A rail transit data analysis method and system based on cloud computing

Abstract

The application discloses a kind of rail transit data analysis method and system based on cloud computing, it is related to rail transit data analysis technical field, the global operation data of target rail transit network is collected, dynamic panoramic topological graph is constructed based on the global operation data, dynamic panoramic topological graph is segmented into N traffic area subgraph, the elastic cloud computing node mapped to the N traffic area subgraph is configured, global operation data is distributed to corresponding elastic cloud computing node, in combination with cloud computing processing model, global operation data initialization screening and cloud association conversion are carried out by data clustering algorithm, determine abnormal data distribution, after cloud platform monitors key event, the abnormal data of each elastic cloud computing node is summarized, the commonality of abnormal data is mined, in combination with real-time passenger flow evolution trend determines predicted abnormal data distribution, obtains abnormal distribution graph, and abnormal distribution graph is mapped in dynamic panoramic topological graph in real time, generates global early warning information.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of rail transit data analysis, and in particular to a rail transit data analysis method and system based on cloud computing. BACKGROUND

[0002] With the continuous expansion of the scale of the rail transit network and the continuous improvement of the complexity of operation, data-driven fine management has become the core demand of the industry development. At present, a large amount of multi-source data is generated in the process of operation of the rail transit system, and the traditional dimensions of multi-source data cover passenger flow distribution, train operation state and equipment working condition, and the public opinion feedback in social media also gradually becomes a key factor affecting the operation efficiency. However, the existing data analysis technology still has significant bottlenecks: on the one hand, the degree of data integration is insufficient, most systems only analyze single type of data in isolation, and cannot realize the deep linkage of multi-source data, resulting in prominent data island phenomenon and difficulty in capturing cross-modal correlation anomalies; on the other hand, the allocation of computing resources is static, which cannot adapt to the passenger flow tide changes in peak and off-peak hours and holiday scenes, and there are often problems of insufficient computing resources in high-load areas and waste of resources in low-load areas; in addition, the abnormal processing is mostly limited to real-time identification, lacks the prediction ability of abnormal diffusion path, and the warning information mostly only includes basic location and type without integrating decision support dimensions such as economic impact evaluation, resulting in insufficient response and passenger experience optimization of the operation party.

[0003] At present, the Chinese invention with the application number CN120589070A discloses a rail transit remote monitoring method and system based on artificial intelligence, which realizes the positioning and remote monitoring of equipment failure through artificial intelligence technology, and assists maintenance and disposal combined with augmented reality technology. The above technology cannot meet the fine analysis needs of the whole rail transit operation: firstly, the data coverage dimension is limited, focusing on the monitoring and fault diagnosis of equipment operation data, without including social media public opinion data in the analysis framework, which cannot capture the cross-modal collaborative anomalies of public opinion heat, passenger flow fluctuation and equipment load; secondly, there is a lack of dynamic adaptive computing power scheduling mechanism, which does not perform regional segmentation and elastic allocation of computing power based on the real-time changes of operation data, making it difficult to cope with the instantaneous data processing pressure brought by passenger flow tide; thirdly, the abnormal response is limited to fault disposal after the event, without building a prediction model combining passenger flow evolution trend and abnormal common characteristics, which cannot predict the potential risks of congestion diffusion and fault chain reaction in advance; fourthly, the warning output dimension is single, without realizing the visualization mapping and multi-end synchronous pushing of abnormal distribution, and also lacks the quantitative evaluation of operation loss, making it difficult to support the precise decision of the operation party and the travel guidance of passengers. SUMMARY

[0004] The technical problem solved by this invention is that existing technologies are unable to integrate passenger flow, equipment and public opinion data, cannot dynamically configure computing power to adapt to tidal passenger flow, lack the ability to predict the spread of anomalies, and the early warning lacks economic assessment and multi-terminal push, making it difficult to support the refined operation decision-making of the entire rail transit system.

[0005] To address the aforementioned technical problems, this invention provides the following technical solution: a cloud computing-based rail transit data analysis method, comprising the following steps:

[0006] Step S1: Collect full-domain operation data of the target rail transit network, and construct a dynamic panoramic topology map based on the full-domain operation data;

[0007] Step S2: Divide the dynamic panoramic topology map into N traffic area sub-maps;

[0008] Step S3: Configure elastic cloud computing nodes mapped to the N traffic area sub-maps, record the service traffic area sub-map range, computing power limit and data receiving port of each elastic cloud computing node, and monitor passenger flow changes and public opinion heat of each traffic area sub-map in real time.

