Railway station key track section identification method based on multilayer coupling network

Abstract

The invention provides a railway station key track section identification method based on a multilayer coupling network. The method comprises the following steps: respectively constructing sub-networks of a train route layer network, a continuation route layer network, a shunting route layer network and the like according to a route list in a station interlocking chart, and constructing a multi-layer coupling network by utilizing the plurality of sub-networks; calculating a multi-dimensional index score for each track section in the multi-layer coupling network, performing normalization and weighted fusion on the index scores of each track section by applying a multi-index decision-making method, and calculating a comprehensive score of each track structure; and determining the track section with the top comprehensive score as a key track section. According to the method provided by the invention, by introducing an inter-layer coupled unified network model and a richer evaluation index system and combining with a multi-index fusion decision, more comprehensive and accurate identification of a key track section and robustness evaluation of a fault scene are realized, so that the structure expression ability of the model and the practicability of an identification result are improved.

Description

Technical Field

[0001] This invention relates to the field of track section identification technology, and in particular to a method for identifying key track sections in railway stations based on a multi-layer coupled network. Background Technology

[0002] Traditional railway station route models typically extract station components and their connections based on interlocking charts (also known as control tables), treating the station as a topological structure composed of a set of interconnected parts. Early research proposed station topology models based on component connections, obtaining the set of permissible routes and signal equipment status through interlocking charts. Subsequently, methods for automatically locating and verifying routes using matrix algorithms, and methods for calculating utilization rates by statistically analyzing track occupancy times, have also emerged. However, most of these studies are limited to abstracting the station into a single-layer network structure, lacking analysis of the interactions between different types of operations (such as normal train operation, continuous routes, and shunting operations), and particularly lacking methods for quantitatively identifying high-occupancy track sections.

[0003] With the development of network science, complex network theory has been introduced into transportation research. Traditional single-layer network models can only capture characteristics from a single perspective (such as connectivity strength or flow), ignoring the inherent multidimensional attributes and inter-layer dependencies in infrastructure. In contrast, multi-layer (multi-level) network frameworks can explicitly represent cross-layer interactions and dependencies, providing richer structural insights for analyzing complex transportation infrastructure. In railway systems, multi-layer network modeling has begun to be used to assess station throughput capacity, node importance, and other parameters. For example, some studies have constructed multi-layer network models including train routes, continuation routes, and shunting operations to analyze station performance, demonstrating that multi-layer coupled networks can significantly improve the accuracy and comprehensiveness of node importance assessment.

[0004] One existing method for identifying critical track sections in railway stations based on a multi-network architecture involves constructing a station network model with three operational layers using interlocking chart data: the Train Routes (TR) layer represents the route connections for normal train operations, the Continuation Routes (ER) layer represents the continued occupancy relationships after a train passes through a station, and the Shunting Routes (SR) layer represents the route relationships for shunting operations. The three layers share the same set of nodes (track sections), and within each layer, track sections sharing adjacent relationships are connected by edges according to the corresponding route type. The literature evaluates the importance of each track section in each layer using various classical centrality indices, including degree centrality (DC), proximity centrality (CC), betweenness centrality (BC), eigenvector centrality (PR), and Kernel-Kind centrality (KC). Furthermore, a customized index for railway stations, Fusion Centrality (FC), is introduced to comprehensively characterize the connection density and connectivity efficiency of nodes. This method addresses the conflicting ranking results that different indicators may provide and employs a multi-indicator decision-making approach to comprehensively evaluate the importance of track segments: First, it uses PCA (Principal Component Analysis) to reduce dimensionality and extract the main components of multiple centrality indicators. Then, it uses the Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) to rank the comprehensive scores after dimensionality reduction.

[0005] The disadvantages of a current method for identifying critical track sections at stations based on multiple networks include:

[0006] (1) Lack of connection between model layers: This method models train routes, continuation routes, and shunting routes as independent three-layer networks, with no direct connection edges between layers. The connection is only indirectly reflected in subsequent analysis through the correlation of indicators of the same node. This loosely coupled model structure does not fully consider the real-time mutual influence between different route types, making it difficult to simulate the transmission of one type of operation failure to other operations in a timely manner.

[0007] (2) Limitations of the indicator system: The indicator used in this method is mainly classical topological centrality. Although the bias is mitigated by fusion, these indicators still focus on the static topology of the network and do not fully quantify the importance of track segments in dynamic usage and failure scenarios. For example, the literature does not involve the evaluation of redundant paths in track segments, and cannot reflect the availability and capability of alternative paths when a certain segment is blocked.

[0008] (3) Lack of failure impact assessment: This method assumes all infrastructure is intact and evaluates based on static efficiency, without simulating route reconfiguration or interruption in the event of track segment failure. Therefore, it cannot verify whether the identified critical segments actually have a significant impact on the overall operation in real failure scenarios. Summary of the Invention

[0009] The embodiments of the present invention provide a method for identifying critical track sections in railway stations based on a multi-layer coupled network, so as to effectively identify critical track sections.

[0010] To achieve the above objectives, the present invention adopts the following technical solution.

[0011] A method for identifying critical track sections in railway stations based on multi-layer coupled networks, comprising:

[0012] The configuration data of track sections, turnouts, signals, and all train routes, extension routes, and shunting routes are extracted from the station interlocking chart, and the collected data is preprocessed and verified.

