Urban rail transit signal system risk point identification method and system

Abstract

The application provides a kind of urban rail transit signal system risk point identification method and system, belongs to urban rail transit operation maintenance technical field, based on the set of urban rail signal system operation safety influence factors and component nodes, constructs the safety characteristic network model of urban rail signal system, obtains the risk characteristic index of urban rail signal system in combination with text data and statistical data;Text sequence is obtained by using operation hidden Markov model to extract text data from the text;Based on the risk characteristic index, in combination with the extracted text sequence, the risk point identification model of urban rail signal system is constructed;Based on the risk point identification model of urban rail signal system, the risk point of urban rail signal system is identified.The application identifies the risk point of urban rail signal system based on the consideration of data characteristics, improves the identification accuracy, provides safe, reliable and stable support for realizing signal system safety risk control, reducing the incidence of urban rail transit accidents and ensuring the safe operation of urban rail transit.

Description

Technical Field

[0001] This invention relates to the field of urban rail transit operation and maintenance technology, specifically to a method and system for identifying risk points in urban rail transit signaling systems. Background Technology

[0002] As the "brain" of the urban rail transit system, the signaling system is a crucial subsystem for ensuring the efficient, safe, and green operation of trains. It is responsible for the information transmission and real-time communication between key components of train operation, while also ensuring the stable operation of train equipment.

[0003] However, the extremely heavy system load has also brought many pain points to the industry. Among them, safety is the prerequisite and foundation for the efficient and stable operation of urban rail transit. Statistical data from surveys show that more than 51% of incidents in the urban rail transit industry, such as people being trapped by platform screen doors, derailments, train stoppages, and train delays of more than five minutes, are caused by signal system failures or signal equipment malfunctions. Therefore, ensuring the safe operation of urban rail transit signal systems has become a top priority for the industry.

[0004] Ensuring the safe and stable operation of the signaling system is a crucial way to reduce the failure and accident rates of urban rail transit systems. As a culmination of advanced technologies, urban rail signaling systems are characterized by structural coupling, unique functions, and a complex operating environment. Different levels of safety risks exist at all levels and in all aspects of the system, presenting critical challenges that urgently need to be addressed, such as difficulties in defining safety risks, identifying key risks, and managing system risks. Summary of the Invention

[0005] The purpose of this invention is to provide a method and system for identifying risk points in urban rail transit signaling systems, so as to solve at least one of the technical problems existing in the background art.

[0006] To achieve the above objectives, the present invention adopts the following technical solution:

[0007] On one hand, the present invention provides a method for identifying risk points in urban rail transit signaling systems, comprising:

[0008] Based on the set of factors affecting the operation safety of urban rail signaling systems and the component nodes, a network model of safety characteristics of urban rail signaling systems is constructed.

[0009] Based on the safety feature network model of urban rail signaling system, combined with text data and statistical data, risk characteristic indicators of urban rail signaling system are obtained;

[0010] The text sequence is obtained by using a hidden Markov model to extract text data.

[0011] Based on risk characteristic indicators and combined with extracted text sequences, a risk point identification model for urban rail signaling systems is constructed.

[0012] Based on the risk point identification model of urban rail signaling system, the risk points of urban rail signaling system are identified.

[0013] Preferably, the risk characteristic indicators are divided into structural indicators and functional indicators. The structural indicators include node importance, which represents the degree value of a node relative to its neighboring nodes. If the current node is connected to other important nodes, then the node is also assigned a high node importance. The functional indicators include impact severity. Considering the data sample characteristics of the urban rail signaling system, the weighted ratio of various indicators such as the central delay duration and the number of trains affecting operational information is calculated to obtain the final impact severity indicator.

[0014] Preferably, a risk point identification model for urban rail signaling systems is constructed, including: using the entropy weight method to calculate the weight of the node importance in the risk characteristic indicators of each LSTM unit, and integrating the node importance and the severity of the impact into each LSTM unit, and calculating the weight parameters.

[0015] Preferably, the calculated risk characteristic index of the weight parameters is integrated into the set of all possible state sequences given by the Hidden Markov Model to form a new vector set, which is then used as the sequence input to the Bi-LSTM network structure. Each bidirectional LSTM unit considers two risk characteristic indices during training and applies them to the calculation process of the next LSTM node unit. By adjusting and optimizing the three gating weight parameters, the probability of the vector to be calculated entering the gating unit is controlled, and finally the hidden layer state at each time step is obtained.

[0016] Preferably, the calculation of weight parameters includes: positively processing the train quantity data affecting operations; standardizing the data to eliminate the influence caused by different ranges of values ​​for different indicators, and constructing a standardization matrix; constructing a probability matrix to calculate the information entropy value of each indicator in the standardization matrix; and obtaining the indicator weight coefficient based on the information entropy value of the indicator.

