Service Complex Network (RCN) S The adjacency information entropy of all nodes in .

[0041] Preferably, in step S5, fusing the global and neighborhood features of the railway network nodes comprises the following steps:

[0042] S501: Setting global and neighborhood feature fusion factors μ and η, which correspond to the global importance feature and neighborhood attribute feature of the network node respectively, and the feature fusion factor satisfies μ+η=1;

[0043] S502: Calculate the comprehensive importance index of the network node. Based on the global importance of the railway network node analyzed in step S3 and the neighborhood attributes of the railway network node output in step S4, combine the global and neighborhood feature fusion factors μ and η to output the comprehensive importance index of the railway transportation network node.

[0044] Node v in the network i The calculation method of the comprehensive importance index is

[0045] CI i =norm(ks i / K×ne i )+norm(H i ), where data standardization dimensionless processing is required before feature fusion, norm(·) refers to the standardization of the data set using the range normalization method, ks i For node v i K-shell value, K is the maximum ks after network decomposition i Value, ne i For node v i The efficiency value, H i For node v i The adjacency information entropy of node v i Comprehensive importance index CI i ∈[0,2];

[0046] Calculate the value of each node V={v1,v2,v3,…,v m}, and form a set of comprehensive importance indices of railway network nodes in the area to be evaluated

[0047] Preferably, in step S6, the importance of the railway network nodes is evaluated according to the comprehensive importance index set of the railway network nodes in the area to be evaluated calculated in step S502. The importance of network nodes is ranked. The larger the comprehensive importance index value is, the higher the comprehensive importance of the node in the overall centralization of the railway network and the influence of neighboring nodes.

[0048] Another object of the present invention is:

[0049] An information data processing terminal is installed on an electronic device and provides a user input interface to implement the railway network node importance evaluation method that integrates global and neighborhood features.

[0050] The beneficial effects of the present invention are:

[0051] 1. This invention integrates the global importance of railway network nodes with neighborhood attribute characteristics to perform importance assessment. Compared with traditional single network feature measurement indicators, its node importance assessment results are more objective and accurate. At the same time, the technical method is easy to implement and has strong generalization capabilities.

[0052] 2. This invention solves the problem of coarse-grained decomposition of complex networks in traditional algorithms and the inability to evaluate and rank the importance of nodes at the same level. By considering the global and neighborhood characteristics of network nodes, it effectively identifies key nodes with high importance in complex railway transportation networks.

[0053] 3. The present invention realizes the precise evaluation of the node importance of the complex railway transportation network from the actual operation dimension. The method has low computational complexity and high application evaluation efficiency. It provides methodological support and evaluation means for researchers to conduct scientific and efficient analysis of the importance of railway transportation network nodes, supports the improvement of the collection and distribution efficiency of the railway transportation network and the construction of a high-resilience transportation system, and has relatively broad application prospects. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] The technical solutions of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. However, it should be understood that these drawings are designed for illustrative purposes only and are not intended to limit the scope of the present invention. Furthermore, unless otherwise specified, these drawings are intended only to conceptually illustrate the structures described herein and are not necessarily drawn to scale.

[0055] Figure 1 Flowchart of the railway network node importance evaluation method integrating global and neighborhood features of the present invention;

[0056] Figure 2 A schematic diagram of a railway transportation facility topology network according to an application example of the present invention;

[0057] Figure 3 A schematic diagram of a complex network of railway transportation services according to an application example of the present invention;

[0058] Figure 4 This is an example diagram of calculating node adjacency information entropy in a simple directed weighted network according to an application example of the present invention;

[0059] Figure 5 This is a diagram showing the calculation results of the comprehensive importance index of railway network nodes in an application example of the present invention;

[0060] Figure 6 This is a distribution diagram of the top 20 nodes in the railway network ranked by comprehensive importance index in an application example of the present invention. DETAILED DESCRIPTION

[0061] First of all, it should be noted that the specific structure, features and advantages of the present invention will be described in detail below by way of example. However, all descriptions are for illustration only and should not be understood as limiting the present invention. In addition, any single technical feature described or implied in the embodiments mentioned herein, or any single technical feature shown or implied in the drawings, can still be combined or deleted between these technical features to obtain more other embodiments of the present invention that may not be directly mentioned herein. In addition, in order to simplify the drawings, the same or similar technical features may be marked in only one place in the same drawing.

