Safety treatment methods, devices, equipment and media for rail transit system lines and networks
By reducing and clustering the initial data of the rail transit system network and determining the target cluster, the problem of difficulty in effectively evaluating and managing the safety of the rail transit system network in the existing technology is solved, and the accuracy and efficiency of safety processing are improved , ensuring operational safety.
Patent Information
- Application Number
- CN202411230496.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-04
- Publication Date
- 2024-12-03
- Estimated Expiration
- 2044-09-04
AI Technical Summary
It is difficult for the existing technology to effectively evaluate and manage the safety of rail transit system networks, especially in a networked environment with multi-standard integration, and the lack of scientific and accurate safety risk assessment methods, resulting in increased operational management difficulties.
By obtaining the initial data group and scene data of the rail transit system network based on the initial safety state dimension, performing dimensionality reduction processing, determining the state similarity between the objects to be processed, clustering to obtain the reference cluster and cluster center, and iteratively update it Obtain the target cluster and then perform safe processing.
It improves the accuracy and efficiency of safety processing of rail transit system networks, ensures that safety risks are in a controllable state, avoids the impact of redundant data on subsequent target cluster determination, and enhances managers' operational safety grasp capabilities.
Smart Images

Figure CN118735280B_ABST
Abstract
Description
Technical Field
[0001] Embodiments of the present invention relate to the field of rail transit technology, and in particular to a method, device, equipment and medium for safely handling a line network of a rail transit system. Background Art
[0002] The safety evaluation of rail transit system networks has always been one of the important contents of rail transit safety management. How to use scientific and effective methods to evaluate the safety of rail transit system networks is of great significance. With the regional development of urban agglomerations, the travel distance of passengers continues to extend, and the scale of the rail transit network with regional rail transit as the core continues to expand. Ensuring the safe operation of the network with multi-standard integration brings greater challenges to managers. The most fundamental purpose of establishing a rail transit operation safety risk evaluation method is to assist rail transit operation managers to control the status of the network operation process and safety risks in real time, comprehensively and accurately, realize early warning and pre-control, so that safety risks are controllable, prevent problems before they happen, and ensure operational safety. Therefore, it is crucial to improve the accuracy of safety processing of rail transit system networks. Summary of the invention
[0003] The present invention provides a method, device, equipment and medium for safely processing a line network of a rail transit system, so as to improve the accuracy of safely processing the line network of the rail transit system.
[0004] According to one aspect of the present invention, a method for safely processing a rail transit system network is provided, comprising:
[0005] Based on the initial safety state dimension, an initial data group of each object to be processed in the rail transit system network and scene data of the object to be processed are obtained; wherein the initial data group includes state attribute data corresponding to the initial safety state dimension;
[0006] Performing dimensionality reduction processing on the initial data group to obtain a target data group for each object to be processed;
[0007] According to each of the target data groups, determining the state similarity between the corresponding objects to be processed, and clustering the objects to be processed according to the state similarity to obtain a reference cluster and a cluster center of the reference cluster;
[0008] Determine the center distance between each of the objects to be processed and the center of the cluster;
[0009] According to the center distance, the reference cluster is iteratively updated to obtain a target cluster, and according to the target cluster and the scene data, the rail transit system network is safely processed.
[0010] According to another aspect of the present invention, a safety processing device for a rail transit system network is provided, characterized in that it comprises:
[0011] A data acquisition module, used to acquire an initial data group of each object to be processed in the rail transit system network and scene data of the object to be processed based on the initial safety state dimension; wherein the initial data group includes state attribute data corresponding to the initial safety state dimension;
[0012] A data dimension reduction module, used for performing dimension reduction processing on the initial data group to obtain a target data group of each object to be processed;
[0013] A data clustering module, used to determine the state similarity between the corresponding objects to be processed according to each of the target data groups, and cluster the objects to be processed according to the state similarity to obtain a reference cluster and a cluster center of the reference cluster;
[0014] A distance determination module, used to determine the center distance between each of the objects to be processed and the cluster center;
[0015] The safety processing module is used to iteratively update the reference cluster according to the center distance to obtain a target cluster, and perform safety processing on the rail transit system network according to the target cluster and the scene data.
[0016] According to another aspect of the present invention, there is provided an electronic device, comprising:
[0017] one or more processors;
[0018] A memory for storing one or more programs;
[0019] When one or more programs are executed by one or more processors, the one or more processors can execute any one of the safety processing methods for the rail transit system network provided by the embodiments of the present invention.
[0020] According to another aspect of the present invention, a computer-readable storage medium is provided, which stores computer instructions, and the computer instructions are used to enable a processor to implement any one of the rail transit system line network safety processing methods provided in the embodiments of the present invention when executed.
[0021] An embodiment of the present invention provides a safety processing solution for a rail transit system network. Based on an initial safety state dimension, an initial data group of each object to be processed in the rail transit system network and scene data of the object to be processed are obtained; wherein the initial data group includes state attribute data corresponding to the initial safety state dimension; the initial data group is subjected to dimensionality reduction processing to obtain a target data group of each object to be processed; according to each target data group, the state similarity between the corresponding objects to be processed is determined, and according to the state similarity, the objects to be processed are clustered to obtain a reference cluster and a cluster center of the reference cluster; the center distance between each object to be processed and the cluster center is determined; according to the center distance, the reference cluster is iteratively updated to obtain a target cluster, and according to the target cluster and the scene data, the rail transit system network is safely processed. The above scheme determines the target cluster by iterating the objects to be processed, thereby improving the accuracy of the determined target cluster, and further improving the accuracy of the safe processing of the rail transit system network based on the target cluster and scene data; at the same time, by reducing the dimension of the initial data group to obtain the target data group, the influence of redundant data on the subsequent determination of the target cluster is avoided, and the efficiency of the subsequent determination of the target cluster based on the target data group is improved.
[0022] It should be understood that the contents described in this section are not intended to identify the key or important features of the embodiments of the present invention, nor are they intended to limit the scope of the present invention. Other features of the present invention will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS
[0023] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.
[0024] Figure 1 This is a flow chart of a safety processing method for a rail transit system network provided by Embodiment 1 of the present invention;
[0025] Figure 2 is a flow chart of a safety processing method for a rail transit system network provided by Embodiment 2 of the present invention;
[0026] Figure 3 It is a schematic diagram of the structure of a safety processing device for a rail transit system network provided by a fourth embodiment of the present invention;
[0027] Figure 4 It is a structural schematic diagram of an electronic device for implementing a safe processing method for a rail transit system network provided in Embodiment 5 of the present invention. DETAILED DESCRIPTION
[0028] The present invention will be further described in detail below in conjunction with the accompanying drawings and embodiments. It is to be understood that the specific embodiments described herein are only used to explain the present invention, rather than to limit the present invention. It should also be noted that, for ease of description, only parts related to the present invention, rather than all structures, are shown in the accompanying drawings.