[0017] Preferably, based on the risk point identification model of the urban rail signaling system, risk points of the urban rail signaling system are identified, including: perceiving the information of fused risk characteristic indicators in the text sequence, the specific formula of which is as follows:

[0018] i t =δ(W i [R t h t-1 ]+m i *R t +n i )

[0019] f t =δ(W f [R t h t-1 ]+m f *R t +n f )

[0020]

[0021]

[0022] o t =δ(W o [R t h t-1 ]+m o *R t +n o )

[0023] h t =o t *tanh(C t )

[0024] Among them, i t f t o t These represent the input gate, forget gate, and output gate in a cell unit, respectively, δ represents the sigmoid function, and R... t h represents the set of input vectors corresponding to time step t. t-1 Indicates the previous hidden state, h t W represents the hidden state at the current time step. i W f W g W o Let n represent the weight matrix. f n g n o m represents the bias of cells in the training data. i m f m o C represents the input gate weight parameter, the forget gate weight parameter, and the output gate weight parameter of the risk feature index, respectively. t This indicates that the LSTM cell storage unit is updated at time t through the input and output gates. This represents the candidate input vector.

[0025] Secondly, the present invention provides a risk point identification system for urban rail transit signaling systems, comprising:

[0026] The first construction module is used to construct a safety feature network model of the urban rail signaling system based on the set of factors affecting the operation safety of the urban rail signaling system and the component nodes.

[0027] The calculation module is used to obtain risk characteristic indicators of urban rail signaling systems based on the safety feature network model of urban rail signaling systems, combined with text data and statistical data.

[0028] The extraction module is used to extract text sequences from text data using a hidden Markov model.

[0029] The second construction module is used to build a risk point identification model for urban rail signaling systems based on risk characteristic indicators and extracted text sequences.

[0030] The identification module is used to identify risk points in the urban rail signaling system based on the risk point identification model.

[0031] Thirdly, the present invention provides a non-transitory computer-readable storage medium for storing computer instructions, which, when executed by a processor, implement the risk point identification method for urban rail transit signaling systems as described above.

[0032] Fourthly, the present invention provides a computer program product, including a computer program that, when run on one or more processors, is used to implement the urban rail transit signaling system risk point identification method as described above.

[0033] Fifthly, the present invention provides an electronic device, comprising: a processor, a memory, and a computer program; wherein the processor is connected to the memory, the computer program is stored in the memory, and when the electronic device is running, the processor executes the computer program stored in the memory to cause the electronic device to execute instructions for implementing the risk point identification method for urban rail transit signaling systems as described above.

[0034] The beneficial effects of this invention are as follows: It elucidates the factors affecting the operational safety of urban rail transit signaling systems, constructs a safety feature network model of urban rail transit signaling systems, and identifies risk points of urban rail transit signaling systems based on data characteristics, thereby improving the accuracy of identification. This provides safe, reliable, and stable support for achieving safety risk control of signaling systems, reducing the accident rate of urban rail transit, and ensuring the safe operation of urban rail transit.

[0035] The advantages of additional aspects of the invention will be set forth more clearly in the following description or will be learned by practice of the invention. Attached Figure Description

[0036] 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.

[0037] Figure 1 This is a flowchart illustrating the risk point identification method for urban rail transit signaling systems according to an embodiment of the present invention.

[0038] Figure 2 This is a schematic diagram of the safety feature network model of the urban rail transit signaling system, which is the method for identifying risk points in the urban rail transit signaling system according to an embodiment of the present invention.

[0039] Figure 3 This is a schematic diagram of the urban rail transit signaling system risk characteristic indicators integrated into the Bi-LSTM network structure in the urban rail transit signaling system risk point identification method described in this embodiment of the invention.

[0040] Figure 4 This is a schematic diagram illustrating the experimental accuracy of the risk point identification model for urban rail transit signaling systems according to the embodiment of the present invention. Detailed Implementation

[0041] 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.

[0042] 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.

[0043] It should also be understood that terms such as those defined in general dictionaries should be understood to have meanings consistent with their meanings in the context of the prior art, and should not be interpreted in an idealized or overly formal sense unless defined as described here.

[0044] 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, and / or groups thereof.

[0045] In the description of this specification, references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the present invention. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of those different embodiments or examples.

[0046] To facilitate understanding of the present invention, the present invention will be further explained and described below with reference to the accompanying drawings and specific embodiments. However, the specific embodiments do not constitute a limitation on the embodiments of the present invention.

[0047] Those skilled in the art should understand that the accompanying drawings are merely schematic diagrams of embodiments, and the components in the drawings are not necessarily essential for implementing the present invention.

[0048] Example 1

[0049] This embodiment 1 provides a risk point identification system for urban rail transit signaling systems, including:

[0050] The first construction module is used to construct a safety feature network model of the urban rail signaling system based on the set of factors affecting the operation safety of the urban rail signaling system and the component nodes.

[0051] The calculation module is used to obtain risk characteristic indicators of urban rail signaling systems based on the safety feature network model of urban rail signaling systems, combined with text data and statistical data.

[0052] The extraction module is used to extract text sequences from text data using a hidden Markov model.

[0053] The second construction module is used to build a risk point identification model for urban rail signaling systems based on risk characteristic indicators and extracted text sequences.

[0054] The identification module is used to identify risk points in the urban rail signaling system based on the risk point identification model.