[0009] Step S4: Distribute the full-domain operation data to the corresponding elastic cloud computing nodes. Combine the cloud computing processing model and use the data clustering algorithm to perform initial screening and cloud association transformation of the full-domain operation data to determine the distribution of abnormal data.

[0010] Step S5: After the cloud platform detects a critical event, it aggregates the abnormal data from each elastic cloud computing node, mines the commonalities of the abnormal data, and determines the predicted distribution of abnormal data by combining it with the real-time passenger flow evolution trend, and obtains an abnormal distribution map.

[0011] Step S6: Map the anomaly distribution map in real time onto the dynamic panoramic topology map to generate global early warning information.

[0012] Preferably, step S1 includes the following sub-steps:

[0013] Step S101: Real-time collection of full-domain operation data of the target rail transit network, including passenger flow data, train operation status data, equipment operating status data, and social media sentiment data;

[0014] Step S102: Based on the global operation data, construct an initial panoramic topology map of the rail transit network;

[0015] Step S103: Based on the real-time stream of the global operation data, dynamically adjust the initial panoramic topology map to obtain a dynamic panoramic topology map.

[0016] Preferably, step S2 includes the following sub-steps:

[0017] Step S201, under the constraint of a preset principle tolerance, combining real-time global operation data, a dynamic panoramic topology graph is preliminarily segmented to obtain N initial traffic area subgraphs;

[0018] The preset principle tolerance includes a passenger flow peak balance principle and a weak dependence principle.

[0019] The passenger flow peak balance principle includes:

[0020] The initial traffic area subgraph passenger flow load fluctuation threshold is within a first preset range.

[0021] The weight of the passenger flow peak balance principle is adjusted to a first percentage during a passenger flow peak or public opinion early warning period.

[0022] The weak dependence principle includes that the data interaction frequency between the initial traffic area subgraphs is within a second preset range, and the initial traffic area subgraph boundary division preferentially maintains the integrity of the same line continuous station.

[0023] Step S202, according to real-time passenger flow changes and dynamic panoramic topology graph update conditions, the coverage range of the initial traffic area subgraph is adaptively optimized to obtain a traffic area subgraph.

[0024] Preferably, the logic of adaptively optimizing the coverage range of the initial traffic area subgraph includes:

[0025] Real-time monitoring of the passenger flow density of each initial traffic area subgraph, if the passenger flow of one of the initial traffic area subgraphs is greater than a preset load threshold, the coverage range of the one of the initial traffic area subgraphs is reduced and the surrounding area is divided into adjacent initial traffic area subgraphs.

[0026] If the load difference between the initial traffic area subgraphs is greater than a second percentage due to passenger flow tides, the initial traffic area subgraph boundary is adjusted.

[0027] If the public opinion heat of one of the initial traffic area subgraphs is greater than a first preset warning line, the coverage range of the one of the initial traffic area subgraphs is preferentially ensured to completely match the public opinion involved station.

[0028] Preferably, the step S3 includes the following sub-steps:

[0029] Step S301, evaluating the business carrying demand of each traffic area subgraph, specifically including:

[0030] Combining the real-time passenger flow load, data processing amount and public opinion correlation degree of the traffic area subgraph, the computing power benchmark value of the cloud computing node required by each traffic area subgraph is determined.

[0031] Step S302, according to the computing power reference value configuration corresponding elastic cloud computing node, the dynamic mapping relationship between the elastic cloud computing node and the traffic area subgraph is established, the service traffic area subgraph range, the upper limit of computing power and the data receiving port of each elastic cloud computing node are recorded, and stored to the elastic cloud computing node mapping library of the cloud platform;

[0032] Step S303, real-time monitoring of passenger flow changes and public opinion heat of each traffic area subgraph, specifically including:

[0033] If the traffic area subgraph load is greater than the third percentage of the computing power reference value, the corresponding elastic cloud computing node is expanded;

[0034] If the traffic area subgraph load is less than the fourth percentage of the computing power reference value, the elastic cloud computing node computing power is reduced.