[0029] Embodiment 1
[0030] Figure 1 It is a flow chart of a method for safety processing of a rail transit system network provided in Example 1 of the present invention. This embodiment is applicable to situations where safety processing of a rail transit system network is performed. The method can be executed by a safety processing device for a rail transit system network. The device can be implemented in software and / or hardware and can be configured in an electronic device that carries the safety processing function of the rail transit system network.
[0031] See also Figure 1 The safety treatment method of the rail transit system network shown includes:
[0032] S110. Based on the initial safety state dimension, an initial data group of each object to be processed in the rail transit system network and scene data of the object to be processed are obtained.
[0033] The initial safety state dimension can be understood as an indicator for safety assessment of the object to be processed. The implementation of the present invention does not impose any limitation on the initial safety state dimension, which can be set by technicians based on experience. Exemplarily, the initial safety state dimension can include at least one of passenger flow distribution balance, station passenger flow congestion, failure rate, average recovery time, average train density, service reliability, and the number of extreme weather occurrences.
[0034] Among them, the object to be processed refers to any object in the rail transit system network that needs to be evaluated for safety status. Exemplarily, the object to be processed can be one of the rail stations, rail station lines and rail lines in the rail transit system network. The initial data group refers to the numerical data group obtained based on the initial safety status dimension. The initial data group includes state attribute data corresponding to the initial safety status dimension. State attribute data refers to the numerical value corresponding to any initial safety status dimension. Scenario data refers to the data of the operating scenario in which the object to be processed is located. Exemplarily, the scenario data may include off-peak hours, morning and evening peak hours, and failure periods.
[0035] In an optional embodiment, based on the initial safety status dimension, an initial data group of each object to be processed in the rail transit system network is obtained, including: determining the type of objects to be processed in the rail transit system network, and determining a target processing type from the types to be processed; determining at least one object to be processed under the target processing type, and based on the initial safety status dimension, obtaining an initial data group of each object to be processed.
[0036] The type to be processed refers to the type of objects included in the rail transit system network. The target processing type refers to the type of objects for safety assessment in the rail transit system network. Exemplarily, the target processing type may be a station type, a station line type, and a network type.
[0037] It should be noted that the objects to be processed in the embodiments of the present invention are of the same processing type, that is, the objects to be processed may be one of the rail sites under the site type, the rail station lines under the station line type, and the rail line networks under the line network type.
[0038] It can be understood that by determining the object to be processed according to the target processing type, the accuracy of the determined object to be processed is improved.
[0039] S120, performing dimensionality reduction processing on the initial data group to obtain a target data group for each object to be processed.
[0040] The target data set refers to the initial data set after dimensionality reduction processing.
[0041] In an optional embodiment, a dimensionality reduction process is performed on the initial data group to obtain a target data group for each object to be processed, including: determining an initial similarity matrix between each object to be processed in terms of security status based on state attribute data of each object to be processed in terms of the initial security status dimension; reducing the dimension of the initial security status dimension to obtain a candidate security status dimension; determining a candidate similarity matrix between each object to be processed in terms of security status based on candidate attribute data of each object to be processed in terms of the candidate security status dimension; determining a target security status dimension from the candidate security status dimension based on the initial similarity matrix, the candidate similarity matrix and a preset similarity threshold; and performing dimensionality reduction process on the initial data group based on the target security status dimension to obtain a target data group for each object to be processed.
[0042] Among them, the initial similarity matrix refers to the matrix of the similarity between the state attribute data of each object to be processed without dimensionality reduction. The candidate security state dimension refers to the security state dimension after the initial security state dimension is reduced in dimension. The candidate attribute data refers to the state attribute data after the candidate security state dimension is reduced in dimension. The candidate similarity matrix refers to the matrix of the similarity between the state attribute data of each object to be processed after dimensionality reduction. The embodiment of the present invention does not impose any limitation on the size of the preset similarity threshold, which can be set by technicians based on experience, or repeatedly determined through a large number of experiments. The target security state dimension refers to the final dimension after the initial data group is reduced in dimension. The target data group includes target attribute data. The target attribute data refers to the state attribute data after the target security state dimension is reduced in dimension.
[0043] In order to better display related content later in the embodiment of the present invention, the target security status dimension may be two-dimensional.
[0044] It can be understood that by determining the target security status dimension based on the initial similarity matrix, the candidate similarity matrix and the preset similarity threshold, the accuracy of the determined target security status dimension is improved, thereby improving the accuracy of dimensionality reduction.
[0045] Exemplarily, according to the initial similarity matrix, the candidate similarity matrix and the preset similarity threshold, the target security state dimension is determined from the candidate security state dimension, including: determining the reference similarity matrix according to the initial similarity matrix and the candidate similarity matrix; determining the similarity coefficient according to the reference similarity matrix and the candidate similarity matrix; determining the target security state dimension from the candidate security state dimension according to the similarity coefficient and the preset similarity threshold. Among them, the reference similarity matrix refers to a similarity matrix determined based on a preset similarity algorithm according to the initial similarity matrix and the candidate similarity matrix. The embodiment of the present invention does not impose any limitation on the preset similarity algorithm, which can be set by a technician based on experience. Exemplarily, the preset similarity algorithm can be a Pool-Adjacent Violator algorithm. Specifically, the candidate similarity matrix corresponds to the reference similarity matrix one-to-one. The similarity coefficient can be understood as the degree of similarity between the candidate similarity matrix and the reference similarity matrix.
[0046] Exemplarily, the target security status dimension is determined from the candidate security status dimensions based on the similarity coefficient and the preset similarity threshold, including: for any candidate security status dimension, if the similarity coefficient under the candidate security status dimension is greater than or equal to the preset similarity threshold, then the candidate security status dimension is prohibited from being used as the target security status dimension; if the similarity coefficient of the candidate security status dimension is less than the preset similarity threshold, then the candidate security status dimension is used as the target security status dimension.
[0047] For example, the embodiment of the present invention can determine the target security status dimension by the following formula:
[0048] ;
[0049] Among them, δ ij Represents the object to be processed G i With the object to be processed G j The spatial distance between them is defined by the Euclidean distance, which is the initial similarity between the two objects to be processed in terms of their security status. The initial similarity represents the similarity of the security status of the two objects to be processed in the high-dimensional space without dimensionality reduction. ij Represents the object to be processed G i With the object to be processed G j The distance in the low-dimensional state space is defined using the Euclidean distance, i.e., d ij Represents the candidate similarity of the security status of the two candidate objects to be processed. The candidate similarity represents the similarity of the security status of the two objects to be processed in the low-dimensional space after dimensionality reduction.