[0062] In the present invention, unless otherwise clearly stipulated and limited, the terms "install", "set", "connect", "fix", "screw" and the like should be understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or an integrated connection; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium; it can be the internal connection of two elements or the interaction relationship between two elements. Unless otherwise clearly defined, ordinary technicians in this field can understand the specific meanings of the above terms in the present invention according to the specific circumstances.

[0063] The following is combined with Figure 1 -Attached Figure 6 The present invention will be described in detail.

[0064] Example 1:

[0065] like Figure 1 As shown in FIG, a railway network node importance evaluation method based on the fusion of global and neighborhood features includes the following steps:

[0066] S1: Obtain railway transportation network node evaluation data, wherein the railway transportation network node evaluation data includes railway station and connection relationship data and railway section transportation data;

[0067] S2: Constructing a railway transportation complex network model, wherein the railway transportation complex network model includes a railway transportation facility topology network and a railway transportation service complex network;

[0068] S3: analyzing the global importance of railway network nodes, wherein the analyzing the global importance of railway network nodes includes K-shell network decomposition and network node efficiency calculation;

[0069] S4: Outputting railway network node neighborhood attributes, wherein the outputting railway network node neighborhood attributes includes network neighborhood node secondary search, network node capability strength calculation, network node adjacency capability strength calculation, and network node adjacency information entropy calculation.

[0070] S5: fusing the global and neighborhood features of the railway network nodes, wherein fusing the global and neighborhood features of the railway network nodes includes setting a fusion factor of the global and neighborhood features and calculating a comprehensive importance index of the network nodes;

[0071] S6: Evaluate the importance of railway network nodes.

[0072] Furthermore, in the embodiment, it can also be considered that step S1 includes the following processing steps:

[0073] S101: The railway station and connection relationship data are the railway stations in the area to be evaluated and their connection relationships, where:

[0074] V={v1,v2,v3,…,v m}, represents the railway station data set in the area to be evaluated, and the railway stations are the nodes in the corresponding network;

[0075] S102: The railway section transport data is the passenger volume or freight volume between railway stations in the area to be evaluated. The data is obtained according to the railway passenger network or railway freight network of the evaluation object, and the passenger volume or freight volume between railway stations in the area to be evaluated is organized into a matrix form, which is recorded as the section transport data matrix

[0076]

[0077] Where: w ij Represents node v i To node v j The section passenger volume or freight volume, the railway section transportation data is used to determine the edge weights of the railway transportation complex network.

[0078] Furthermore, in the embodiment, it can also be considered that in step S101, the connection relationship between the railway stations is represented by the adjacency matrix A m×m describe;

[0079] If the node v i To node v j If there is a connection between them, then the element in the i-th row and j-th column of the adjacency matrix is a ij =1;

[0080] If the node v i To node v j If there is no connection between them, then the element in row i and column j of the adjacency matrix is a ij =0, recorded as Where m = |V| is the number of railway stations in the area to be evaluated.

[0081] Furthermore, in the embodiment, it can also be considered that in step S2, constructing the railway transportation complex network model includes the following steps:

[0082] S201: Based on the topological mapping method of the transportation network structure, using the Space L spatial modeling rule, based on the railway station data set V of the area to be evaluated obtained in step S1 and the adjacency matrix A m×m , establish the railway transportation facilities topology network RCN T =(V,E), where E={e1, e2, e3, ..., e n}, indicating that according to the adjacency matrix A m×m The generated network edge set, n = |E| is the number of edges in the railway transportation facility topology network.