[0013] Based on the route list in the station interlocking chart, construct the train route layer network, the continuation route layer network, and the shunting route layer network sub-networks respectively, and use multiple sub-networks to construct a multi-layer coupled network;

[0014] For each track segment in the multi-layer coupled network, multi-dimensional index scores are calculated. A multi-index decision method is applied to normalize and weightedly fuse the multiple index scores of each track segment to calculate the comprehensive score of each track structure.

[0015] The track sections with the highest overall scores are identified as critical track sections.

[0016] Preferably, the step of extracting configuration data for track sections, turnouts, signals, and all train routes, extension routes, and shunting routes from the station interlocking chart, and preprocessing and verifying the collected data, includes:

[0017] Obtain the station interlocking chart, which includes track section number and location, turnout number and turning status, signal type and control range, as well as configuration data for all train routes, continuation routes and shunting routes, and clearly records the interlocking relationships between equipment;

[0018] The configuration data of track sections, turnouts, signals, and all train routes, extension routes, and shunting routes are extracted from the station interlocking chart, and the collected data is preprocessed and verified.

[0019] Preferably, the construction of train route layer network, continuous route layer network, and shunting route layer network sub-networks based on the route list in the station interlocking chart, and the construction of a multi-layer coupled network using multiple sub-networks, includes:

[0020] Based on the route list in the station interlocking chart, train route layer sub-network, continuation route layer sub-network, and shunting route layer sub-network are constructed respectively. Using the train route layer network, continuation route layer network, and shunting route layer network, a multi-layer coupling network is constructed. This multi-layer coupling network simultaneously presents three logical relationships: train route, continuation route, and shunting route.

[0021] Each sub-network in the multi-layer coupled network uses multiple different track segments as nodes and the sequential connection relationship between track segments as edges. Each sub-network contains: track segments, directed edges between nodes, logical positions of nodes, and naming information. Nodes corresponding to the same track segment in different sub-networks are considered to be interconnected.

[0022] Preferably, the calculation of multi-dimensional index scores for each track segment in the multi-layer coupled network includes:

[0023] For each track segment in a multi-layer coupled network, a multi-dimensional index score is calculated. The multi-dimensional index includes: connection frequency F, degree centrality D, in-degree centrality I, out-degree centrality O, proximity centrality C, betweenness centrality B, Katz centrality K, and improved PageRank centrality P.

[0024] The connection frequency F of a node track segment represents the number of times the node track segment appears in all interlocking charts. This indicator is a benefit-oriented indicator, and its calculation formula is as follows:

[0025]

[0026] in: Z represents the number of times node track segment i appears in the interlocking graph corresponding to the m-th layer network; + V represents a positive integer, V represents the set of nodes, and m represents the m-th layer of the network;

[0027] The degree centrality D of a node's track segment represents the density of connections between its surrounding neighboring nodes. This indicator is a benefit-oriented indicator, and its calculation formula is as follows:

[0028] D i =I i +O i

[0029] Where: i represents the node track segment; D i I represents the degree centrality of a node. i Indicates the in-degree centrality of a node, O iThis indicates the out-degree centrality of a node;

[0030] The in-degree centrality I of a node track segment represents the connection between other track segments and this track segment. This indicator is a benefit-oriented indicator, and its calculation formula is as follows:

[0031]

[0032] Where: i represents the node track segment; AM (m) Let N represent the adjacency matrix of the m-th layer network; N represents the number of nodes; and V represents the set of nodes.

[0033] The out-degree centrality O of a node track segment represents the external connectivity of that node track segment. This indicator is a benefit-oriented indicator, and its calculation formula is as follows:

[0034]

[0035] Where: i represents the node track segment; AM (m) Let N represent the adjacency matrix of the m-th layer network; N represents the number of nodes; and V represents the set of nodes.

[0036] The proximity centrality C of a node's track segment represents the ease with which the track segment of this node can reach other nodes in the network. This indicator is a benefit-oriented indicator, and its calculation formula is as follows:

[0037]

[0038] Where: n-1 represents the number of nodes reachable from node i in the m-th layer network; N represents the shortest path distance between node segment j and node i in the m-th layer network; N represents the number of nodes; V represents the set of nodes.

[0039] The betweenness centrality B of a nodal track segment represents the mediating capacity of that nodal track segment. This indicator is a benefit-oriented indicator, and its calculation formula is as follows:

[0040]

[0041] in: This represents the number of shortest paths from node track segment s to t in the m-th layer network; V represents the number of shortest paths passing through node i; V represents the set of nodes.

[0042] The Katz centrality K of a node's orbital segment measures the relative impact of network nodes by considering neighboring nodes and all other nodes reached through these neighbors. This metric is a benefit-based indicator, and its calculation formula is as follows:

[0043]

[0044] Where α represents the attenuation factor, and satisfies α < 1 / λ MAX ;λ MAX AM (m) Maximum eigenvalue; AM (m) Let represent the adjacency matrix of the m-th layer network; β represent the adjustment coefficient; N represent the number of nodes; and V represent the node set.