[0050] The embodiment of the present invention can use the Shepard-Kruskal algorithm to determine the target security state dimension. Specifically, the object to be processed G under the candidate security state dimension is arbitrarily selected. p (where p can be used to represent the identification information of the object to be processed. For example, when p is i, it can represent the object to be processed G i ; When p is j, the object to be processed G can be identified j ) ; Based on the candidate attribute data of the object to be processed in the candidate security state dimension, calculate the corresponding candidate similarity matrix , k=0. Where k represents the number of iterations to determine the target security state dimension. Based on the initial similarity matrix of the initial security state dimension and the candidate similarity matrix , using the Pool-Adjacent Violator algorithm, determine the reference similarity matrix Further, the candidate security status dimension is updated, and the candidate attribute data of the object to be processed under the updated candidate security status dimension is determined. The object to be processed G can be updated by the following formula: p Candidate attribute data :
[0051] ;
[0052] Where n represents the number of objects to be processed; θ represents the step size of the iteration.
[0053] Furthermore, the candidate similarity matrix is updated according to the updated candidate attribute data to obtain the candidate similarity matrix corresponding to the new candidate security status dimension , determine the candidate similarity matrix The corresponding reference similarity matrix .
[0054] Furthermore, the similarity coefficient S is determined by the following formula:
[0055] ;
[0056] Furthermore, if the similarity coefficient S is less than a preset similarity threshold, the candidate security state dimension at this time is used as the target security state dimension; if not, it is prohibited to use the candidate security state dimension at this time as the target security state dimension, and the above-mentioned similarity coefficient determination process is repeated, that is, the target security state dimension is iteratively determined.
[0057] S130 , determining the state similarity between the corresponding objects to be processed according to each target data group, and clustering the objects to be processed according to the state similarity to obtain a reference cluster and a cluster center of the reference cluster.
[0058] Among them, the state similarity refers to the similarity of the security states between the objects to be processed. The reference cluster refers to the cluster obtained by clustering the objects to be processed. The cluster center refers to the center of the reference cluster.
[0059] S140: Determine the center distance between each object to be processed and the cluster center.
[0060] The center distance refers to the distance between any object to be processed and any cluster center.
[0061] In the embodiment of the present invention, there may be multiple objects to be processed, multiple reference clusters, and correspondingly, multiple cluster centers. For any object to be processed, the distance between the object to be processed and each cluster center is determined, which is the center distance.
[0062] S150. Iteratively update the reference cluster according to the center distance to obtain the target cluster, and perform safety processing on the rail transit system network according to the target cluster and scene data.
[0063] The target cluster refers to the cluster of the objects to be processed that is finally determined. In the embodiment of the present invention, the number of iterative updates is determined according to the number of changes in the cluster center.
[0064] Exemplarily, for any object to be processed, the reference cluster where the object to be processed is located is used as the current reference cluster, the cluster center of the current reference cluster is used as the current cluster center, the center distance between the object to be processed and the current cluster center is used as the current center distance, and the center distance between the object to be processed and the non-current cluster center in the non-current reference cluster is used as the non-current center distance; if the current center distance is less than or equal to the non-current center distance, then the reference cluster where the object to be processed is located in this iterative update is correct and does not need to be replaced; if the current center distance is greater than the non-current center distance, then the reference cluster where the object to be processed is located in this iterative update is incorrect and needs to be replaced. Further, the object to be processed is replaced to the reference cluster corresponding to the smallest non-current center distance.
[0065] It can be understood that in the embodiment of the present invention, the accuracy of the determined target cluster is improved by iteratively updating the reference cluster according to the number of changes in the cluster center.
[0066] Exemplarily, performing safety processing on the rail transit system network may be performing safety assessment on the rail transit system network.
[0067] In an optional embodiment, after iteratively updating the reference cluster according to the center distance to obtain the target cluster, the method further includes: visually presenting the target cluster to determine the presentation degree of the target cluster; and updating the target cluster according to the presentation degree.
[0068] The presentation degree can be used to quantify the degree of visualization effect of the target cluster.
[0069] Exemplarily, the clustering process of the target cluster can be adjusted according to the presentation degree to update the target cluster. The advantage is that the target cluster obtained by clustering can improve the presentability of the target cluster while ensuring the accuracy.
[0070] It can be understood that by visually displaying the target cluster and determining the presentation degree of the target cluster, an intuitive presentation of the target cluster is achieved, and the presentability of the clustering effect of the target cluster can be judged according to the presentation degree, thereby ensuring the clustering accuracy of the target cluster while improving the presentability of the target cluster and improving the user experience.
[0071] An embodiment of the present invention provides a safety processing solution for a rail transit system network. Based on an initial safety state dimension, an initial data group of each object to be processed in the rail transit system network and scene data of the object to be processed are obtained; wherein the initial data group includes state attribute data corresponding to the initial safety state dimension; the initial data group is subjected to dimensionality reduction processing to obtain a target data group of each object to be processed; according to each target data group, the state similarity between the corresponding objects to be processed is determined, and according to the state similarity, the objects to be processed are clustered to obtain a reference cluster and a cluster center of the reference cluster; the center distance between each object to be processed and the cluster center is determined; according to the center distance, the reference cluster is iteratively updated to obtain a target cluster, and according to the target cluster and the scene data, the rail transit system network is safely processed. The above scheme determines the target cluster by iterating the objects to be processed, thereby improving the accuracy of the determined target cluster, and further improving the accuracy of the safe processing of the rail transit system network based on the target cluster and scene data; at the same time, by reducing the dimension of the initial data group to obtain the target data group, the influence of redundant data on the subsequent determination of the target cluster is avoided, and the efficiency of the subsequent determination of the target cluster based on the target data group is improved.
[0072] Embodiment 2
[0073] Figure 2 It is a flow chart of a method for safely handling a rail transit system network provided in the second embodiment of the present invention. Based on the above embodiments, this embodiment further refines the operation of "determining the state similarity between the corresponding objects to be processed according to each target data group, and clustering the objects to be processed according to the state similarity to obtain a reference cluster" into "if the state similarity includes object state similarity and cluster state similarity, determine the object state similarity between the objects to be processed according to the target data group, and perform initial clustering on the objects to be processed according to the object state similarity to obtain an initial cluster; determine the cluster state similarity between the initial clusters, and iteratively cluster the initial clusters according to the cluster state similarity to obtain a reference cluster" to improve the reference cluster determination mechanism. It should be noted that for the parts not described in detail in the embodiments of the present invention, please refer to the descriptions of other embodiments.