[0035] Preferably, the step S4 includes the following sub-steps:

[0036] Step S401, through the cloud computing processing model, simultaneously reading the global operation data of multi-period rail transit, mining abnormal features, building a pre-check database for parallel access of multiple elastic cloud computing nodes, the pre-check database updating threshold according to seasonal passenger flow and public opinion hotspots, the abnormal features including passenger flow fluctuation threshold, equipment parameter threshold and negative emotion threshold of public opinion;

[0037] The cloud computing processing model includes a distributed preprocessing layer and a cloud conversion layer, the preprocessing layer integrates a data cleaning module, and the cloud conversion layer embeds a data clustering algorithm based on machine learning;

[0038] Step S402, distributing the global operation data to each elastic cloud computing node according to the traffic area subgraph corresponding relationship, calling the pre-check database from the preprocessing layer, traversing the passenger flow data, equipment working condition data and social media public opinion data respectively through multiple elastic cloud computing nodes in parallel, and screening out abnormal data of rail transit matched with abnormal features;

[0039] Step S403, the screened abnormal data is transmitted into the cloud conversion layer, classified according to the main data mode through the machine learning clustering algorithm, the intra-modal data correlation rules and cross-modal correlation rules are established, the abnormal data distribution is determined, and the main data mode includes passenger flow congestion, equipment failure and public opinion warning.

[0040] Preferably, the step S5 includes the following sub-steps:

[0041] Step S501, the cloud platform monitors the key events in the rail transit network in real time, and starts the data aggregation process as soon as a key event is detected. When starting the data aggregation process, the cloud platform will notify all the elastic cloud computing nodes corresponding to the traffic area subgraph, and let the elastic cloud computing nodes upload the abnormal data processed by themselves to the cloud platform. The elastic cloud computing nodes preferentially upload the abnormal data of the event-related area;

[0042] The key events include:

[0043] The third percentage of the real-time passenger flow of the station exceeding the preset passenger flow threshold, the train failure and stagnation, and the negative sentiment index of the rail transit-related public opinion on the social media being greater than the second preset warning line;

[0044] The event-related area includes:

[0045] When the train fails, the abnormal data of the station that the faulty train will pass through is preferentially uploaded;

[0046] When the public opinion is warned, the abnormal data of the key station mentioned in the public opinion is preferentially uploaded;

[0047] Step S502, after the cloud platform aggregates the abnormal data uploaded by each elastic cloud computing node, the correlation strength between the elastic cloud computing nodes is learned according to the machine learning algorithm, the common characteristics of the abnormal data are analyzed, and the correlation graph of the abnormal data and the key events is constructed;

[0048] The common characteristics include intra-modal common characteristics and cross-modal common characteristics;

[0049] The intra-modal common characteristics include the propagation path characteristics of passenger flow congestion and the chain reaction type of equipment failure;

[0050] The cross-modal common characteristics include the cooperative change law of the public opinion heat and the passenger flow congestion degree;

[0051] Step S503, according to the correlation graph, combining the real-time passenger flow evolution trend, the abnormal distribution is predicted, the predicted abnormal data is obtained, the real-time abnormal data and the predicted abnormal data are integrated, and the abnormal distribution graph is generated. The real-time passenger flow evolution trend includes the passenger flow direction and the passenger flow growth rate.

[0052] Preferably, the logic of predicting the abnormal distribution includes:

[0053] The passenger flow direction is fused with the spatial characteristics, specifically including:

[0054] The passenger flow direction is converted into a spatial adjacency weight matrix, a weight of a station pointed by a preset flow direction is a first spatial connection weight, a weight of a station pointed by a reverse direction of the preset flow direction is a second spatial connection weight, the first spatial connection weight and the second spatial connection weight are taken as spatial inputs of the prediction model, and the first spatial connection weight is greater than the second spatial connection weight;

[0055] The passenger flow growth rate is fused with time characteristics, specifically including:

[0056] The passenger flow growth rate is taken as a time sequence input of the prediction model, when the passenger flow growth rate is greater than a third preset range, it is predicted that the abnormality covers adjacent first number of stations in a first unit of time, and when the passenger flow growth rate is less than the third preset range, it is predicted that the abnormality covers adjacent second number of stations in a second unit of time, the first number is greater than the second number;