[0045] The improved PageRank centrality P of node track segments quantifies the importance of nodes in the network through iterative distribution influenced by connectivity and link structure. This metric is a benefit-based indicator, and its calculation formula is as follows:

[0046]

[0047] Where: μ represents the damping coefficient, AM (m) Let represent the adjacency matrix of the m-th layer network; e represents an N×N matrix with all elements equal to 1; N represents the number of nodes. 5. The method according to claim 4, characterized in that the application of the multi-index decision-making method normalizes and weights the scores of multiple indices for each track segment to calculate the comprehensive score of each track structure, including:

[0048] Multiple indicators from the following criteria are selected as decision indicators: connection frequency F, degree centrality D, in-degree centrality I, out-degree centrality O, proximity centrality C, betweenness centrality B, Katz centrality K, and improved PageRank centrality P. A multi-indicator decision method is applied to normalize and weightedly fuse multiple indicators for each track segment to calculate the comprehensive score for each track structure.

[0049] The processing flow of the multi-index decision-making method includes:

[0050] Obtain the input parameters, including the number of M network layers, the set of V nodes, the number of N nodes, and the number of Q importance evaluation indicators. Iterate through and sequentially select the q-th decision indicator.

[0051] Determine whether the indicator type is a benefit indicator or a cost indicator, and write the determination result into the indicator type matrix T:

[0052] T=(t q ) 1×Q =[t F t D t I t O t C t B t K t P …]

[0053]

[0054] Calculate the comprehensive evaluation index of each layer of the network:

[0055] 4. Traverse and sequentially select the m-th layer of the network.

[0056] 5. Construct the multi-attribute initial decision matrix D (m )

[0057] a) Initialize D (m) It is an N*Q zero matrix;

[0058] b) Iterate through the q-th decision index value of each node i.

[0059] c) Read the type T of decision indicator q q If T q =1, then Unchanged; if T q =-1, then

[0060] d) Obtain the multi-attribute matrix D (m) ,like:

[0061]

[0062] 6. Calculate the weights of each indicator using the entropy weight method.

[0063] a) Traverse and sequentially select the original decision matrix D of the m-th layer network. (m )

[0064] b) Regarding D (m) Standardization is performed to eliminate differences in the dimensions of the indicators.

[0065] The decision matrix D is analyzed using the proportional normalization method. (m) Standardization is performed for the q-th decision.

[0066] The formula for calculating the proportion of each unit in the policy indicator is as follows:

[0067]

[0068] In the formula: D represents the initial decision matrix with multiple attributes. (m) The original data in the i-th row and q-th position;

[0069] That is, the standardized data, and when At that time, it was agreed

[0070] c) Calculate the entropy value of each indicator.

[0071] For each decision indicator q, its information entropy is calculated based on the standardization results, and the calculation formula is...

[0072] The formula is:

[0073]

[0074] In the formula: It reflects the dispersion of the q-th index across all decision-making units, where k = 1 / ln(N) is a constant to ensure that the entropy value is normalized to the range of [0,1].

[0075] d) Calculate the difference coefficient for each decision indicator.

[0076] Define the coefficient of variation for decision indicators as:

[0077]

[0078] In the formula: This reflects the degree of contribution of the q-th decision indicator to the decision. The larger the value, the stronger the discriminative power of the indicator;

[0079] e) Calculate the weight of each decision indicator based on the difference coefficient of each decision indicator.

[0080] The objective weights of each indicator are determined based on the inverse principle of the coefficient of variation, and the calculation formula is as follows:

[0081]

[0082] In the formula: Let represent the weight of the q-th decision indicator in the m-th layer network, satisfying ...

[0083]

[0084] Calculate the comprehensive evaluation value of each node in each layer:

[0085] f) Traverse and sequentially select the m-th layer network

[0086] g) Traverse and sequentially select the multi-attribute matrix D of the i-th node. (m )

[0087] h) Calculate the overall network evaluation index for each node. The calculation formula is as follows:

[0088]

[0089] In the formula: This represents the comprehensive evaluation value of all nodes in the m-th layer network; This represents the comprehensive evaluation index of the i-th node in the m-th layer network; Represents the multi-attribute decision matrix D (m) The score data of the i-th node and the q-th indicator; This indicates the weight corresponding to the indicator q;

[0090] Calculation of comprehensive evaluation index for multi-layer networks

[0091] 2. Construct a multi-attribute matrix A

[0092] a) Initialize A as an N*M zero matrix;

[0093] b) Traverse the m-th layer importance evaluation index value of each node i

[0094] c) Obtain the multi-attribute matrix (D) N*P ),like:

[0095]

[0096] 4. Determine the weights of each layer of the network, assuming they are...

[0097] ω=[ω1ω2…ω M ]

[0098] 5. The comprehensive evaluation index of the node can be calculated using the following formula:

[0099]

[0100] Where: LEVEL i This represents the final comprehensive evaluation score of node i.

[0101] Preferably, the step of identifying the track segments with higher overall scores as critical track segments includes:

[0102] The comprehensive evaluation scores of each track segment are ranked, and a set number of track segments with the highest comprehensive evaluation scores are identified as critical track segments.