[0055] In this embodiment 1, the above-described system is used to implement a risk point identification method for urban rail transit signaling systems, including:

[0056] Based on the set of factors affecting the operation safety of urban rail signaling systems and the component nodes, a network model of safety characteristics of urban rail signaling systems is constructed.

[0057] Based on the safety feature network model of urban rail signaling system, combined with text data and statistical data, risk characteristic indicators of urban rail signaling system are obtained;

[0058] The text sequence is obtained by using a hidden Markov model to extract text data.

[0059] Based on risk characteristic indicators and combined with extracted text sequences, a risk point identification model for urban rail signaling systems is constructed.

[0060] Based on the risk point identification model of urban rail signaling system, the risk points of urban rail signaling system are identified.

[0061] The risk characteristic indicators are divided into structural indicators and functional indicators. The structural indicators include node importance, which represents the degree value of a node relative to its neighboring nodes. If the current node is connected to other important nodes, then the node is also assigned a high node importance. The functional indicators include impact severity. Considering the data sample characteristics of the urban rail signaling system, the weighted ratio of various indicators such as the central delay duration and the number of trains affecting operational information is calculated to obtain the final impact severity indicator.

[0062] Construct a risk point identification model for urban rail signaling systems, including: using the entropy weight method to calculate the weight of node importance in the risk characteristic indicators of each LSTM unit, and integrating node importance and impact severity into each LSTM unit, and calculating weight parameters.

[0063] The calculated risk characteristic index of the weight parameters is integrated into the set of all possible state sequences given by the Hidden Markov Model to form a new vector set, which is then used as the sequence input to the Bi-LSTM network structure. Each bidirectional LSTM unit considers two risk characteristic indices during training and applies them to the calculation process of the next LSTM node unit. By adjusting and optimizing the three gating weight parameters, the probability of the vector to be calculated entering the gating unit is controlled, and finally the hidden layer state at each time step is obtained.

[0064] The calculation of weight parameters includes: positively processing the train quantity data affecting operations; standardizing the data to eliminate the influence caused by different ranges of values ​​for different indicators, and constructing a standardization matrix; constructing a probability matrix to calculate the information entropy value of each indicator in the standardization matrix; and obtaining the indicator weight coefficient based on the information entropy value of the indicator.

[0065] Based on the risk point identification model of urban rail transit signaling system, risk points of urban rail transit signaling system are identified, including: perceiving the information of fused risk characteristic indicators in text sequences, the specific formula is expressed as follows:

[0066] it =δ(W i [R t h t-1 ]+m i *R t +n i )

[0067] f t =δ(W f [R t h t-1 ]+m f *R t +n f )

[0068]

[0069]

[0070] o t =δ(W o [R t h t-1 ]+m o *R t +n o )

[0071] h t =o t *tanh(C t )

[0072] Among them, i t f t o t These represent the input gate, forget gate, and output gate in a cell unit, respectively, δ represents the sigmoid function, and R... t h represents the set of input vectors corresponding to time step t. t-1 Indicates the previous hidden state, h t W represents the hidden state at the current time step. i W f W g W o Let n represent the weight matrix. f n g n o m represents the bias of cells in the training data. i m f m o C represents the input gate weight parameter, the forget gate weight parameter, and the output gate weight parameter of the risk feature index, respectively. t This indicates that the LSTM cell storage unit is updated at time t through the input and output gates. This represents the candidate input vector.

[0073] Example 2

[0074] This embodiment 2 provides a method for identifying risk points in an urban rail transit signaling system, including the following steps:

[0075] Step 1: Based on the set of factors affecting the operation safety of the urban rail signaling system and the component nodes, construct a network model of the safety characteristics of the urban rail signaling system.

[0076] Step 2: Based on the network model, combined with text data and statistical data, construct risk characteristic indicators for urban rail signaling systems, and extract text data using a hidden Markov model.

[0077] Step 3: Based on risk characteristic indicators and combined with the text sequence output by the Hidden Markov Model, construct a risk point identification model for the urban rail signaling system;

[0078] Step 4: Based on the risk point identification model of the urban rail signaling system, the risk points of the urban rail signaling system are identified.

[0079] In this embodiment 2, the risk characteristic indicators of the urban rail signaling system in step two include:

[0080] Node Importance: This is a structural indicator. If each point in the risk point topology network model of a signal system is considered as a feature vector, then node importance refers to the centrality of this feature vector. This importance indicator represents the degree value of a node relative to its neighboring nodes. If the current node is connected to other important nodes, then that node is also assigned a higher node importance.

[0081] For a given ordinary node, the degree is the number of nodes connected to it in the topology. The formula for calculating the degree of the i-th node is:

[0082] d i =∑ j a ij d j (j=,1,2,...,n) (1)

[0083] Among them, a ij This indicates the connection state between node i and node j; n represents the number of nodes connected to node i. When there is a connection between two nodes, a... ij The value of a is 1; when there is no connection between two nodes, a ij The value is 0, and this calculation method is applicable to binary networks.