[0083] S202: Railway Transport Facilities Topology Network (RCN) T As a carrier, combined with the railway section transportation data matrix W m×m , giving the actual passenger or freight flow direction to the railway transportation facility topology network, and superimposing weights on the corresponding edges to establish the railway transportation service complex network RCN S =(V,E,W);

[0084] If site v i To site v j There is a connection relationship (a ij =1), node v i To node v j The edge weight between them is assigned as w ij .

[0085] Furthermore, in the embodiment, it is also possible to consider establishing a railway transportation facility topology network RCN T is an undirected unweighted graph, and the adjacency matrix A m×m is a symmetric square matrix; Railway transportation service complex network RCN S is a directed weighted graph, and the section transportation data matrix W m×m is an asymmetric matrix.

[0086] In this embodiment, the railway transport service complex network established is a directed weighted graph. Due to the directionality of railway section transportation and the difference in uplink and downlink transportation volume, the railway transport service complex network RCN S Medium general w ij ≠w ji , that is, W m×m is an asymmetric matrix.

[0087] Furthermore, in the embodiment, it can also be considered that in step S3, the specific steps of analyzing the global importance of railway network nodes are:

[0088] S301: The railway transportation facility topology network RCN established in step S201 TPerform K-shell algorithm decomposition and get each node v after decomposition i ks i The K-shell algorithm searches for key nodes in the network center from a global perspective, and obtains the ks value of the node to represent the node in the railway transportation facility topology network RCN. T levels within the whole;

[0089] S302: For the railway transportation facility topology network RCN established in step S201 T All nodes v in i , calculate its node efficiency where d ij For node v i To node v j The shortest path length is m, and m is the number of railway stations in the area to be evaluated.

[0090] In this embodiment, the path length between nodes in the topological network is the number of edges of the shortest path connecting the nodes, and the node efficiency ne i Reflect node v i Reach Network RCN T The difficulty of other nodes in the node v i In the network RCN T The centralization level in the network, the greater the efficiency value of the node, the more important it is in the network RCN T The higher the centrality, the more important the node is in the global network.

[0091] Network RCN T The decomposition process of the K-shell algorithm is shown in Table 1, that is, for the node v in the network i Decompose the degree until each node v in the network i Give it ks i value.

[0092] Table 1 K-shell decomposition of railway transportation facility topology network

[0093]

[0094] Furthermore, in the embodiment, it can also be considered that in step S4, outputting the neighborhood attributes of the railway network node includes the following steps:

[0095] S401: For the railway transportation service complex network RCN established in step S202 S Perform secondary search of neighboring nodes based on RCN S The topological connection relationship and the directionality of the connection between nodes, node v i The set of neighboring nodes connected to it during the first search is recorded as θi ;

[0096] When v j ∈θ i When node v j The set of neighboring nodes connected to the second search is recorded as θ j , the railway transport service complex network RCN S All nodes in the node complete the secondary search of their neighboring nodes;

[0097] S402: Calculate the directed weighted network RCN S Node capability strength, node v i Node capability in-degree intensity and node capability out-degree intensity Using the in-degree strength of directed networks The influence coefficient κ is calculated after comprehensive calculation of node v i Node capability strength

[0098] where v i Node capability in-degree strength v i Node capability out-degree strength

[0099] S403: Calculate the directed weighted network RCN S Node adjacency strength, combined with the secondary search results of the network neighborhood nodes in step S401, node v i The adjacent node v j (v j ∈θ i ) Corresponding adjacency strength

[0100] S404: Calculate the directed weighted network RCN S Node adjacency information entropy uses a probability function to characterize the probability of an evaluation node being selected among its adjacent nodes, i.e., p ij For node v i In the network, the neighboring node v j The probability of selection;

[0101] According to step S402 and step S403, the directed weighted network RCN S Midpoint v i Adjacency Information Entropy in Calculating the Railway Transport Service Complex Network (RCN) S The adjacency information entropy of all nodes in .