[0103] As can be seen from the technical solutions provided by the embodiments of the present invention above, the present invention provides an improved method for multi-layer network modeling of stations and identification of critical track sections. In view of the shortcomings of the prior art, by introducing a unified network model with inter-layer coupling and a richer evaluation index system, and combining multi-index fusion decision-making, the present invention achieves more comprehensive and accurate identification of critical track sections and robust assessment of fault scenarios, thereby improving the structural expressiveness of the model and the practicality of the identification results.

[0104] Additional aspects and advantages of the invention will be set forth in part in the description which follows, and will become apparent from the description or may be learned by practice of the invention. Attached Figure Description

[0105] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0106] Figure 1 A flowchart illustrating a method for identifying critical track sections in railway stations based on a multi-layer coupled network, provided in an embodiment of the present invention.

[0107] Figure 2 This is a schematic diagram of a multi-layer network structure for a station, provided as an embodiment of the present invention. Detailed Implementation

[0108] Embodiments of the present invention are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and should not be construed as limiting the present invention.

[0109] Those skilled in the art will understand that, unless specifically stated otherwise, the singular forms “a,” “an,” “the,” and “the” used herein may also include the plural forms. It should be further understood that the term “comprising” as used in this specification means the presence of the stated features, integers, steps, operations, elements, and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or groups thereof. It should be understood that when we say an element is “connected” or “coupled” to another element, it can be directly connected or coupled to the other element, or there may be intermediate elements. Furthermore, “connected” or “coupled” as used herein can include wireless connections or couplings. The term “and / or” as used herein includes any and all combinations of one or more of the associated listed items.

[0110] It will be understood by those skilled in the art that, unless otherwise defined, all terms used herein (including technical and scientific terms) have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. It should also be understood that terms such as those defined in general dictionaries should be understood to have the same meaning as in the context of the prior art, and should not be interpreted in an idealized or overly formal sense unless defined as herein.

[0111] To facilitate understanding of the embodiments of the present invention, the following will provide further explanation and description with reference to the accompanying drawings and several specific embodiments. These embodiments do not constitute a limitation on the embodiments of the present invention.

[0112] This invention adds inter-layer coupling connections to a multi-layer coupled network (i.e., realizes the coupling relationship mapping of corresponding nodes in different layers through virtual connections), directly connecting cross-layer nodes corresponding to the same track segment to form a unified network structure, thereby enabling the unified representation of multiple types of path logic and inter-layer dependencies.

[0113] This invention introduces new evaluation metrics such as topological entropy and alternative path ratio. Topological entropy measures the complexity and uncertainty of track segment connections, reflecting the global role of nodes in the network structure. The alternative path ratio assesses the proportion of routes that originally passed through a track segment that can be completed via other paths when that segment fails, quantifying the impact of nodes on network connectivity. By adding these dimensions of metrics, this invention constructs a more comprehensive multi-dimensional indicator system to improve the accuracy of critical segment identification.

[0114] This invention explicitly simulates the impact of track section failures on station routes by using alternative path ratio indicators and interlayer coupled network structures, thereby incorporating the fault propagation effect into the critical section identification decision and making the identification results equally reliable under abnormal operating conditions.

[0115] The processing flow of a method for identifying critical track sections in railway stations based on a multi-layer coupled network provided in this embodiment of the invention is as follows: Figure 1 As shown, the processing includes data acquisition, network modeling, indicator calculation, fusion analysis, and decision-making, and includes the following processing steps:

[0116] Step S10: Data Acquisition: Extract configuration data of track sections, turnouts, signals, and all train routes, extension routes, and shunting routes from the station interlocking chart, and preprocess and verify the acquired data.

[0117] The station interlocking chart is a core technical document of the railway signaling system, used to describe the logical interlocking relationships between signaling equipment within a station. The station interlocking chart includes track section numbers and locations, turnout numbers and turning statuses, signal types and control ranges, and configuration data for all train routes (receiving / departing / passing), extended routes (main route extensions), and shunting routes. It clearly records the interlocking relationships between equipment (such as opposing routes, turnout locking conditions, and unlocking rules), and after preprocessing and verification, provides fundamental support for track static occupancy analysis.

[0118] Step S20: Establish a multi-layer coupled network: Based on the route list in the station interlocking chart, construct multiple sub-networks such as the train route layer network, the continuation route layer network, and the shunting route layer network, and use these multiple sub-networks to construct a multi-layer coupled network.

[0119] A multi-layered coupled network is a collection of all different sub-networks, containing information such as the logical coordinates and names of node track segments. Each sub-network uses multiple different track segments as nodes, and the sequential connections between track segments are represented by edges. Each sub-network contains: track segments (nodes), directed edges between nodes, node logical positions, and names. The nodes, their logical positions, and names are completely identical across all sub-networks, but the directed edges between nodes differ depending on the nature of the task.