[0084] In this embodiment 2, the importance of the i-th node in the signal system topology network model can be calculated using the adjacency matrix, and the final importance calculation result of each node is obtained. The specific calculation formula is as follows:

[0085]

[0086] This formula means that the eigenvector centrality of the i-th node is equal to the sum of the eigenvector centralities of all its neighbors, where To determine the centrality of nodes in the topological network, this value must be a positive number. The specific calculation formula is as follows:

[0087]

[0088] Among them, E i (x,y) represents the number of shortest paths between nodes x and y via node i; E(x,y) represents the number of shortest paths between nodes x and y. This parameter indicates that if a node has high centrality, it means that the shortest path between other points must pass through that node, and the higher the importance of that node.

[0089] In summary, the formula for node importance is:

[0090]

[0091] Impact Severity: This is a functional indicator, based on the data characteristics of the original data and incorporating an entropy weighting algorithm to analyze and define the weights of impact factors. The impact severity fully considers the data sample characteristics of the urban rail signaling system, extracting various indicators from the dataset, including delay duration and the number of trains affecting operational information (mainly including the number of trains delayed by less than 5 minutes α, the number of trains delayed by more than 5 minutes β, the number of through trains γ, the number of cancelled trains δ, and the number of trains offline ε), and calculating their weighted ratios to obtain the final impact severity index. The formula is as follows:

[0092]

[0093] Where n is the risk factor S involved i Total number of data entries Delay duration for each data failure, Indicates risk factor S i Similarly, the number of trains delayed by less than 5 minutes in the k-th data point... Indeed. a , λ b , λ c , λ d , λ eThe weighted parameters ε represent the number of trains delayed by less than 5 minutes, the number of trains delayed by more than 5 minutes, the number of through trains, the number of cancelled trains, and the number of disconnected trains, respectively, and are calculated by the entropy weight algorithm.

[0094] In this embodiment 2, step three includes:

[0095] The entropy weight method is used to calculate the weight of structural importance in the risk characteristic indicators of each LSTM unit, and the two indicators are integrated into each LSTM unit. The steps for calculating the weight parameters are as follows:

[0096] (1) Data forward processing

[0097] The interval-type, extremely small, and intermediate-type indicators will be processed to convert them into extremely large-type indicator data. Since the indicator values ​​in this paper represent the number of trains affecting operations, we hope that all indicators are as close to zero as possible. Therefore, this paper uses the intermediate-type indicator formula as follows:

[0098]

[0099] Where, x i This represents the value of the i-th row in the original data index; x b This represents the optimal data value for the current indicator column. M is the sum of the values ​​in the current data sample column and x. b The result is obtained by taking the absolute value of the difference, then taking the maximum value from the set. The specific formula is as follows:

[0100] M = ∑max(|x i -x b |) (7)

[0101] (2) Data standardization processing

[0102] Data standardization is primarily aimed at eliminating the impact of different ranges of values ​​for different indicators. Therefore, a standardization matrix Z is constructed, and its calculation method is shown below:

[0103]

[0104] Among them, Z ij x represents each element in the normalized matrix; ij This represents each element in the matrix after the forwarding process.

[0105] (3) Calculate the probability matrix

[0106] After standardizing the data in the normalization matrix, a probability matrix is ​​needed to calculate the information entropy value of each indicator. The specific formula is as follows:

[0107]

[0108] Among them, Z ij This represents each element in the normalized matrix.

[0109] (4) Calculate the information entropy value of the indicator to obtain the indicator weight coefficient.

[0110] For each indicator, the information entropy value is understood as a quantification of the information value of each indicator. The specific formula is as follows, where n is the number of indicators in the matrix:

[0111]

[0112] The weight coefficient S of each indicator in the current data can be calculated based on the information entropy value of the indicator. j The larger the calculated weight coefficient, the greater the information content and the greater the impact of the indicator.

[0113]

[0114] The calculated risk characteristic indicators of the weight parameters are integrated into the set of all possible state sequences given by the Hidden Markov Model, forming a new vector set R. This R is then input into the Bi-LSTM network structure. Each bidirectional LSTM unit considers two risk characteristic indicators during training and applies them to the computation process of the next LSTM node. By adjusting and optimizing the three gating weight parameters, the probability of each vector entering the gating unit is controlled, ultimately obtaining the hidden layer state at each time step.

[0115] This method can effectively perceive information about fusion risk characteristic indicators in text sequences. The specific formula is as follows:

[0116] i t =δ(W i [R t h t-1 ]+m i *R t +n i (12)

[0117] f t =δ(W f [R t h t-1 ]+m f *R t +n f (13)

[0118]

[0119]

[0120] o t=δ(W o [R t h t-1 ]+m o *R t +n o (16)

[0121] h t =o t *tanh(C t (17)

[0122] Among them, i t f t o t These represent the input gate, forget gate, and output gate in a cell unit, respectively, δ represents the sigmoid function, and R... t h represents the set of input vectors corresponding to time step t. t-1 Indicates the previous hidden state, h t W represents the hidden state at the current time step. i W f W g W o Let n represent the weight matrix. f n g n o m represents the bias of cells in the training data. i m f m o C represents the input gate weight parameter, the forget gate weight parameter, and the output gate weight parameter of the risk feature index, respectively. t This indicates that the LSTM cell storage unit is updated at time t through the input and output gates. This represents the candidate input vector. In a traditional LSTM, i... t It can only control whether a new sequence can be input, while the RB-LSTM network controls the parameter m i Optimization: Input i according to the optimal ratio t This makes the results more targeted.