[0074] See also Figure 2 The safety treatment method of the rail transit system network shown includes:
[0075] S210. Based on the initial safety state dimension, an initial data group of each object to be processed in the rail transit system network and scene data of the object to be processed are obtained.
[0076] The initial data group includes state attribute data corresponding to the initial security state dimension;
[0077] S220, performing dimensionality reduction processing on the initial data group to obtain a target data group for each object to be processed.
[0078] S230 , determining the object state similarity between the objects to be processed according to the target data group, and performing initial clustering on the objects to be processed according to the object state similarity to obtain initial clusters.
[0079] The object state similarity refers to the feature similarity between any two objects to be processed in terms of security status, that is, the object state similarity can be understood as the feature similarity between the target data groups of each object to be processed. The initial cluster refers to the cluster obtained by the initial clustering of the objects to be processed.
[0080] For example, if there are U objects to be processed in the rail transit system network, and the number of clusters of the clustered reference clusters is S. The present invention uses the hybrid hierarchical K-means clustering algorithm HHKMC to perform cluster analysis on the target data group of the objects to be processed output by the MDS (Multidimensional Scaling) algorithm, and divides them into S reference clusters. The safety status of the objects to be processed belonging to the same reference cluster has great similarity, and the safety status of the objects to be processed belonging to different reference clusters is quite different.
[0081] Continuing with the above example, in the dimension of initial security status, each of the U objects to be processed is considered as a cluster (or class), that is, it is assumed that the security features of the U objects to be processed are different. The distances between all the objects to be processed are calculated to determine the similarity of the security features between the two objects to be processed. The Euclidean distance is used to define the object G to be processed. i With the object to be processed G j The distance between ij (i.e., object state similarity). The distance d between the objects to be processed ij It can be used to measure the similarity of security features between objects to be processed. i is the object to be processed G i The target data set, . Z j is the object to be processed G j The target data set, . Where T represents transpose.
[0082] Exemplarily, the object state similarity can be determined by the following formula:
[0083] ;
[0084] Among them, dist represents distance; n represents the number of target attribute data in the target data group; m represents the nth target attribute data in the target data group, which can be understood as the identifier of the target attribute data.
[0085] S240 , determining the cluster state similarity between the initial clusters, and iteratively clustering the initial clusters according to the cluster state similarity to obtain a reference cluster and a cluster center of the reference cluster.
[0086] Among them, cluster state similarity refers to the feature similarity between clusters in terms of security state.
[0087] Use the Average Linkage (K-means) strategy to define the initial cluster C v With the initial cluster C e The distance D ve (i.e. cluster state similarity), the distance between the initial clusters is the average distance between all the objects to be processed in the cluster. Cluster state similarity D ve Used to measure the similarity of security features between clusters, D ve The smaller the value, the more similar the security features between clusters. v Represents cluster C v The number of objects to be processed in t e Represents cluster C e The number of objects to be processed.
[0088] Exemplarily, the cluster state similarity can be determined by the following formula:
[0089] ;
[0090] Furthermore, the two initial clusters with the shortest distance are merged into a new cluster, that is, the two initial clusters with the most similar security features are merged into a new cluster, namely, the reference cluster. The reference clusters are iteratively clustered and merged until the U objects to be processed are divided into S reference clusters.
[0091] In an optional embodiment, the number of iterative clustering is determined according to a preset number of reference clusters or cluster state similarity, wherein the preset number of reference clusters refers to the number of pre-set reference clusters.
[0092] Exemplarily, when the number of reference clusters obtained by iterative clustering reaches a preset number of reference clusters, the iterative clustering may be stopped; or, when the cluster state similarities of the reference clusters obtained by iterative clustering are all similar, the iterative clustering may be stopped.
[0093] It can be understood that by determining the number of iterative clustering times according to the preset reference cluster number or cluster state similarity, the accuracy of the determined number of iterative clustering times is improved.
[0094] Exemplarily, S reference clusters C are determined by the following formula: s The respective cluster centers P s :
[0095] ;
[0096] in, Represents the reference cluster C s The i-th object to be processed in t s Represents the reference cluster C s The number of objects to be processed in; Indicates the object to be processed Target attribute data; Represents the reference cluster C s Target attribute data.
[0097] S250: Determine the center distance between each object to be processed and the cluster center.
[0098] For example, each object G to be processed is calculated i With S cluster centers P s The objects to be processed are respectively assigned to the reference clusters with the smallest distance, that is, the objects to be processed are respectively assigned to the reference clusters with the most similar safety status.
[0099] For example, the center distance can be determined by the following formula:
[0100] ;
[0101] The center distance The Euclidean distance definition is used to measure the similarity between the object to be processed and the cluster center; The smaller the value, the more similar the object to be processed is to the cluster center; Represents the object to be processed G i The corresponding reference cluster is C s , that is, relative to other reference clusters, the object to be processed G i With reference cluster C s The security status is most similar.
[0102] S260. Iteratively update the reference cluster according to the center distance to obtain the target cluster, and perform safety processing on the rail transit system network according to the target cluster and scene data.
[0103] Exemplarily, based on the clustering results, the centers of the S reference clusters are updated; the center distances between each object to be processed and the cluster center are re-determined; the reference cluster where the object to be processed is located is updated based on the center distances; until the cluster centers of the S reference clusters do not change, the target cluster is obtained.
[0104] The embodiment of the present invention provides a safety processing scheme for the line network of a rail transit system, which determines the state similarity between the corresponding objects to be processed according to each target data group, and clusters the objects to be processed according to the state similarity to obtain a reference cluster operation, which is refined into: if the state similarity includes object state similarity and cluster state similarity, the object state similarity between the objects to be processed is determined according to the target data group, and the objects to be processed are initially clustered according to the object state similarity to obtain an initial cluster; the cluster state similarity between the initial clusters is determined, and the initial clusters are iteratively clustered according to the cluster state similarity to obtain a reference cluster, thereby improving the reference cluster determination mechanism. The above scheme improves the accuracy of the reference cluster obtained by iterative clustering by introducing object state similarity and cluster state similarity to iteratively cluster the objects to be processed.
[0105] Embodiment 3
[0106] Based on the above embodiments, the embodiments of the present invention propose a rail transit system safety status big data visualization model (Graphical Visualization Model for Big Data of Evaluating Safety, GVM-BDES), i.e., a rail line network classification model, based on the above-mentioned rail transit system line network safety processing method. The rail line network classification model refers to a model that clusters the objects to be processed in the rail transit system line network and displays the clustering results.