[0120] The adjacency relationships of each track segment in different route categories are represented using graph theory nodes and edges. Figure 2 This is a schematic diagram of a multi-layer network structure for a station, provided as an embodiment of the present invention. Figure 2 The network uses three horizontal levels to represent the train route layer (TR layer network), the continuation route layer (ER layer network), and the shunting route layer (SR layer network). Nodes in each layer (circled as T1, T2, T3, etc.) represent track sections of a station. Solid arrows within a layer indicate direct connections between track sections under that layer's route logic (i.e., two track sections appearing sequentially in the corresponding route type). Black dashed lines connect different layers to represent inter-layer coupling; nodes corresponding to the same track section in different layers are considered interconnected. These inter-layer edges unify the node mapping of the three route layers into a single multi-layer coupled network. This multi-layer coupled network can simultaneously represent three logical relationships: train routes, continuation routes, and shunting routes. On one hand, the internal topology of each layer describes the sequential connections of track sections for the corresponding route type; on the other hand, inter-layer coupling edges link physically identical track sections, representing the combined effect of the same track section under different operating modes. The multi-layered coupled network constructed in this way more realistically reflects the interactive effects of station track sections under various operating conditions, laying the foundation for subsequent multi-dimensional index calculation and key section identification.

[0121] Step S30: Indicator Calculation: Calculate multi-dimensional importance indicators for each track segment in the multi-layer coupled network, including topological centrality indicators such as degree, betweenness, and proximity centrality, as well as custom indicators such as topological entropy and alternative path ratio.

[0122] ① The connection frequency (F) of the node track section indicates the number of times the node track section appears in all interlocking charts. This indicator is a benefit-oriented indicator, and the calculation formula is as follows:

[0123]

[0124] in: Z represents the number of times node track segment i appears in the interlocking graph corresponding to the m-th layer network; + Represents a positive integer.

[0125] ② Degree centrality (D) of a node track segment represents the density of connections between neighboring nodes (track segments). This indicator is a benefit-oriented indicator, and its calculation formula is as follows:

[0126] D i =I i +O i

[0127] Where: i represents the node track segment; D i I represents the degree centrality of a node. i Indicates the in-degree centrality of a node, O i This indicates the out-degree centrality of a node.

[0128] ③ The in-degree centrality (I) of a node track segment represents the number of other track segments connected to that node (this track segment). This indicator is a benefit-oriented indicator, and its calculation formula is as follows:

[0129]

[0130] Where: i represents the node track segment; AM (m) Let N represent the adjacency matrix of the m-th layer network; N represent the number of nodes; V represent the set of nodes; and m represent the m-th layer network.

[0131] ④ The out-degree centrality (O) of a node track segment indicates the external connectivity of this node track segment. This indicator is a benefit-oriented indicator, and its calculation formula is as follows:

[0132]

[0133] Where: i represents the node track segment; AM (m) Let N represent the adjacency matrix of the m-th layer network; N represent the number of nodes; V represent the set of nodes; and m represent the m-th layer network.

[0134] ⑤ The proximity centrality (C) of a node's track segment indicates the ease with which the track segment of this node can reach other nodes in the network. This indicator is a benefit-oriented indicator, and its calculation formula is as follows:

[0135]

[0136] Where: n-1 represents the number of nodes reachable from node i in the m-th layer network; Let N represent the shortest path distance between node segment j and node i in the m-th layer network; N represents the number of nodes; V represents the set of nodes; and m represents the m-th layer network.

[0137] ⑥ The betweenness centrality (B) of a node track segment represents the mediating capacity of that node track segment. This indicator is a benefit-type indicator, and its calculation formula is as follows:

[0138]

[0139] in: This represents the number of shortest paths from node track segment s to t in the m-th layer network; V represents the number of shortest paths passing through node i; V represents the set of nodes.

[0140] ⑦ Katz centrality (K) of a node's orbital segment measures the relative influence of network nodes by considering neighboring nodes and all other nodes reached through these neighbors. This is a benefit-based indicator, and its calculation formula is as follows:

[0141]

[0142] Where α represents the attenuation factor, and must satisfy α < 1 / λ MAX ;λ MAX AM (m) Maximum eigenvalue; AM (m) represents the adjacency matrix of the m-th layer network; β represents the adjustment coefficient, which is a constant, usually set to 1; N represents the number of nodes; V represents the node set; m represents the m-th layer network.

[0143] ⑧ Improved PageRank centrality (P) of node track segments quantifies the importance of nodes in the network by iteratively distributing the influence of connectivity and link structure. This indicator is a benefit-based indicator, and its calculation formula is as follows:

[0144]

[0145] Where: μ represents the damping coefficient, used to solve the PageRank spider trap problem, with a value range of (0,1), and is generally set to 0.85; AM (m) Let represent the adjacency matrix of the m-th layer network; e represents an N×N matrix with all elements equal to 1; N represents the number of nodes; V represents the set of nodes; and m represents the m-th layer network.

[0146] Step S40: Fusion Analysis: Apply multi-index decision-making methods (such as TOPSIS) to normalize and weightedly fuse multiple indicators for each track segment, calculate the comprehensive score of each track structure, and evaluate the global importance of each track segment based on the comprehensive score of the track structure.

[0147] Multi-indicator decision-making process:

[0148] Step 1: Input parameters

[0149] ■M: Number of network layers

[0150] ■V: Node set

[0151] ■N: Number of nodes

[0152] ■Q: Number of importance evaluation indicators Step 2: Determine the type of indicator

[0153] 1. Iterate through and sequentially select the q-th decision indicator.

[0154] 2. Determine the type of indicator: For benefit-type indicators, the larger the calculated value, the more important the node. Conversely, for cost-type indicators, the smaller the calculated value, the more important the node.