[0123] Example 3

[0124] like Figure 1 As shown in the figure, this embodiment 3 provides a method for identifying risk points in an urban rail transit signaling system, including the following steps:

[0125] Step 1: Based on the set of factors affecting the operation safety of the urban rail signaling system and the component nodes, construct a network model of the safety characteristics of the urban rail signaling system.

[0126] Step 2: Based on the network model, combined with text data and statistical data, construct risk characteristic indicators for urban rail signaling systems, and extract text data using a hidden Markov model.

[0127] Step 3: Based on risk characteristic indicators and combined with the text sequence output by the Hidden Markov Model, construct a risk point judgment model for the urban rail signaling system;

[0128] Step 4: Based on the risk point judgment model of the urban rail signaling system, the risk points of the urban rail signaling system are identified.

[0129] In step one, unsafe human behavior, unsafe equipment conditions, and environmental maladaptation can all cause urban rail transit signaling systems to malfunction and fail to operate safely. Therefore, this invention considers the impact of personnel and the environment on its physical components and categorizes the human-related, environmental, and physical component nodes of urban rail transit signaling systems (as shown in Tables 1, 2, and 3).

[0130] Table 1 Physical Component Nodes

[0131]

[0132]

[0133]

[0134]

[0135] Table 2 Personnel Category Nodes

[0136]

[0137] Table 3 Environmental Component Nodes

[0138]

[0139] Simultaneously, based on the set of operational safety impact factors and expert experience, a system was built using Gephi, such as... Figure 2 The safety feature network model of the urban rail signaling system is shown.

[0140] Furthermore, based on the safety feature network model diagram of the urban rail transit signaling system, and combined with textual and statistical data, a risk point identification model for urban rail transit was constructed. The model construction process is as follows:

[0141] 1. The construction of risk characteristic indicators for urban rail signaling systems includes:

[0142] The aforementioned risk characteristic indicators are divided into structural indicators and functional indicators.

[0143] Node Importance: This is a structural indicator. If each point in the risk point topology network model of a signal system is considered as a feature vector, then node importance refers to the centrality of this feature vector. This importance indicator represents the degree value of a node relative to its neighboring nodes. If the current node is connected to other important nodes, then that node is also assigned a higher node importance.

[0144] For a given ordinary node, the degree is the number of nodes connected to it in the topology. The formula for calculating the degree of the i-th node is:

[0145] d i =∑ j a ij d j (j=,1,2,...,n) (1)

[0146] Among them, a ij This indicates the connection state between node i and node j; n represents the number of nodes connected to node i. When there is a connection between two nodes, a... ij The value of a is 1; when there is no connection between two nodes, a ij The value is 0, and this calculation method is applicable to binary networks.

[0147] The importance of the i-th node in the signal system topology network model can be calculated using the adjacency matrix, ultimately yielding the importance calculation result for each node. The specific calculation formula is as follows:

[0148]

[0149] This formula means that the eigenvector centrality of the i-th node is equal to the sum of the eigenvector centralities of all its neighbors, where To determine the centrality of nodes in the topological network, this value must be a positive number. The specific calculation formula is as follows:

[0150]

[0151] Among them, E i (x,y) represents the number of shortest paths between nodes x and y via node i; E(x,y) represents the number of shortest paths between nodes x and y. This parameter indicates that if a node has high centrality, it means that the shortest path between other points must pass through that node, and the higher the importance of that node.

[0152] In summary, the formula for node importance is:

[0153]

[0154] Impact Severity: This is a functional indicator, based on the data characteristics of the original data and incorporating an entropy weighting algorithm to analyze and define the weights of impact factors. The impact severity fully considers the data sample characteristics of the urban rail signaling system, extracting various indicators from the dataset, including delay duration and the number of trains affecting operational information (mainly including the number of trains delayed by less than 5 minutes α, the number of trains delayed by more than 5 minutes β, the number of through trains γ, the number of cancelled trains δ, and the number of trains offline ε), and calculating their weighted ratios to obtain the final impact severity index. The formula is as follows:

[0155]

[0156] Where n is the risk factor S involved i Total number of data entries Delay duration for each data failure, Indicates risk factor S i Similarly, the number of trains delayed by less than 5 minutes in the k-th data point... Indeed. a , λ b , λ c , λ d , λ e The weighted parameters ε represent the number of trains delayed by less than 5 minutes, the number of trains delayed by more than 5 minutes, the number of through trains, the number of cancelled trains, and the number of disconnected trains, respectively, and are calculated by the entropy weight algorithm.