[0102] In this embodiment, information entropy is used to measure the uncertainty of random variables. The higher the value, the greater the uncertainty of information. Information entropy can characterize the disorder of the system by using probability and statistical methods based on the uncertainty of the system. The information entropy of discrete random variables X is Where X is a random variable; p(x) represents the probability function of the random variable distribution; i is the base of the logarithm. When the base i = 2, the unit of information is bit. At this time, the calculated information entropy H(X) is the Shannon entropy. In the calculation of information entropy, 0log0 = 0 is specified.

[0103] In order to better use information entropy to measure the importance of complex network nodes, a probability function is used to characterize the probability of the evaluation node being selected among its adjacent nodes, that is, p ij For node v i In the network, the neighboring node v j The selection probability of a node is usually that of a node with large information entropy, which means that the connection pattern of the node is more complex, which is more conducive to the dissemination of information between nodes. That is, the greater the influence of the node, the node can be regarded as an important node in the network.

[0104] According to step S402 and step S403, the directed weighted network RCN S Midpoint v i Adjacency Information Entropy in The calculation process is shown in Table 2. The calculation of the railway transportation service complex network RCN S The adjacency information entropy of all nodes in .

[0105] Table 2 Calculation of node adjacency information entropy of railway transportation service complex network

[0106]

[0107]

[0108] Furthermore, in the embodiment, it can also be considered that in step S5, fusing the global and neighborhood features of the railway network node includes the following steps:

[0109] S501: Setting global and neighborhood feature fusion factors μ and η, which correspond to the global importance feature and neighborhood attribute feature of the network node respectively, and the feature fusion factor satisfies μ+η=1;

[0110] S502: Calculate the comprehensive importance index of the network node. Based on the global importance of the railway network node analyzed in step S3 and the neighborhood attributes of the railway network node output in step S4, combine the global and neighborhood feature fusion factors μ and η to output the comprehensive importance index of the railway transportation network node.

[0111] Node v in the network iThe calculation method of the comprehensive importance index is

[0112] CI i =norm(ks i / K×ne i )+norm(H i ), where data standardization dimensionless processing is required before feature fusion, norm(·) refers to the standardization of the data set using the range normalization method, ks i For node v i K-shell value, K is the maximum ks after network decomposition i Value, ne i For node v i The efficiency value, H i For node v i The adjacency information entropy of node v i Comprehensive importance index CI i ∈[0,2];

[0113] Calculate the value of each node V={v1,v2,v3,…,v m}, and form a set of comprehensive importance indices of railway network nodes in the area to be evaluated

[0114] In this embodiment, norm(·) refers to normalizing a data set using the range normalization method, mapping the data to the range of 0-1. The dot (·) in the brackets represents an arbitrary expression, such as norm(X) represents normalizing X.

[0115] Furthermore, in the embodiment, it can be considered that in step S6, the importance of the railway network node is evaluated according to the comprehensive importance index set of the railway network nodes in the area to be evaluated calculated in step S502. The importance of network nodes is ranked. The larger the comprehensive importance index value is, the higher the comprehensive importance of the node in the overall centralization of the railway network and the influence of neighboring nodes.

[0116] In this embodiment, the larger the comprehensive importance index value is, the higher the comprehensive importance of the node in the overall centralization of the railway network and the influence of neighboring nodes is, thus achieving the goal of evaluating the importance of nodes in the complex railway transportation network.

[0117] The present invention proposes a railway network node importance assessment method that integrates global and neighborhood features. Compared with traditional single network feature measurement indicators, the node importance assessment results of this method are more objective and accurate. This invention solves the problems of coarse-grained decomposition of complex networks in traditional algorithms and the inability to perform importance assessment and ranking of nodes at the same level. By considering the global and neighborhood characteristics of network nodes, it effectively identifies key nodes with higher importance in the complex network of railway transportation, and at the same time accurately assesses the node importance of the complex network of railway transportation from the actual operation dimension. The method has low computational complexity and high application evaluation efficiency, supporting the construction of a highly resilient transportation system and the high-quality development of railway transportation in the context of a modern integrated three-dimensional transportation network.