[0057] The spatial characteristics are station and line correlation characteristics, and the time characteristics are passenger flow evolution characteristics;

[0058] Common characteristics in the correlation relationship graph are introduced as constraint conditions of the prediction model, if it is an intra-modal common characteristic, the diffusion range is corrected according to the propagation path characteristics of the passenger flow congestion and the chain reaction type of the equipment failure, and if it is a cross-modal common characteristic, the prediction result is adjusted in combination with real-time public opinion data;

[0059] When the prediction model is predicted, the stations in the event correlation area are preferentially included in the preferential coverage range;

[0060] The stations and lines covered by the abnormality are predicted in stages according to a preset time period, and prediction abnormal data of each time period is generated.

[0061] Preferably, the step S6 includes the following sub-steps:

[0062] Step S601, a coordinate mapping relationship between the abnormal distribution graph and the dynamic panoramic topology graph is established, and real-time abnormal areas and prediction abnormal areas in the abnormal distribution graph are matched with corresponding station and line positions in the dynamic panoramic topology graph respectively;

[0063] The matching mode includes: adopting differential visual identification according to the abnormal type, marking passenger flow congestion with red, marking equipment failure with yellow, and marking public opinion early warning with orange;

[0064] Step S602, global early warning information is generated based on a mapping result of the coordinate mapping relationship between the abnormal distribution graph and the dynamic panoramic topology graph, the early warning information includes an abnormal type, an abnormal occurrence position, an abnormal influence range and an economic influence evaluation;

[0065] The economic influence evaluation includes an increase of per capita commuting cost caused by train delay and an estimation of surrounding business revenue loss caused by station congestion;

[0066] The abnormal influence range is divided into a first warning, a second warning and a third warning in descending order;

[0067] The first warning is a gateway associated anomaly, the second warning is a cross-line anomaly, and the third warning is a single station anomaly;

[0068] In step S603, the global warning information is synchronized to the cloud platform visual display interface, and is pushed to the rail transit operation management system, the station broadcasting system and the passenger service APP through an interface, and the warning information is refreshed in real time according to the update of the dynamic panoramic topological graph.

[0069] A rail transit data analysis system based on cloud computing includes a construction module, a segmentation module, a configuration module, a processing module, a monitoring module and a warning module.

[0070] The construction module is configured to collect real-time global operation data of a target rail transit network, construct an initial panoramic topological graph based on the global operation data, and dynamically adjust the initial panoramic topological graph according to real-time update data to obtain a dynamic panoramic topological graph.

[0071] The segmentation module is configured to segment the dynamic panoramic topological graph into N traffic area subgraphs under the constraint of a preset principle tolerance.

[0072] The configuration module is configured to configure elastic cloud computing nodes mapped to the N traffic area subgraphs.

[0073] The processing module is configured to distribute the global operation data to corresponding elastic cloud computing nodes, combine a cloud computing processing model, perform monitoring data initialization filtering and cloud association conversion through a data clustering algorithm, and determine an abnormal data distribution.

[0074] The monitoring module is configured to monitor key events in real time, aggregate abnormal data of each elastic cloud computing node, mine commonalities of the abnormal data, determine a predicted abnormal distribution in combination with a real-time passenger flow evolution trend, and obtain an abnormal distribution graph.

[0075] The warning module is configured to real-time map the abnormal distribution graph in the dynamic panoramic topological graph to generate global warning information.

[0076] The beneficial effects of the present application are: by integrating passenger flow, train operation state, equipment working condition and social media public opinion data, combining the space-time synchronization characteristics of the dynamic panoramic topological map, effectively solving the traditional multi-source data island problem, improving data integrity, improving cross-modal anomaly recognition rate, accurately capturing the chain correlation from the rise of public opinion heat to the intensification of passenger congestion to the overload of equipment load, and avoiding one-sidedness of single data analysis. At the same time, based on the elastic cloud computing node scheduling mechanism of the traffic area subgraph load, the computing power is dynamically adapted, the utilization rate of computing power resources is improved, the waste of invalid computing power is reduced, and the expansion response time of computing power in peak period is shortened, which not only avoids data processing delay in high-load areas, but also reduces energy consumption during off-peak periods, balancing efficiency and economy. The prediction algorithm using spatiotemporal graph neural network combined with common constraints upgrades the abnormal treatment from passive response to active prediction, improves the accuracy of abnormal prediction, and can predict the abnormal diffusion range and time in advance, shortens the emergency response time, and the correlation graph can intuitively present the linkage of events, abnormalities and affected areas, helping the operator to quickly locate the core problem. The multi-end collaborative early warning system constructed can synchronize the early warning information to the operation management system, station broadcast and passenger service APP, shorten the delay of multi-end synchronization, speed up the passengers to obtain abnormal information, improve passenger satisfaction, and the operator can also quantify the loss by means of economic impact evaluation, decision-making is more accurate, and provides core support for fine operation of large-scale rail transit network. BRIEF DESCRIPTION OF DRAWINGS