[0157] 2. The construction of the risk point identification model for urban rail signaling systems specifically includes:

[0158] The entropy weight method is used to calculate the weight of structural importance in the risk characteristic indicators of each LSTM unit, and the two indicators are integrated into each LSTM unit. The steps for calculating the weight parameters are as follows:

[0159] (1) Data forward processing

[0160] The interval-type, extremely small, and intermediate-type indicators will be processed to convert them into extremely large-type indicator data. Since the indicator values ​​in this paper represent the number of trains affecting operations, we hope that all indicators are as close to zero as possible. Therefore, this paper uses the intermediate-type indicator formula as follows:

[0161]

[0162] Where, x i This represents the value of the i-th row in the original data index; x b This represents the optimal data value for the current indicator column. M is the result of taking the maximum value from the set of absolute differences between each value in the current data sample column and xb, as shown in the following formula:

[0163] M = ∑max(|x i -x b |) (7)

[0164] (2) Data standardization processing

[0165] Data standardization is primarily aimed at eliminating the impact of different ranges of values ​​for different indicators. Therefore, a standardization matrix Z is constructed, and its calculation method is shown below:

[0166]

[0167] Among them, Z ij x represents each element in the normalized matrix; ij This represents each element in the matrix after the forwarding process.

[0168] (3) Calculate the probability matrix

[0169] After standardizing the data in the normalization matrix, a probability matrix is ​​needed to calculate the information entropy value of each indicator. The specific formula is as follows:

[0170]

[0171] Among them, Z ij This represents each element in the normalized matrix.

[0172] (4) Calculate the information entropy value of the indicator to obtain the indicator weight coefficient.

[0173] For each indicator, the information entropy value is understood as a quantification of the information value of each indicator. The specific formula is as follows, where n is the number of indicators in the matrix:

[0174]

[0175] The weight coefficient S of each indicator in the current data can be calculated based on the information entropy value of the indicator. j The larger the calculated weight coefficient, the greater the information content and the greater the impact of the indicator.

[0176]

[0177] The calculated risk characteristic indicators of the weight parameters are integrated into the set of all possible state sequences given by the Hidden Markov Model, forming a new vector set R. This R is then input into the Bi-LSTM network structure. Each bidirectional LSTM unit considers two risk characteristic indicators during training and applies them to the computation process of the next LSTM node. By adjusting and optimizing the three gating weight parameters, the probability of each vector entering the gating unit is controlled, ultimately obtaining the hidden layer state at each time step.

[0178] This method can effectively perceive information about fusion risk characteristic indicators in text sequences. The specific formula is as follows:

[0179] i t =δ(W i [R t h t-1 ]+m i *R t +n i (12)

[0180] f t =δ(W f [R t h t-1 ]+m f *R t +n f (13)

[0181]

[0182]

[0183] o t =δ(W o [R t h t-1 ]+m o *R t +n o (16)

[0184] h t =o t *tanh(C t (17)

[0185] Among them, i t f t o t These represent the input gate, forget gate, and output gate in a cell unit, respectively, δ represents the sigmoid function, and R... t h represents the set of input vectors corresponding to time step t. t-1 Indicates the previous hidden state, h t W represents the hidden state at the current time step. i W f W g W o Let n represent the weight matrix. f n g n o m represents the bias of cells in the training data. i m f m oC represents the input gate weight parameter, the forget gate weight parameter, and the output gate weight parameter of the risk feature index, respectively. t This indicates that the LSTM cell storage unit is updated at time t through the input and output gates. This represents the candidate input vector. In a traditional LSTM, i... t It can only control whether a new sequence can be input, while the RB-LSTM network controls the parameter m i Optimization: Input i according to the optimal ratio t This makes the results more targeted.

[0186] 3. The constructed urban rail signal risk point identification model was experimentally verified.

[0187] The experiment first required determining the optimal weight parameters for the RB-LSTM network model in different gating units. During the experiment, the weight parameters of the three gating units were represented by i, f, and o, respectively. The model was trained using 2000 original dataset entries and validated using 500 dataset entries. A control experiment was first set up, with the weight parameters of all three gates set to 0, meaning no control was made over the input parameter ratio. The results showed that the distribution of the network model's loss rate, accuracy, and validation set accuracy was very uneven. For example, only one instance had a loss rate below 30%, far from meeting the experimental requirements. Subsequently, parameter tuning was performed. After multiple experiments, it was found that when i = 0.5, f = 0.5, and o = 0.5, the three metrics performed best, with the loss rate concentrated between 15% and 20%, and the accuracy stabilizing above 95% after multiple training rounds, meeting the expected experimental results. After 20 rounds of validation on the validation set data, as shown... Figure 4 As shown, the RB-LSTM network model proposed in this invention has a loss rate of 1.51% (line ①), an accuracy of 98.72% (line ②), and a validation set accuracy of 82.1% (line ③). Experiments were conducted using the constructed RB-LSTM model on text data related to faults and accidents in urban rail transit signaling systems, resulting in 107 risk point keywords as follows:

[0188] Extreme cold / vehicle-mounted personnel / workstation display / automatic gate opener / signal tower operator / ATS extension / bottom shell / motor / accelerometer sensor;

[0189] Signal / Axle counter / Rainfall / Filament conversion relay / Train driver / Tuning unit (TU) / Server / Train departure timer / BTM antenna / Depot control operator / Grounding wire;

[0190] High temperature / Wireless switch / Switch machine / Vehicle switch / Industrial control computer / Axle counter sensor / Maintenance personnel / Network communication equipment / ATO brake relay / Free wave antenna;