[0155] 3. Write the judgment result into the indicator type matrix T:

[0156] T=(t q ) 1×Q =[t F t D t I t O t C t B t K t P …]

[0157]

[0158] Step 3: Calculation of comprehensive evaluation indicators for each layer of the network

[0159] 7. Traverse and sequentially select the m-th layer of the network.

[0160] 8. Constructing a multi-attribute initial decision matrix (D (m ))

[0161] a) Initialize D (m) It is an N*Q zero matrix;

[0162] b) Iterate through the q-th evaluation index value of each node i.

[0163] c) Read the type T of the evaluation index q q If T q =1, then Unchanged; if T q =-1, then

[0164] d) Obtain the multi-attribute matrix (D) (m) ),like:

[0165]

[0166] 9. Calculate the weights of each indicator using the entropy weight method.

[0167] a) Traverse and sequentially select the original decision matrix (D) of the m-th layer network.(m) )

[0168] b) Regarding D (m) Standardization is performed to eliminate differences in the dimensions of the indicators.

[0169] To eliminate the influence of dimensions, the decision matrix D is normalized using a proportionality normalization method. (m) Standardization is performed. For the q-th column, the formula for calculating the proportion of each cell is:

[0170]

[0171] In the formula: Represents the multi-attribute initial decision matrix (D) (m) The original data in the i-th row and q-th position of the array. That is, the standardized data, and when At that time, it was agreed

[0172] c) Calculate the entropy value of each indicator.

[0173] For each decision attribute q, its information entropy is calculated based on the standardization results. The calculation formula is as follows:

[0174]

[0175] In the formula: It reflects the dispersion of the q-th index across all decision-making units, where k = 1 / ln(N) is a constant to ensure that the entropy value is normalized to the range [0,1].

[0176] d) Calculate the coefficient of variation for each indicator.

[0177] The higher the information entropy of each indicator, the more dispersed the information of that indicator, and the weaker its discriminative ability. Therefore, the difference coefficient of the indicators is defined as:

[0178]

[0179] In the formula: This reflects the degree of contribution of the q-th indicator to the decision-making process. The larger the value, the stronger the discriminative power of the indicator.

[0180] e) Calculate the weights based on the difference coefficients of each indicator.

[0181] The objective weights of each indicator are determined based on the inverse principle of the coefficient of difference, forming a weight vector. The weights reflect the relative importance of each indicator in the comprehensive evaluation, and their calculation formula is as follows:

[0182]

[0183] In the formula: satisfy Furthermore, each weight value objectively reflects the information differences between the indicators.

[0184] 10. Calculate the comprehensive evaluation value of each node in each layer:

[0185] a) Traverse and sequentially select the m-th layer of the network

[0186] b) Traverse and sequentially select the multi-attribute matrix (D) of the i-th node. (m ))

[0187] c) Calculate the overall network evaluation index for each node. The calculation formula is as follows:

[0188]

[0189] In the formula: Represents a multi-attribute decision matrix (D) (m) The data of the q-th indicator in the i-th node; This indicates the weight corresponding to the indicator q.

[0190] Step 4: Calculation of Comprehensive Evaluation Indicators for Multilayer Networks

[0191] 3. Construct a multi-attribute matrix (A)

[0192] a) Initialize A as an N*M zero matrix;

[0193] b) Traverse the m-th layer importance evaluation index value of each node i

[0194] c) Obtain the multi-attribute matrix (D) N*P ),like:

[0195]

[0196] 6. Calculate the weights of each network layer based on the actual on-site operation, assuming they are...

[0197] ω=[ω1ω2…ω M ]

[0198] 7. The comprehensive evaluation index of the node can be calculated using the following formula:

[0199]

[0200] Where: LEVEL i This represents the final comprehensive evaluation score of node i. The larger the value, the more important the corresponding track segment.

[0201] Step S50: Critical Section Identification and Output: Track sections are sorted according to their comprehensive scores. The track sections with the highest comprehensive scores are identified as critical track sections; for example, the top 10 track sections are identified as critical track sections. The results are then output for operational decision support. The entire process, from data preparation to result output, is executed automatically in sequence, achieving systematic identification of critical track sections at stations.

[0202] In summary, the method of the embodiments of the present invention has the following beneficial effects:

[0203] (1) Enhanced Model Expressiveness: The multi-layer coupled network constructed in this invention adds connections between layers, enabling the model to explicitly represent the mutual influence of track sections under different operating modes. In contrast, existing methods process each layer independently, without considering real-time connections between layers, making it difficult to comprehensively describe the complex relationships in real-world stations.

[0204] (2) More comprehensive indicator system: This invention adds topological entropy and alternative path ratio to the evaluation indicators, which not only consider the global role of nodes in the topology, but also reflect the changes in network connectivity when nodes fail, and evaluate the importance of track segments from multiple perspectives. In contrast, existing schemes mainly rely on traditional centrality indicators, which do not cover dimensions such as the impact of failure, and the indicator dimensions are relatively simple.

[0205] (3) More accurate and robust identification results: By introducing multi-indicator fusion decision-making (such as TOPSIS) and rich indicator information, this invention can effectively reduce the impact of single indicator bias on the results. Compared with the methods in the literature, the key section identification results of this invention maintain higher consistency and effectiveness in different scenarios, can more accurately locate track sections that are crucial to the overall operation, and can still maintain high reliability under simulated fault conditions.