[0118] Application Example 1:

[0119] In order to further illustrate the embodiment, the present invention is further illustrated by combining actual application examples. The specific steps of the present invention are described in detail using a railway network in a certain region as an example:

[0120] S1: Obtain railway transport network node assessment data, including railway station and connection relationship data, and railway section transport data. This embodiment obtains and organizes relevant data through data collection and research. The specific implementation process is as follows:

[0121] S101: Arrange the railway stations and their connection relationship data in the area to be evaluated, where V = {v1, v2, v3, ..., v m} represents the railway station data set in the area to be evaluated. Railway stations are nodes in the corresponding network. The adjacency matrix A m×m Describe the connection relationship between railway stations, denoted as Wherein, m=|V|=132, that is, the number of railway stations in the area to be evaluated in this embodiment is 132.

[0122] S102: Arrange the railway section transport data. In this embodiment, the railway network in the region to be evaluated mainly provides passenger services. Therefore, the passenger volume between railway stations in the region to be evaluated in the latest full year is arranged in a matrix form and recorded as the section transport data matrix. Where w ij Represents node v i To node v j Section passenger volume (unit: 10,000 people). Railway section transportation data is used to determine the edge weights of the railway transportation complex network.

[0123] S2: Construct a complex network model for railway transportation, and establish a railway transportation facility topology network and a railway transportation service complex network. The specific implementation process is as follows:

[0124] S201: Based on the topological mapping method of the transportation network structure, using the Space L spatial modeling rule, based on the railway station data set V (size: 132) obtained in S1 and the connection relationship data adjacency matrix A m×m =A 132×132 , establish the railway transportation facilities topology network RCN T =(V,E), where E={e1, e2, e3, ..., e n} indicates that according to the adjacency matrix A 132×132 The generated network edge set.

[0125] In this embodiment, the number of edges in the railway transportation facility topology network is n=|E|=167. The established railway transportation facility topology network is an undirected unweighted graph. The adjacency matrix A m×m It is a symmetrical square matrix.

[0126] like Figure 2 The railway transportation facility topology network constructed in this application example S201 is shown.

[0127] S202: Further develop the Railway Transport Facilities Topology Network (RCN) T As a carrier, combined with the railway section transportation data matrix W in S102 m×m =W 132×132 , giving the actual passenger flow direction to the railway transportation facility topology network, and superimposing weights on the corresponding edges to establish the railway transportation service complex network RCN S =(V,E,W).

[0128] The railway transport service complex network established is a directed weighted graph. Due to the directionality of railway section transportation and the difference in upstream and downstream transportation volume, the railway transport service complex network RCN S Medium general w ij ≠w ji , that is, W m×m is an asymmetric matrix.

[0129] like Figure 3 The railway transportation service complex network constructed in this application example S202 is given.

[0130] S3: Analyze the global importance of railway network nodes, perform K-shell network decomposition and calculate network node efficiency for the railway network in the evaluation area. The specific implementation process is as follows:

[0131] S301: The railway transportation facilities topology network RCN established in S201 T Perform K-shell algorithm decomposition according to the process in Table 1, and get each node v after decomposition. i ks i Value. That is, for the node v in the networki Perform degree decomposition until each node v in the network i Give it ks i Value, get the ks value of the node to represent the node in the railway transportation facility topology network RCN T The level in the whole, get the set of node ks values Φ = {1, 1, ..., ks i ,…}(size:132).

[0132] S302: Railway transportation facilities topology network RCN established in S201 T All nodes v in i , calculate its node efficiency where d ij For node v i To node v j The shortest path length is m=132.

[0133] In a topological network, the path length between nodes is the number of edges in the shortest path connecting the nodes, and the node efficiency ne i Reflect node v i Reach Network RCN T The difficulty of other nodes in the node v i In the network RCN T The centralization level in the network, the greater the efficiency value of the node, the more important it is in the network RCN T The higher the centrality, the higher the node efficiency value set Ψ={0.241,0.226,…,ne i ,…}(size:132), reflecting the importance of the node in the global network.