[0191] Junction box / Large screen host / Transponder / Receiver / Moving contact / JTC power supply / Gateway computer / Shift contactor / CPU board / Amplifier board / Transformer;

[0192] Gearbox / Power Supply Panel / Flange / Analog Board / Amplifier Filter Board / Inspection Column / Lightning Protection Equipment / Snowfall / Spindle Counting Head / Energy Saver / Cable / Cover / Waveguide;

[0193] Block information board / dust / relay / power board / speed sensor / track circuit / indicator pole / platform screen door / AP motherboard / optical modem / TRE box / train dispatcher;

[0194] Passengers / Vehicle Controllers / Vehicle Equipment Human-Machine Interfaces / Electromagnetic Interference / BTM Power Amplifier Boards / ATO Vehicle Control Boards / Dispatch Room Personnel / Bolts / VOBC Equipment;

[0195] Console / Run Diagram / ZC / MMI / MDT Board / EB Board / AP Board / Output Board / Planet Antenna / Flood / Beacon / Loop Detection Board / Battery / I / O Board;

[0196] Bracket / Point Rail / Base / Static Contact / Coil / XB Box / Central Screen / Zeroing Board / Receiving Antenna / ATP Main Control Board / SA Board / Repeater / Turnout;

[0197] Adjustment plate / external locking frame / light bulb filament / cable / BTM transmitter board / construction personnel / CPAM module / axle counter / frost / railside AP.

[0198] Example 4

[0199] Embodiment 4 of the present invention provides a non-transitory computer-readable storage medium for storing computer instructions. When executed by a processor, the computer instructions implement a method for identifying risk points in an urban rail transit signaling system. The method includes:

[0200] Based on the set of factors affecting the operation safety of urban rail signaling systems and the component nodes, a network model of safety characteristics of urban rail signaling systems is constructed.

[0201] Based on the safety feature network model of urban rail signaling system, combined with text data and statistical data, risk characteristic indicators of urban rail signaling system are obtained;

[0202] The text sequence is obtained by using a hidden Markov model to extract text data.

[0203] Based on risk characteristic indicators and combined with extracted text sequences, a risk point identification model for urban rail signaling systems is constructed.

[0204] Based on the risk point identification model of urban rail signaling system, the risk points of urban rail signaling system are identified.

[0205] Example 5

[0206] Embodiment 5 of the present invention provides a computer program (product), including a computer program that, when run on one or more processors, is used to implement a risk point identification method for urban rail transit signaling systems. The method includes:

[0207] Based on the set of factors affecting the operation safety of urban rail signaling systems and the component nodes, a network model of safety characteristics of urban rail signaling systems is constructed.

[0208] Based on the safety feature network model of urban rail signaling system, combined with text data and statistical data, risk characteristic indicators of urban rail signaling system are obtained;

[0209] The text sequence is obtained by using a hidden Markov model to extract text data.

[0210] Based on risk characteristic indicators and combined with extracted text sequences, a risk point identification model for urban rail signaling systems is constructed.

[0211] Based on the risk point identification model of urban rail signaling system, the risk points of urban rail signaling system are identified.

[0212] Example 6

[0213] Embodiment 6 of the present invention provides an electronic device, including: a processor, a memory, and a computer program; wherein, the processor is connected to the memory, and the computer program is stored in the memory. When the electronic device is running, the processor executes the computer program stored in the memory to cause the electronic device to execute instructions for implementing a risk point identification method for urban rail transit signaling systems, the method including:

[0214] Based on the set of factors affecting the operation safety of urban rail signaling systems and the component nodes, a network model of safety characteristics of urban rail signaling systems is constructed.

[0215] Based on the safety feature network model of urban rail signaling system, combined with text data and statistical data, risk characteristic indicators of urban rail signaling system are obtained;

[0216] The text sequence is obtained by using a hidden Markov model to extract text data.

[0217] Based on risk characteristic indicators and combined with extracted text sequences, a risk point identification model for urban rail signaling systems is constructed.

[0218] Based on the risk point identification model of urban rail signaling system, the risk points of urban rail signaling system are identified.

[0219] In summary, the risk point identification method for urban rail transit signaling systems provided in this embodiment of the present invention mainly includes the following steps: constructing a safety feature network model for urban rail transit signaling systems based on the set of operational safety impact factors and component nodes; constructing risk characteristic indicators for urban rail transit signaling systems based on this network model, combined with text data and statistical data, while simultaneously extracting text data using a Hidden Markov Model; constructing a risk point judgment model for urban rail transit signaling systems based on the risk characteristic indicators and the text sequence output by the Hidden Markov Model; and identifying risk points in the urban rail transit signaling systems based on this risk point judgment model. The method provided by this invention has the following advantages: it fully considers the situation where different urban rail transit signaling systems have different models of equipment when constructing the safety feature network model, and fully considers the data characteristics of urban rail transit signaling system faults and accidents. It emphasizes the construction of a risk point identification model for urban rail transit signaling systems based on the integration of risk characteristic indicators, thereby enhancing the accuracy of risk point identification and providing methodological support for achieving safety risk management of signaling systems, reducing the incidence of urban rail transit accidents, and ensuring the safe operation of urban rail transit. A dynamic calculation model for the full life cycle cost of urban rail transit was constructed, which can more effectively calculate the full life cycle cost of urban rail transit and provide an important research foundation for optimizing the decision-making of the full life cycle cost of urban rail transit.