[0206] In summary, this invention is superior to the existing best solutions in terms of network model structure, indicator selection, and decision-making methods, and can provide higher quality key track section identification results, providing strong support for the safe and efficient operation of railway stations.

[0207] Those skilled in the art will understand that the accompanying drawings are merely schematic diagrams of one embodiment, and the modules or processes shown in the drawings are not necessarily essential for implementing the present invention.

[0208] As can be seen from the above description of the embodiments, those skilled in the art can clearly understand that the present invention can be implemented by means of software plus necessary general-purpose hardware platforms. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in various embodiments or some parts of the embodiments of the present invention.

[0209] The various embodiments in this specification are described in a progressive manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, for apparatus or system embodiments, since they are basically similar to method embodiments, the description is relatively simple; relevant parts can be referred to the descriptions in the method embodiments. The apparatus and system embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without creative effort.

[0210] The above description is merely a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for identifying key track sections in railway stations based on multi-layer coupled networks, characterized in that, include: The configuration data of track sections, turnouts, signals, and all train routes, extension routes, and shunting routes are extracted from the station interlocking chart, and the collected data is preprocessed and verified. Based on the route list in the station interlocking chart, construct the train route layer network, the continuation route layer network, and the shunting route layer network sub-networks respectively, and use multiple sub-networks to construct a multi-layer coupled network; For each track segment in the multi-layer coupled network, multi-dimensional index scores are calculated. A multi-index decision method is applied to normalize and weightedly fuse the multiple index scores of each track segment to calculate the comprehensive score of each track structure. The track sections with the highest overall scores are identified as critical track sections.

2. The method according to claim 1, characterized in that, The aforementioned extraction of configuration data for track sections, turnouts, signals, and all train routes, extension routes, and shunting routes from the station interlocking chart, followed by preprocessing and verification of the collected data, includes: Obtain the station interlocking chart, which includes track section number and location, turnout number and turning status, signal type and control range, as well as configuration data for all train routes, continuation routes and shunting routes, and clearly records the interlocking relationships between equipment; The configuration data of track sections, turnouts, signals, and all train routes, extension routes, and shunting routes are extracted from the station interlocking chart, and the collected data is preprocessed and verified.

3. The method according to claim 2, characterized in that, The construction of train route layer network, continuation route layer network, and shunting route layer network sub-networks based on the route list in the station interlocking chart, and the use of multiple sub-networks to construct a multi-layer coupled network, includes: Based on the route list in the station interlocking chart, train route layer sub-network, continuation route layer sub-network, and shunting route layer sub-network are constructed respectively. Using the train route layer network, continuation route layer network, and shunting route layer network, a multi-layer coupling network is constructed. This multi-layer coupling network simultaneously presents three logical relationships: train route, continuation route, and shunting route. Each sub-network in the multi-layer coupled network uses multiple different track segments as nodes and the sequential connection relationship between track segments as edges. Each sub-network contains: track segments, directed edges between nodes, logical positions of nodes, and naming information. Nodes corresponding to the same track segment in different sub-networks are considered to be interconnected.

4. The method according to claim 3, characterized in that, The calculation of multi-dimensional index scores for each orbital segment in the multi-layer coupled network includes: For each track segment in a multi-layer coupled network, a multi-dimensional index score is calculated. The multi-dimensional index includes: connection frequency F, degree centrality D, in-degree centrality I, out-degree centrality O, proximity centrality C, betweenness centrality B, Katz centrality K, and improved PageRank centrality P. The connection frequency F of a node track segment represents the number of times the node track segment appears in all interlocking charts. This indicator is a benefit-oriented indicator, and its calculation formula is as follows: F i =f i (m) ,i∈V,f i ∈Z + Where: f i (m) Z represents the number of times node track segment i appears in the interlocking graph corresponding to the m-th layer network; + V represents a positive integer, V represents the set of nodes, and m represents the m-th layer of the network; The degree centrality D of a node's track segment represents the density of connections between its surrounding neighboring nodes. This indicator is a benefit-oriented indicator, and its calculation formula is as follows: D i =I i +O i Where: i represents the node track segment; D i I represents the degree centrality of a node. i Indicates the in-degree centrality of a node, O i This indicates the out-degree centrality of a node; The in-degree centrality I of a node track segment represents the connection between other track segments and this track segment. This indicator is a benefit-oriented indicator, and its calculation formula is as follows: Where: i represents the node track segment; AM (m) Let N represent the adjacency matrix of the m-th layer network; N represents the number of nodes; and V represents the set of nodes. The out-degree centrality O of a node track segment represents the external connectivity of that node track segment. This indicator is a benefit-oriented indicator, and its calculation formula is as follows: Where: i represents the node track segment; AM( m ) represents the adjacency matrix of the m-th layer network; N represents the number of nodes; V represents the set of nodes; The proximity centrality C of a node's track segment represents the ease with which the track segment of this node can reach other nodes in the network. This indicator is a benefit-oriented indicator, and its calculation formula is as follows: Where: n-1 represents the number of nodes reachable from node i in the m-th layer network; N represents the shortest path distance between node segment j and node i in the m-th layer network; N represents the number of nodes; V represents the set of nodes. The betweenness centrality B of a nodal track segment represents the mediating capacity of that nodal track segment. This indicator is a benefit-oriented indicator, and its calculation formula is as follows: in: This represents the number of shortest paths from node track segment s to t in the m-th layer network; V represents the number of shortest paths passing through node i; V represents the set of nodes. The Katz centrality K of a node's orbital segment measures the relative impact of network nodes by considering neighboring nodes and all other nodes reached through these neighbors. This metric is a benefit-based indicator, and its calculation formula is as follows: Where α represents the attenuation factor, and satisfies α < 1 / λ MAX ;λ MAX AM (m) Maximum eigenvalue; AM (m) Let represent the adjacency matrix of the m-th layer network; β represent the adjustment coefficient; N represent the number of nodes; and V represent the node set. The improved PageRank centrality P of node track segments quantifies the importance of nodes in the network through iterative distribution influenced by connectivity and link structure. This metric is a benefit-based indicator, and its calculation formula is as follows: Where: μ represents the damping coefficient, AM (m) Let represent the adjacency matrix of the m-th layer network; e represents an N×N matrix with all elements equal to 1; N represents the number of nodes.