[0107] At present, the research on rail transit system network mainly focuses on network structure, service level, network performance, reliability and other aspects. The methods of safety risk assessment are relatively simple, mostly using qualitative, quantitative or a combination of the two methods. For example, the fault tree analysis method is used for relatively simple analysis of engineering system safety or accident risks, and is not suitable for regional comprehensive risk assessment; the probability risk assessment method analyzes the risk data and experimental results data that have occurred and evaluates various risks, but it has many uncertain variables that make the calculation amount large; the Monte Carlo method is for risk events with a certain probability distribution or a large number of samples, but it can only solve the risk assessment of determined risks and relatively simple risk problems. Most statistical models cannot solve complex situations with more qualitative and quantitative objectives.
[0108] With the continuous advancement of rail transit informatization and the widespread use of more inspection and testing methods, safety-related status data is becoming increasingly abundant. However, managers currently mainly use simple graphics such as line charts and bar charts to intuitively display analysis results, which makes it difficult to deeply explore the potential patterns in the data and to intuitively and clearly display relevant information in massive data. Data visualization analysis is an important research direction in big data analysis. The basic idea of data visualization analysis is to organize and integrate data with the help of computer graphics and image processing technology, observe data from different dimensions, and conduct a deeper analysis of data to achieve the acquisition of three types of knowledge: (1) Anomaly: determine whether there is "problem" data in the data set; (2) Relationship: analyze the correlation between different data; (3) Pattern: analyze certain patterns in the data.
[0109] In the embodiment of the present invention, accurate quantitative comprehensive assessment of the safety status of the rail transit system is of great significance for rail transit managers to grasp the changes in safety status during the operation of multi-standard rail transit and to reasonably allocate transportation capacity and operation and maintenance resources. The embodiment of the present invention proposes a big data visualization method for the safety status of the rail transit system for the point-line-network multi-level evaluation objects in the rail transit system, and uses a multidimensional scaling analysis algorithm and a hybrid hierarchical K-means clustering algorithm to visualize the distribution of the safety status of the rail system, intuitively reflecting the similarities or differences of the characteristics of each state. The evaluation results can avoid the interference of human subjective factors and are more in line with the actual situation.
[0110] Specifically, the safety processing model of the rail transit system network uses a multidimensional scaling analysis algorithm and a hybrid hierarchical K-means clustering algorithm to visualize the distribution of the rail system's safety status, optimize parameters to reduce the amount of calculation, avoid the influence of human factors, and conduct a more accurate and comprehensive comprehensive risk assessment of the rail transit system. The safety processing model of the rail transit system network intuitively displays the similarities or differences in the comprehensive safety status of each evaluation object, allowing managers to comprehensively and intuitively grasp the distribution of safety status in different time-space ranges, and provide decision support for rail transit operation management.
[0111] In an embodiment of the present invention, a rail network classification model is provided, which can visually display the similarities or differences of the safety state characteristics of the rail system while better maintaining the similarity relationship of the safety state of the rail transit system in the original high-dimensional state space.
[0112] For example, in the rail network classification model, the multi-level evaluation objects of the rail transit system, such as the midpoint (rail station) - line (rail station line) - network (rail line network), are used as the objects to be processed, so that users can understand the safety status of the rail network, line, section and other levels; the multi-state evaluation index of the rail transit system safety is reduced in dimension. In the multi-dimensional space composed of the multi-state evaluation index (i.e., the initial safety state dimension), the distance between the objects to be evaluated is used to define the similarity of the safety state of the objects to be processed, and the multi-dimensional scaling analysis algorithm (Multidimensional Scaling, MDS) is used to reduce the dimension of the multi-dimensional state evaluation index of the regional rail transit system safety, i.e., determine the target safety state dimension, while maintaining the similarity of the safety state of the objects to be processed; the rail transit system safety feature clustering is performed using the hybrid hierarchical K-means clustering algorithm (Hybrid Hierarchical K-means Clustering (HHKMC) is used to perform cluster analysis on the safety status characteristics of each object to be processed based on the dimensionality reduction calculation results. At the same time, the spatial distribution of the rail transit system state is visualized in the low-dimensional state space, and the similarities or differences in the safety status of each object to be processed in the original high-dimensional state space are analyzed.
[0113] In summary, the calculation process of the rail network classification model includes: obtaining high-dimensional safety status assessment index data of the rail transit system, that is, based on the initial safety status dimension, obtaining the initial data group of each object to be processed in the rail transit system network; based on the multidimensional scaling analysis algorithm, reducing the dimension of the initial data group; newly constructed low-dimensional safety status assessment index data, that is, the target data group; clustering the objects to be processed based on the hybrid hierarchical K-means clustering algorithm; clustering characteristics of the rail transit system safety status, that is, the target cluster; visualization of the low-dimensional state space distribution, that is, displaying the target cluster.
[0114] In the embodiment of the present invention, in the track network classification model, the process of performing dimensionality reduction processing on the initial data group to obtain the target data group is as follows:
[0115] For example, if It represents the distance matrix of the rail transit system in the multi-dimensional space composed of the original multi-state evaluation indicators. The spatial dimension is represented by m, which is determined according to the number of state evaluation indicators. That is, ∆ can be used to represent the initial similarity matrix. ij Represents the object to be processed G i With the object to be processed G j The spatial distance between them is defined by Euclidean distance, which is the initial similarity between the two objects to be processed in terms of security status. The initial similarity indicates the similarity of the security status of the two objects to be processed in the high-dimensional space without dimensionality reduction.
[0116] It should be noted that δij The smaller it is, the more similar the security status characteristics of the objects to be processed are. represents the distance matrix of the rail transit system in the newly generated low-dimensional state space, and the spatial dimension is represented by n, that is, D represents the candidate similarity matrix. In the embodiment of the present invention, the spatial dimension m>n of the initial similarity matrix. d ij Represents the object to be processed G i With the object to be processed G j The distance in the low-dimensional state space is defined using the Euclidean distance, i.e., d ij Represents the candidate similarity of the security status of the two candidate objects to be processed. The candidate similarity represents the similarity of the security status of the two objects to be processed in the low-dimensional space after dimensionality reduction.
[0117] The purpose of the multidimensional scaling (MDS) algorithm is to make the distance d of the object to be processed in the low-dimensional space ij As close as possible to the security status of the object to be processed ij Close. Use the distance d of the object to be processed in the low-dimensional state space ij , which approximately represents the similarity of the security status of each object to be processed. ij The similarities and differences between the security states of the objects to be processed can be studied.
[0118] The training process of the track network classification model in the embodiment of the present invention may adopt the target cluster determination process in other embodiments.