[0134] S4: Output the neighborhood attributes of railway network nodes, perform secondary search of network neighborhood nodes, calculate network node capability strength, calculate network node adjacency capability strength, and calculate network node adjacency information entropy for the railway network in the area to be evaluated. The specific implementation process is as follows:

[0135] S401: Railway transport service complex network (RCN) established in S202 S Perform secondary search of neighboring nodes based on RCN S The topological connection relationship and the directionality of the connection between nodes, node v i The set of neighboring nodes connected to it during the first search is recorded as θ i ;

[0136] When v j ∈θ i When node v j The set of neighboring nodes connected to the second search is recorded as θ j, the railway transport service complex network RCN S All nodes in the node complete the secondary search of their neighboring nodes;

[0137] S402: Calculate the directed weighted network RCN S Node capability strength, for RCN S Midpoint v i Node capability in-degree intensity and out-degree strength Calculate the node v by using the directed network in-degree strength influence coefficient κ i Node capability strength where v i Node capability in-degree strength v i Node capability out-degree strength The penetration strength influence coefficient κ=0.75.

[0138] S403: Calculate the directed weighted network RCN S Node adjacency strength, combined with the secondary search results of S401 network neighborhood nodes, node v i The adjacent node v j (v j ∈θ i ) Corresponding adjacency strength

[0139] S404: Calculate the directed weighted network RCN S Node adjacency information entropy uses a probability function to characterize the probability of an evaluation node being selected among its adjacent nodes, i.e., p ij For node v i In the network, the neighboring node v j The selection probability of . According to S4-2 and S4-3 directed weighted network RCN S Midpoint v i The calculation method of adjacency information entropy is: in

[0140] Calculate the railway transport service complex network RCN according to Table 2 S The adjacency information entropy of all nodes in the node is obtained, and the set of node adjacency information entropy values Π={3.039, 2.549, …, H i ,…}(size:132).

[0141] like Figure 4 An example of calculating node adjacency information entropy in a simple directed weighted network is shown.

[0142] S5: Integrate the global and neighborhood characteristics of railway network nodes, set the global and neighborhood characteristic fusion factors respectively, and calculate the comprehensive importance index of network nodes. The specific implementation process is as follows:

[0143] S501: Set the global and neighborhood feature fusion factors μ=0.25 and η=0.75, corresponding to the global importance feature and neighborhood attribute feature of the network node respectively, and the feature fusion factor satisfies μ+η=1.

[0144] S502: Calculate the comprehensive importance index of network nodes. According to the global importance of railway network nodes analyzed in S3 and the neighborhood attributes of railway network nodes output in S4, the comprehensive importance index of railway transportation network nodes is output by combining the global and neighborhood feature fusion factors μ and η. Node v i Comprehensive importance index CI i =norm(ks i / K×ne i )+norm(H i ), where data standardization dimensionless processing is required before feature fusion, norm(·) refers to the use of range normalization method to standardize the data set and map the data to the range of 0-1, ks i For node v i K-shell value, K is the maximum ks after network decomposition i Value, ne i For node v i The efficiency value, H i For node v i The node ks value set Φ, node efficiency value set Ψ, and node adjacency information entropy value set Π in the embodiment are standardized, and the node v i Comprehensive importance index CI i ∈[0,2], calculate the value of each node V={v1,v2,v3,…,v m}(m=132), and the railway network node comprehensive importance index set is obtained

[0145] like Figure 5 The calculation results of the comprehensive importance index of the railway network nodes in the area to be evaluated in this application example are shown.

[0146] S6: Evaluate the importance of railway network nodes, based on the comprehensive importance index set of railway network nodes in the area to be evaluated calculated in S502 The importance of network nodes is ranked, that is, the larger the comprehensive importance index value is, the higher the comprehensive importance of the node in the overall centralization of the railway network and the influence of neighboring nodes.

[0147] like Figure 6 The distribution of the top 20 nodes ranked by comprehensive importance index of the railway network nodes in the area to be evaluated in this application example is shown, achieving the goal of evaluating the importance of complex railway transportation networks.