[0077] Figure 1 A step flow chart of a rail transit data analysis method based on cloud computing provided for an embodiment of the present application;

[0078] Figure 2 A basic flow diagram of a rail transit data analysis system based on cloud computing provided for an embodiment of the present application. DETAILED DESCRIPTION

[0079] In order to make the above-mentioned objects, features and advantages of the present application more obvious and easy to understand, the specific embodiments of the present application will be described in detail below with reference to the drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present application, rather than all embodiments.

[0080] Embodiment 1, refer to Figure 1 , provides a rail transit data analysis method based on cloud computing, characterized in that it comprises the following steps:

[0081] Step S1, collecting global operation data of the target rail transit network, and constructing a dynamic panoramic topological map based on the global operation data.

[0082] Step S2, dividing the dynamic panoramic topological map into N traffic area subgraphs.

[0083] Step S3, configure elastic cloud computing nodes mapped to N traffic area subgraphs, record the service traffic area subgraph range, computing power upper limit and data receiving port of each elastic cloud computing node, and monitor the passenger flow changes and public opinion heat of each traffic area subgraph in real time.

[0084] Step S4, distribute the global running data to the corresponding elastic cloud computing nodes, combine the cloud computing processing model, and perform global running data initialization filtering and cloud association conversion through a data clustering algorithm to determine the abnormal data distribution.

[0085] Step S5, after the cloud platform detects a key event, the abnormal data of each elastic cloud computing node is summarized, the commonality of the abnormal data is mined, the predicted abnormal data distribution is determined in combination with the real-time passenger flow evolution trend, and the abnormal distribution map is obtained.

[0086] Step S6, real-time mapping of the abnormal distribution map in the dynamic panoramic topology map to generate global early warning information.

[0087] Step S1 includes the following sub-steps:

[0088] Step S101, real-time collection of global running data of the target rail transit network, the global running data including passenger flow data, train running state data, equipment working condition data and social media public opinion data, realizing multi-source data second-level collection and unified convergence, improving data integrity and avoiding information gap problems of traditional single data collection.

[0089] Step S102, based on the global running data, an initial panoramic topology map of the rail transit network is constructed. The initial topology map can intuitively present the line correlation, providing a precise spatial basis for subsequent regional segmentation, and the topology construction efficiency is improved compared with traditional CAD drawing. The initial panoramic topology map is constructed by using the graph structure modeling tool NetworkX, setting the station as the node and the line as the edge, and marking the node attributes and edge attributes. The node attributes include the station type and the transfer level, and the edge attributes include the line length and the design speed.

[0090] The calculation formula of the node weight is:

[0091] W n =C t ×C d ;

[0092] Wherein, W n is the node weight, C t is the transfer level coefficient, and C d为 is the daily passenger flow coefficient.

[0093] The calculation formula of the edge weight is:

[0094] W e =(L / V d )×F;

[0095] wherein, W e is the edge weight, L represents the line length, V d represents the design speed, and F is the train passing frequency.

[0096] In step S103, the initial panoramic topology graph is dynamically adjusted according to the real-time flow of the global operation data, and a dynamic panoramic topology graph is obtained. The dynamic panoramic topology graph is marked with a dynamic directed mark based on real-time train scheduling and passenger flow direction. The directed mark is dynamically updated according to the scheduling instruction, train running speed and passenger flow boarding and alighting.