[0119] Exemplarily, the initial data set of the object to be processed is collected according to the initial safety state dimension. Among them, the initial safety state dimension may include passenger flow distribution balance, station passenger flow congestion, failure rate, average recovery time, average train density, service reliability and the number of extreme weather occurrences, then the initial safety state dimension is 7 dimensions. The embodiment of the present invention does not impose any restrictions on the time of collecting data, which can be set by technicians based on experience. Exemplarily, the data can be collected during peak hours in the morning and evening, off-peak hours and during failure periods.
[0120] For example, the initial data set is input into the trained track network classification model. The MDS algorithm in the track network classification model reduces the initial safety state dimension from 7 dimensions to 2 dimensions (i.e., the target safety state dimension). After 23 iterations, the change in pressure stress is less than 1×10 -6, stop iteration. Stress = 0.1098 < 0.2, which shows that the dimensionality reduction effect of the model is good. The newly constructed 2D state space of rail transit system safety can better maintain the similarity relationship of the safety state of the rail transit system in the original 7D state space. The 2D state index obtained by dimensionality reduction has no clear physical meaning. The Euclidean distance calculated by it is used to measure the similarity of the safety state of the rail transit system.
[0121] Furthermore, a 2-dimensional spatial distribution diagram of the objects to be processed in the target security status dimension can be constructed. The horizontal axis in the diagram represents the candidate attribute data y1 newly constructed by the model, and the vertical axis represents the newly constructed candidate attribute data y2. The dots in the diagram represent the objects to be processed. The closer the spatial distance between the two objects to be processed in the diagram, the more similar the security status of the two objects to be processed.
[0122] The embodiment of the present invention uses the R package FactoMineR to implement the hybrid hierarchical K-means clustering algorithm in the rail network classification model GVM-BDES. By comparing the size of the overall silhouette coefficient under different cluster numbers, when the number of clusters is 5, the overall silhouette coefficient is the largest and the clustering effect is the best. Therefore, the safety features of the regional rail transit system in this embodiment are divided into 5 categories (5 clusters), that is, the preset number of reference clusters is 5. In the image showing the target cluster, the objects to be processed belonging to different target clusters can be represented by points of different colors and / or shapes.
[0123] The clustering results of the rail network classification model GVM-BDES, that is, the target clusters and scene data, were superimposed and analyzed. In the image where the target clusters and scene data are superimposed, the objects to be processed belonging to different target clusters are represented by different colors, such as red for the objects to be processed in target cluster Cluster=1, yellow for the objects to be processed in target cluster Cluster=2, green for the objects to be processed in target cluster Cluster=3, blue for the objects to be processed in target cluster Cluster=4, and purple for the objects to be processed in target cluster Cluster=5. The scene data superimposed in different operation scenarios are represented by different numbers, and the objects to be processed in the off-peak period are represented by "1", the objects to be processed in the morning and evening peak periods are represented by "2", and the objects to be processed in the fault period are represented by "3".
[0124] For example, see the confusion matrix of the operation scenario of the object to be processed and the target cluster shown in Table 1. The operation scenario of the object to be processed is displayed in the corresponding row, and the target cluster is displayed in the corresponding column.
[0125] By analyzing Table 1, we can see that the objects to be processed during the off-peak period mainly correspond to target clusters = 2, 3; the objects to be processed during the morning and evening peak periods mainly correspond to Cluster = 1, 5; the objects to be processed during the failure period mainly correspond to Cluster = 4. It can be clearly seen that in this example, the safety status of the objects to be processed judged by the proposed model is basically consistent with the classification of the operation scenario. The main difference is that the proposed model further subdivides the state of the objects to be processed during the morning and evening peak periods and the off-peak period into two small categories. Through further analysis, the main reason is that the safety characteristics of the morning peak and the evening peak are different, and the safety characteristics of the off-peak period on holidays and weekdays are also different.
[0126] Table 1 Confusion matrix of the operation scenarios and target clusters of the objects to be processed
[0127]
[0128] To sum up, the rail network classification model GVM-BDES can achieve different degrees of distinction between objects with large status differences, perform good clustering and resolution of objects with similar status, and present the distribution of safety status intuitively and clearly.
[0129] The embodiment of the present invention starts from the perspective of rail transit system safety level evaluation, and proposes a rail transit system safety status big data visualization model for point-line-network multi-level processing objects in the rail transit system. It adopts a multidimensional scaling analysis algorithm and a hybrid hierarchical K-means clustering algorithm to visualize the distribution of the rail system safety status, intuitively reflects the similarities or differences of the characteristics of each state, and clearly shows the similarities or differences of the safety status of the objects to be processed at each level, thereby improving the managers' ability to analyze and mine data and improving the managers' overall grasp of the distribution of the rail system safety status.
[0130] The embodiment of the present invention accurately conducts a quantitative comprehensive assessment of the safety status of the rail transit system, which is of great significance for rail transit managers to grasp the changes in safety status during the operation of multi-standard rail transit and to reasonably allocate transportation capacity and operation and maintenance resources. The embodiment of the present invention proposes a big data visualization model of the safety status of the rail transit system for the point-line-network multi-level objects to be processed in the rail transit system. It uses a multidimensional scaling analysis algorithm and a hybrid hierarchical K-means clustering algorithm to visualize the distribution of the safety status of the rail system, and intuitively reflects the similarities or differences of the characteristics of each state. The result can avoid the interference of human subjective factors and is more in line with the actual situation. The embodiment of the present invention can intuitively display the similarities or differences of the comprehensive safety status of each object to be processed, allowing managers to comprehensively and intuitively grasp the distribution of safety status in different time-space ranges, and provide decision support for rail transit operation management.
[0131] Embodiment 4
[0132] Figure 3 This is a schematic diagram of the structure of a safety processing device for a rail transit system network provided in Embodiment 4 of the present invention. This embodiment is applicable to the case of performing safety processing on a rail transit system network. The method can be executed by a safety processing device for a rail transit system network. The device can be implemented in software and / or hardware and can be configured in an electronic device that carries the safety processing function of the rail transit system network.
[0133] like Figure 3 As shown, the device includes: a data acquisition module 310, a data dimension reduction module 320, a data clustering module 330, a distance determination module 340 and a security processing module 350. Among them,
[0134] The data acquisition module 310 is used to acquire the initial data group of each object to be processed in the rail transit system network and the scene data of the object to be processed based on the initial safety state dimension; wherein the initial data group includes the state attribute data corresponding to the initial safety state dimension;
[0135] A data dimension reduction module 320 is used to perform dimension reduction processing on the initial data group to obtain a target data group for each object to be processed;
[0136] A data clustering module 330 is used to determine the state similarity between the objects to be processed according to each of the target data groups, and cluster the objects to be processed according to the state similarity to obtain a reference cluster and a cluster center of the reference cluster;
[0137] A distance determination module 340 is used to determine the center distance between each of the objects to be processed and the cluster center;
[0138] The safety processing module 350 is used to iteratively update the reference cluster according to the center distance to obtain a target cluster, and perform safety processing on the rail transit system network according to the target cluster and the scene data.