[0148] The above embodiments describe the present invention in detail, but the contents described are only preferred embodiments of the present invention and should not be considered to limit the scope of the present invention. All equivalent changes and improvements made within the scope of the present invention should still fall within the scope of the patent coverage of the present invention.

Claims

1. A railway network node importance evaluation method based on the fusion of global and neighborhood features, characterized by: The following steps are involved: S1: Obtain railway transportation network node evaluation data, wherein the railway transportation network node evaluation data includes railway station and connection relationship data and railway section transportation data; S2: Constructing a railway transportation complex network model, wherein the railway transportation complex network model includes a railway transportation facility topology network and a railway transportation service complex network; S3: analyzing the global importance of railway network nodes, wherein the analyzing the global importance of railway network nodes includes K-shell network decomposition and network node efficiency calculation; S4: Outputting railway network node neighborhood attributes, wherein the outputting railway network node neighborhood attributes includes network neighborhood node secondary search, network node capability strength calculation, network node adjacency capability strength calculation, and network node adjacency information entropy calculation; S5: fusing the global and neighborhood features of the railway network nodes, wherein fusing the global and neighborhood features of the railway network nodes includes setting a fusion factor of the global and neighborhood features and calculating a comprehensive importance index of the network nodes; S6: Evaluate the importance of railway network nodes.

2. The railway network node importance assessment method based on the fusion of global and neighborhood features according to claim 1 is characterized by: Step S1 includes the following processing steps: S101: The railway station and connection relationship data are the railway stations in the area to be evaluated and their connection relationships, where: V={v1,v2,v3,…,v m }, represents the railway station data set in the area to be evaluated, and the railway stations are the nodes in the corresponding network; S102: The railway section transport data is the passenger volume or freight volume between railway stations in the area to be evaluated. The data is obtained according to the railway passenger network or railway freight network of the evaluation object, and the passenger volume or freight volume between railway stations in the area to be evaluated is organized into a matrix form, which is recorded as the section transport data matrix Where: w ij Represents node v i To node v j The section passenger volume or freight volume, the railway section transportation data is used to determine the edge weights of the railway transportation complex network.

3. The railway network node importance assessment method based on the fusion of global and neighborhood features according to claim 2 is characterized by: In step S101, the connection relationship between the railway stations is represented by the adjacency matrix A m×m describe; If the node v i To node v j If there is a connection between them, then the element in the i-th row and j-th column of the adjacency matrix is a ij =1; If the node v i To node v j If there is no connection between them, then the element in row i and column j of the adjacency matrix is a ij =0, recorded as Where m = |V| is the number of railway stations in the area to be evaluated.

4. The railway network node importance assessment method based on the fusion of global and neighborhood features according to claim 1 is characterized by: In step S2, constructing a railway transportation complex network model includes the following steps: S201: Based on the topological mapping method of the transportation network structure, using the Space L spatial modeling rule, based on the railway station data set V of the area to be evaluated obtained in step S1 and the adjacency matrix A m×m , establish the railway transportation facilities topology network RCN T =(V,E), where E={e1, e2, e3, ..., e n }, indicating that according to the adjacency matrix A m×m The generated network edge set, n = |E| is the number of edges in the railway transportation facility topology network. S202: Railway Transport Facilities Topology Network (RCN) T As a carrier, combined with the railway section transportation data matrix W m×m , giving the actual passenger or freight flow direction to the railway transportation facility topology network, and superimposing weights on the corresponding edges to establish the railway transportation service complex network RCN S =(V,E,W); If site v i To site v j There is a connection relationship (a ij =1), node v i To node v j The edge weight between them is assigned as w ij .

5. The railway network node importance assessment method based on the fusion of global and neighborhood features according to claim 4 is characterized by: Railway Transportation Facilities Topology Network (RCN) T is an undirected unweighted graph, and the adjacency matrix A m×m is a symmetric square matrix; Railway transportation service complex network RCN S is a directed weighted graph, and the section transportation data matrix W m×m is an asymmetric matrix.