[0220] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0221] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0222] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0223] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment, whereby a series of operational steps are performed to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0224] While the specific embodiments of the present invention have been described above in conjunction with the accompanying drawings, this is not intended to limit the scope of protection of the present invention. Those skilled in the art should understand that, based on the technical solutions disclosed in the present invention, various modifications or variations that can be made by those skilled in the art without creative effort should be included within the scope of protection of the present invention.

Claims

1. A method for identifying risk points in an urban rail transit signaling system, characterized in that, include: Based on the set of factors affecting the operation safety of urban rail signaling systems and the component nodes, a network model of safety characteristics of urban rail signaling systems is constructed. Based on the safety feature network model of urban rail transit signaling systems, and combining textual and statistical data, risk characteristic indicators of urban rail transit signaling systems are obtained. These risk characteristic indicators are divided into structural indicators and functional indicators. The structural indicators include node importance, which represents the degree value of a node relative to its neighboring nodes. If a node is connected to other important nodes, it is also assigned a higher node importance. The functional indicators include impact severity. Considering the data sample characteristics of urban rail transit signaling systems, various indicators such as the central delay duration and the number of trains affecting operational information are extracted from the data, and their weighted ratios are calculated to obtain the final impact severity indicator. The text sequence is obtained by using a hidden Markov model to extract text data. Based on risk characteristic indicators and combined with extracted text sequences, a risk point identification model for urban rail signaling systems is constructed. Based on the risk point identification model of urban rail transit signaling system, risk points of urban rail transit signaling system are identified. The construction of the risk point identification model includes: calculating the weights of node importance in the risk characteristic indicators of each LSTM unit using the entropy weight method, and integrating node importance and impact severity into each LSTM unit to calculate weight parameters; integrating the calculated weighted risk characteristic indicators into the set of all possible state sequences given by the Hidden Markov Model to form a new vector set, which is then used as the sequence input to the Bi-LSTM network structure. Each bidirectional LSTM unit considers two risk characteristic indicators during training and applies them to the calculation process of the next LSTM node unit; controlling the probability of the vector to be calculated entering the gated unit by adjusting and optimizing three gating weight parameters, ultimately obtaining the hidden layer state at each time step; calculating weight parameters includes: positively processing the train quantity data affecting operation; standardizing the data to eliminate the influence caused by different value ranges of different indicator data, constructing a standardization matrix; constructing a probability matrix to calculate the information entropy value of each indicator in the standardization matrix; and obtaining the indicator weight coefficient based on the information entropy value of the indicator.

2. The method for identifying risk points in urban rail transit signaling systems according to claim 1, characterized in that, Based on the risk point identification model of urban rail transit signaling system, risk points of urban rail transit signaling system are identified, including: perceiving the information of fused risk characteristic indicators in text sequences, the specific formula is expressed as follows: ； ； ； ； ； ； in, , , These represent the input gate, forget gate, and output gate in a cell unit, respectively. This represents the sigmoid function. This represents the set of input vectors corresponding to time step t. This indicates the previous hidden state. This represents the hidden state at the current time step. , , , Represents the weight matrix. , , This indicates the bias of cells in the training data. , , These represent the input gate weight parameters, the forget gate weight parameters, and the output gate weight parameters of the risk characteristic indicators, respectively. This indicates that the LSTM cell storage unit is updated at time t through the input and output gates. This represents the candidate input vector.

3. A risk point identification system for urban rail transit signaling systems based on the method described in claim 1 or 2, characterized in that, include: The first construction module is used to construct a safety feature network model of the urban rail signaling system based on the set of factors affecting the operation safety of the urban rail signaling system and the component nodes. The calculation module is used to obtain risk characteristic indicators of urban rail signaling systems based on the safety feature network model of urban rail signaling systems, combined with text data and statistical data. The extraction module is used to extract text sequences from text data using a hidden Markov model. The second construction module is used to build a risk point identification model for urban rail signaling systems based on risk characteristic indicators and extracted text sequences. The identification module is used to identify risk points in the urban rail signaling system based on the risk point identification model.

4. A non-transitory computer-readable storage medium, characterized in that, The non-transitory computer-readable storage medium is used to store computer instructions, which, when executed by a processor, implement the risk point identification method for urban rail transit signaling systems as described in claim 1 or 2.

5. A computer program product, characterized in that, Includes a computer program, which, when run on one or more processors, is used to implement the risk point identification method for urban rail transit signaling systems as described in claim 1 or 2.

6. An electronic device, characterized in that, include: The device includes a processor, a memory, and a computer program; wherein the processor is connected to the memory, and the computer program is stored in the memory. When the electronic device is running, the processor executes the computer program stored in the memory to cause the electronic device to execute instructions for implementing the risk point identification method for urban rail transit signaling systems as described in claim 1 or 2.

Images