[0139] An embodiment of the present invention provides a safety processing solution for a rail transit system network. Based on an initial safety state dimension, an initial data group of each object to be processed in the rail transit system network and scene data of the object to be processed are obtained; wherein the initial data group includes state attribute data corresponding to the initial safety state dimension; the initial data group is subjected to dimensionality reduction processing to obtain a target data group of each object to be processed; according to each target data group, the state similarity between the corresponding objects to be processed is determined, and according to the state similarity, the objects to be processed are clustered to obtain a reference cluster and a cluster center of the reference cluster; the center distance between each object to be processed and the cluster center is determined; according to the center distance, the reference cluster is iteratively updated to obtain a target cluster, and according to the target cluster and the scene data, the rail transit system network is safely processed. The above scheme determines the target cluster by iterating the objects to be processed, thereby improving the accuracy of the determined target cluster, and further improving the accuracy of the safe processing of the rail transit system network based on the target cluster and scene data; at the same time, by reducing the dimension of the initial data group to obtain the target data group, the influence of redundant data on the subsequent determination of the target cluster is avoided, and the efficiency of the subsequent determination of the target cluster based on the target data group is improved.
[0140] Optionally, if the state similarity includes object state similarity and cluster state similarity, the data clustering module 330 includes:
[0141] an initial cluster determination unit, configured to determine the object state similarity between the objects to be processed according to the target data group, and to perform initial clustering on the objects to be processed according to the object state similarity to obtain initial clusters;
[0142] The reference cluster determining unit is used to determine the cluster state similarity between the initial clusters, and iteratively cluster the initial clusters according to the cluster state similarity to obtain the reference cluster.
[0143] Optionally, the number of iterative clustering is determined according to a preset number of reference clusters or cluster state similarity.
[0144] Optionally, the number of iterative updates is determined according to the number of changes in the cluster center.
[0145] Optionally, the data acquisition module 310 includes:
[0146] A processing type determination unit, used to determine the types to be processed in the rail transit system network, and determine the target processing type from the types to be processed;
[0147] The initial data group acquisition unit is used to determine at least one to-be-processed object under the target processing type, and acquire an initial data group of each to-be-processed object based on the initial security state dimension.
[0148] Optionally, the data dimension reduction module 320 includes:
[0149] An initial similarity matrix determining unit, used to determine an initial similarity matrix between the objects to be processed in terms of security status according to the state attribute data of each object to be processed in terms of the initial security status dimension;
[0150] A candidate dimension determination unit, configured to reduce the dimension of the initial security state dimension to obtain a candidate security state dimension;
[0151] A candidate similarity matrix determining unit, configured to determine a candidate similarity matrix between the objects to be processed in terms of security status according to the candidate attribute data of each object to be processed in terms of the candidate security status dimension;
[0152] a target dimension determination unit, configured to determine a target security state dimension from the candidate security state dimensions according to the initial similarity matrix, the candidate similarity matrix and a preset similarity threshold;
[0153] The target data group determination unit is used to perform dimensionality reduction processing on the initial data group according to the target security status dimension to obtain the target data group of each object to be processed.
[0154] Optionally, after iteratively updating the reference cluster according to the center distance to obtain the target cluster, the device further includes:
[0155] A target cluster display module, used for visually displaying the target cluster to determine the presentation degree of the target cluster;
[0156] The target cluster updating module is used to update the target cluster according to the presentation degree.
[0157] The safety processing device for the rail transit system network provided in the embodiment of the present invention can execute the safety processing method for the rail transit system network provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects for executing the safety processing method for each rail transit system network.
[0158] In the technical solution of the present invention, the collection, storage, use, processing, transmission, provision and disclosure of the initial data groups and scene data involved are in compliance with the provisions of relevant laws and regulations and do not violate public order and good morals.
[0159] Embodiment 5
[0160] Figure 4It is a structural schematic diagram of an electronic device for implementing a safe processing method for a rail transit system line network provided by Embodiment 5 of the present invention. The electronic device 410 is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workbenches, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device may also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices (such as helmets, glasses, watches, etc.) and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present invention described and / or required herein.
[0161] like Figure 4 As shown, the electronic device 410 includes at least one processor 411, and a memory connected to the at least one processor 411 in communication, such as a read-only memory (ROM) 412, a random access memory (RAM) 413, etc., wherein the memory stores a computer program that can be executed by at least one processor, and the processor 411 can perform various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 412 or the computer program loaded from the storage unit 418 to the random access memory (RAM) 413. In the RAM 413, various programs and data required for the operation of the electronic device 410 can also be stored. The processor 411, the ROM 412, and the RAM 413 are connected to each other via a bus 414. An input / output (I / O) interface 415 is also connected to the bus 414.
[0162] Multiple components in the electronic device 410 are connected to the I / O interface 415, including: an input unit 416, such as a keyboard, a mouse, etc.; an output unit 417, such as various types of displays, speakers, etc.; a storage unit 418, such as a disk, an optical disk, etc.; and a communication unit 419, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 419 allows the electronic device 410 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.
[0163] The processor 411 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the processor 411 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, etc. The processor 411 executes the various methods and processes described above, such as a safety processing method for a rail transit system network.
[0164] In some embodiments, the safety processing method for the rail transit system network can be implemented as a computer program, which is tangibly contained in a computer-readable storage medium, such as a storage unit 418. In some embodiments, part or all of the computer program can be loaded and / or installed on the electronic device 410 via the ROM 412 and / or the communication unit 419. When the computer program is loaded into the RAM 413 and executed by the processor 411, one or more steps of the safety processing method for the rail transit system network described above can be performed. Alternatively, in other embodiments, the processor 411 can be configured to execute the safety processing method for the rail transit system network in any other appropriate manner (for example, by means of firmware).
[0165] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.
[0166] Computer programs for implementing the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that when the computer program is executed by the processor, the functions / operations specified in the flow chart and / or block diagram are implemented. The computer program may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.
[0167] In the context of the present invention, a computer-readable storage medium may be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, device, or equipment. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, or equipment, or any suitable combination of the foregoing. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.
[0168] To provide interaction with a user, the systems and techniques described herein may be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and a pointing device (e.g., a mouse or trackball) through which the user can provide input to the electronic device. Other types of devices may also be used to provide interaction with the user; for example, the feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form (including acoustic input, voice input, or tactile input).