5. The method according to claim 4, characterized in that, The aforementioned multi-index decision-making method normalizes and weights the scores of multiple indicators for each track segment, calculating a comprehensive score for each track structure, including: Multiple indicators from the following criteria are selected as decision indicators: connection frequency F, degree centrality D, in-degree centrality I, out-degree centrality O, proximity centrality C, betweenness centrality B, Katz centrality K, and improved PageRank centrality P. A multi-indicator decision method is applied to normalize and weightedly fuse multiple indicators for each track segment to calculate the comprehensive score for each track structure. The processing flow of the multi-index decision-making method includes: Obtain the input parameters, including the number of M network layers, the set of V nodes, the number of N nodes, and the number of Q importance evaluation indicators. Iterate through and sequentially select the q-th decision indicator. Determine whether the indicator type is a benefit indicator or a cost indicator, and write the determination result into the indicator type matrix T: T=(t q ) 1×Q =[t F t D t I t O t C t B t K t P …] Calculate the comprehensive evaluation index of each layer of the network:

1. Traverse and sequentially select the m-th layer of the network.

2. Construct the multi-attribute initial decision matrix D (m) a) Initialize D (m) It is an N*Q zero matrix; b) Iterate through the q-th decision index value of each node i. c) Read the type T of decision indicator q q If T q =1, then Unchanged; if T q =-1, but d) Obtain the multi-attribute matrix D (m) ,like:

3. Calculate the weights of each indicator using the entropy weight method. a) Traverse and sequentially select the original decision matrix D of the m-th layer network. (m) b) Regarding D (m) Standardization is performed to eliminate differences in the dimensions of the indicators. The decision matrix D is analyzed using the proportional normalization method. (m) For standardization, the formula for calculating the proportion of each unit for the q-th decision indicator is as follows: In the formula: D represents the initial decision matrix with multiple attributes. (m) The original data in the i-th row and q-th position; That is, the standardized data, and when At that time, it was agreed c) Calculate the entropy value of each indicator. For each decision index q, its information entropy is calculated based on the standardization results, and the calculation formula is as follows: In the formula: This reflects the dispersion of the q-th indicator across all decision-making units. k = 1 / ln(N) is a constant to ensure that the entropy value is normalized to the range of [0,1]. d) Calculate the difference coefficient for each decision indicator. Define the coefficient of variation for decision indicators as: In the formula: This reflects the degree of contribution of the q-th decision indicator to the decision. The larger the value, the stronger the discriminative power of the indicator; e) Calculate the weight of each decision indicator based on the difference coefficient of each decision indicator. The objective weights of each indicator are determined based on the inverse principle of the coefficient of variation, and the calculation formula is as follows: In the formula: Let represent the weight of the q-th decision indicator in the m-th layer network, satisfying ... Calculate the comprehensive evaluation value of each node in each layer: f) Traverse and sequentially select the m-th layer network g) Traverse and sequentially select the multi-attribute matrix D of the i-th node. (m ) h) Calculate the overall network evaluation index for each node. The calculation formula is as follows: In the formula: r represents the comprehensive evaluation value of all nodes in the m-th layer network; i (m) This represents the comprehensive evaluation index of the i-th node in the m-th layer network; Represents the multi-attribute decision matrix D (m) The score data of the i-th node and the q-th indicator; This indicates the weight corresponding to the indicator q; Calculation of comprehensive evaluation index for multi-layer networks 1. Construct a multi-attribute matrix A a) Initialize A as an N*M zero matrix; b) Traverse the m-th layer importance evaluation index value r of each node i i (m) ; c) Obtain the multi-attribute matrix (D) N*P ),like:

2. Determine the weights of each layer of the network, assuming they are... ω=[ω1ω2…ω M ] 3. The comprehensive evaluation index of the node can be calculated using the following formula: Where: LEVEL i This represents the final comprehensive evaluation score of node i.

6. The method according to claim 1, characterized in that, The determination of the track segments with higher overall scores as critical track segments includes: The comprehensive evaluation scores of each track segment are ranked, and a set number of track segments with the highest comprehensive evaluation scores are identified as critical track segments.

Images