6. The railway network node importance assessment method based on the fusion of global and neighborhood features according to claim 5 is characterized by: In step S3, the specific steps for analyzing the global importance of railway network nodes are as follows: S301: The railway transportation facility topology network RCN established in step S201 T Perform K-shell algorithm decomposition and get each node v after decomposition i ks i The K-shell algorithm searches for key nodes in the network center from a global perspective, and obtains the ks value of the node to represent the node in the railway transportation facility topology network RCN. T levels within the whole; S302: For the railway transportation facility topology network RCN established in step S201 T All nodes v in i , calculate its node efficiency where d ij For node v i To node v j The shortest path length is m, and m is the number of railway stations in the area to be evaluated.

7. The railway network node importance evaluation method based on the fusion of global and neighborhood features according to claim 5 is characterized in that: In step S4, outputting the neighborhood attributes of the railway network nodes includes the following steps: S401: For the railway transportation service complex network RCN established in step S202 S Perform secondary search of neighboring nodes based on RCN S The topological connection relationship and the directionality of the connection between nodes, node v i The set of neighboring nodes connected to it during the first search is recorded as θ i ; When v j ∈θ i When node v j The set of neighboring nodes connected to the second search is recorded as θ j , the railway transport service complex network RCN S All nodes in the node complete the secondary search of their neighboring nodes; S402: Calculate the directed weighted network RCN S Node capability strength, node v i Node capability in-degree intensity and node capacity out-degree intensity Using the in-degree strength of directed networks The influence coefficient κ is calculated after comprehensive calculation of node v i Node capability strength where v i Node capability in-degree strength v i Node capability out-degree strength S403: Calculate the directed weighted network RCN S Node adjacency strength, combined with the secondary search results of the network neighborhood nodes in step S401, node v i The adjacent node v j (v j ∈θ i ) Corresponding adjacency strength S404: Calculate the directed weighted network RCN S Node adjacency information entropy uses a probability function to characterize the probability of an evaluation node being selected among its adjacent nodes, i.e., p ij For node v i In the network, the neighboring node v j The probability of selection; According to step S402 and step S403, the directed weighted network RCN S Midpoint v i Adjacency Information Entropy in Calculating the Railway Transport Service Complex Network (RCN) S The adjacency information entropy of all nodes in .

8. The railway network node importance assessment method based on the fusion of global and neighborhood features according to claim 1 is characterized by: In step S5, fusing the global and neighborhood features of railway network nodes includes the following steps: S501: Setting global and neighborhood feature fusion factors μ and η, which correspond to the global importance feature and neighborhood attribute feature of the network node respectively, and the feature fusion factor satisfies μ+η=1; S502: Calculate the comprehensive importance index of the network node. Based on the global importance of the railway network node analyzed in step S3 and the neighborhood attributes of the railway network node output in step S4, combine the global and neighborhood feature fusion factors μ and η to output the comprehensive importance index of the railway transportation network node. Node v in the network i The calculation method of the comprehensive importance index is CI i =norm(ks i / K×ne i )+norm(H i ), where data standardization dimensionless processing is required before feature fusion, norm(·) refers to the standardization of the data set using the range normalization method, ks i For node v i K-shell value, K is the maximum ks after network decomposition i Value, ne i For node v i The efficiency value, H i For node v i The adjacency information entropy of node v i Comprehensive importance index CI i ∈[0,2]; Calculate the value of each node V={v1,v2,v3,…,v m }, and organize them into a set of comprehensive importance indices of railway network nodes in the area to be evaluated 9. The railway network node importance assessment method based on the fusion of global and neighborhood features according to claim 1 is characterized by: In step S6, the importance of railway network nodes is evaluated based on the comprehensive importance index set of railway network nodes in the area to be evaluated calculated in step S502. The importance of network nodes is ranked. The larger the comprehensive importance index value is, the higher the comprehensive importance of the node in the overall centralization of the railway network and the influence of neighboring nodes.

10. An information data processing terminal, characterized in that: The information data processing terminal is installed on an electronic device and provides a user input interface to implement the railway network node importance evaluation method based on the fusion of global and neighborhood features as described in any one of claims 1 to 9.

Images