[0169] The systems and techniques described herein may be implemented in a computing system that includes backend components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes frontend components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such backend components, middleware components, or frontend components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), a blockchain network, and the Internet.
[0170] A computing system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The client and server relationship is generated by computer programs running on the corresponding computers and having a client-server relationship with each other. The server may be a cloud server, also known as a cloud computing server or cloud host, which is a host product in the cloud computing service system to solve the defects of difficult management and weak business scalability in traditional physical hosts and VPS services.
[0171] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps described in the present invention can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solution of the present invention can be achieved, and this document does not limit this.
[0172] The above specific implementations do not constitute a limitation on the protection scope of the present invention. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims
1. A method for safely handling a rail transit system network. It is characterized in that include: Based on the initial safety state dimension, an initial data group of each object to be processed in the rail transit system network and scene data of the object to be processed are obtained; wherein the initial data group includes state attribute data corresponding to the initial safety state dimension; Performing dimensionality reduction processing on the initial data group to obtain a target data group for each object to be processed; According to each of the target data groups, determining the state similarity between the corresponding objects to be processed, and clustering the objects to be processed according to the state similarity to obtain a reference cluster and a cluster center of the reference cluster; Determine the center distance between each of the objects to be processed and the center of the cluster; Iteratively updating the reference cluster according to the center distance to obtain a target cluster, and performing safety processing on the rail transit system network according to the target cluster and the scene data; The step of performing dimensionality reduction processing on the initial data group to obtain a target data group of each object to be processed includes: Determine an initial similarity matrix between the objects to be processed in terms of security status according to the state attribute data of the objects to be processed in the initial security status dimension; Performing dimension reduction on the initial security state dimension to obtain a candidate security state dimension; Determine a candidate similarity matrix in terms of security status between the objects to be processed according to the candidate attribute data of the objects to be processed in the candidate security status dimension; Determining a target security state dimension from the candidate security state dimensions according to the initial similarity matrix, the candidate similarity matrix and a preset similarity threshold; According to the target security status dimension, the initial data group is subjected to dimensionality reduction processing to obtain a target data group of each object to be processed; The step of determining the target security state dimension from the candidate security state dimensions according to the initial similarity matrix, the candidate similarity matrix and the preset similarity threshold comprises: Determine a reference similarity matrix according to the initial similarity matrix and the candidate similarity matrix; Determining a similarity coefficient according to the reference similarity matrix and the candidate similarity matrix; The target security state dimension is determined from the candidate security state dimensions according to the similarity coefficient and a preset similarity threshold.
2. The method according to claim 1, It is characterized in that If the state similarity includes object state similarity and cluster state similarity, then determining the state similarity between the corresponding objects to be processed according to each of the target data groups, and clustering the objects to be processed according to the state similarity to obtain a reference cluster, including: Determining the object state similarity between the objects to be processed according to the target data group, and performing initial clustering on the objects to be processed according to the object state similarity to obtain initial clusters; The cluster state similarities between the initial clusters are determined, and the initial clusters are iteratively clustered according to the cluster state similarities to obtain the reference clusters.
3. The method according to claim 2, It is characterized in that The number of iterative clustering is determined according to the preset number of reference clusters or the similarity of cluster states.
4. The method according to claim 1, It is characterized in that The number of iterative updates is determined according to the number of changes in the cluster center.
5. The method according to any one of claims 1 to 4, It is characterized in that The step of obtaining an initial data group of each object to be processed in the rail transit system network based on the initial safety state dimension includes: Determine the types to be processed in the rail transit system network, and determine the target processing type from the types to be processed; At least one to-be-processed object under the target processing type is determined, and based on the initial security state dimension, an initial data group of each to-be-processed object is acquired.
6. The method according to any one of claims 1 to 4, It is characterized in that After the reference cluster is iteratively updated according to the center distance to obtain the target cluster, the method further includes: Visually presenting the target cluster to determine the degree of presentation of the target cluster; The target cluster is updated according to the presentation degree.
7. A safety processing device for a rail transit system network, It is characterized in that include: A data acquisition module, used to acquire an initial data group of each object to be processed in the rail transit system network and scene data of the object to be processed based on the initial safety state dimension; wherein the initial data group includes state attribute data corresponding to the initial safety state dimension; A data dimension reduction module, used for performing dimension reduction processing on the initial data group to obtain a target data group of each object to be processed; A data clustering module, used to determine the state similarity between the corresponding objects to be processed according to each of the target data groups, and cluster the objects to be processed according to the state similarity to obtain a reference cluster and a cluster center of the reference cluster; A distance determination module, used to determine the center distance between each of the objects to be processed and the cluster center; A safety processing module, used for iteratively updating the reference cluster according to the center distance to obtain a target cluster, and performing safety processing on the rail transit system network according to the target cluster and the scene data; Wherein, the data dimension reduction module includes: An initial similarity matrix determining unit, used to determine an initial similarity matrix between the objects to be processed in terms of security status according to the state attribute data of each object to be processed in terms of the initial security status dimension; A candidate dimension determination unit, configured to reduce the dimension of the initial security state dimension to obtain a candidate security state dimension; A candidate similarity matrix determining unit, configured to determine a candidate similarity matrix between the objects to be processed in terms of security status according to the candidate attribute data of each object to be processed in terms of the candidate security status dimension; a target dimension determination unit, configured to determine a target security state dimension from the candidate security state dimensions according to the initial similarity matrix, the candidate similarity matrix and a preset similarity threshold; A target data group determination unit, configured to perform dimensionality reduction processing on the initial data group according to the target security state dimension to obtain a target data group for each of the objects to be processed; Wherein, the target dimension determination unit is specifically used to: Determine a reference similarity matrix according to the initial similarity matrix and the candidate similarity matrix; Determining a similarity coefficient according to the reference similarity matrix and the candidate similarity matrix; The target security state dimension is determined from the candidate security state dimensions according to the similarity coefficient and a preset similarity threshold.
8. An electronic device, It is characterized in that include: one or more processors; A memory for storing one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors implement a safe processing method for a rail transit system network as described in any one of claims 1-6.
9. A computer-readable storage medium having a computer program stored thereon, It is characterized in that When the program is executed by a processor, a safety processing method for a rail transit system network as described in any one of claims 1 to 6 is implemented.
Citation Information
Patent Citations
Two-stage hybrid particle swarm optimization clustering method
CN102663100A
Intersection blind area early warning method and system based on dynamic image perception
CN110992683A
Subway lighting system abnormal energy consumption analysis method based on multi-attribute clustering
CN111695792A
Decentralized federated clustering learning method and device, equipment and medium
CN115545215A
Rail health state evaluation method
CN116862308A