[0097] The real-time updating rules of the dynamic directed mark include:

[0098] According to the scheduling instruction trigger, if a temporary speed limit instruction of the train is received, the color of the directed mark of the corresponding line changes from green to yellow;

[0099] The directed mark is linked with the train speed. When the train speed is less than 20 km / h, it is an inbound station or a fault, and the line width increases. When the train speed is greater than 80 km / h, it is normal driving, and the line width returns to the default;

[0100] If the number of passengers boarding the platform is greater than the number of passengers alighting, it indicates that the passenger flow is concentrated, and the station node flashes red. Otherwise, it flashes green.

[0101] The operation personnel can dynamically identify the line through the initial panoramic topology Figure 1 The dynamic information acquisition efficiency of the initial panoramic topology is improved compared with the traditional static topology graph, and it provides a time and space synchronized topology basis for subsequent initial traffic area subgraph segmentation and abnormal mapping.

[0102] Step S2 includes the following substeps:

[0103] In step S201, under the constraint of a preset principle tolerance, the dynamic panoramic topology graph is preliminarily segmented in combination with real-time global operation data, and N initial traffic area subgraphs are obtained.

[0104] The preset principle tolerance includes a passenger flow peak balance principle and a weak dependence principle.

[0105] The passenger flow peak balance principle includes:

[0106] The initial traffic area subgraph passenger flow load fluctuation threshold is within a first preset range.

[0107] The weight of the passenger flow peak balance principle is improved by a first percentage during a passenger flow peak or public opinion early warning period, so as to ensure that the processing pressure of each initial traffic area subgraph tends to be balanced during a high load period.

[0108] The calculation formula of the initial traffic area subgraph load balance degree is:

[0109] L0=1-(F 0max -F 0min ) / F0;

[0110] wherein, L0 is an initial traffic area subgraph load balancing degree, F 0max is an initial traffic area subgraph maximum passenger flow load, F 0min is an initial traffic area subgraph minimum passenger flow load, and F0 is an initial traffic area subgraph average passenger flow load.

[0111] The weak dependence principle includes that the data interaction frequency between the initial traffic area subgraphs is within a second preset range, and the initial traffic area subgraph boundary division preferentially maintains the integrity of the same line continuous station.

[0112] Step S202, according to the real-time passenger flow change and the update of the dynamic panoramic topology graph, adaptively optimizing the coverage range of the initial traffic area subgraph, and obtaining a traffic area subgraph.

[0113] The logic of adaptively optimizing the coverage range of the initial traffic area subgraph includes:

[0114] Real-time monitoring of the passenger flow density of each initial traffic area subgraph, if the passenger flow of one of the initial traffic area subgraphs is greater than a preset load threshold, then the coverage range of one of the initial traffic area subgraphs is reduced and the surrounding area is divided to the adjacent initial traffic area subgraph.

[0115] If the load difference between the initial traffic area subgraphs is greater than a second percentage due to passenger flow tide, the initial traffic area subgraph boundary is adjusted.

[0116] If the public opinion heat associated with one of the initial traffic area subgraphs is greater than a preset warning line, the coverage range of one of the initial traffic area subgraphs is preferentially ensured to completely match the station involved in the public opinion.

[0117] After optimization, the load balancing degree of the initial traffic area subgraph is improved, the response speed of the public opinion associated anomaly is improved, and the problem that the traditional static segmentation cannot adapt to passenger flow tide and public opinion burst is solved.

[0118] Step S3 includes the following sub-steps:

[0119] Step S301, evaluating the business carrying demand of each traffic area subgraph, specifically including:

[0120] Combining the real-time passenger flow load, data processing amount and public opinion correlation of the traffic area subgraph, determining the computing power benchmark value of the cloud computing node required by each traffic area subgraph.

[0121] The computing power benchmark value is calculated by using a multiple linear regression model.

[0122] The calculation formula of the computing power benchmark value is:

[0123] C = a x + b x + g x ;

[0124] Wherein, C is the computing power benchmark value, is the passenger flow data processing amount, a is the passenger flow data weight, is the device data processing amount, b is the device data weight, is the public opinion data processing amount, g is the public opinion data weight.

[0125] The computing power evaluation error is small, and the precision is significantly improved compared with the traditional fixed computing power configuration.

[0126] Step S302, according to the computing power benchmark value, the corresponding elastic cloud computing node is configured, the dynamic mapping relationship between the elastic cloud computing node and the traffic area subgraph is established, the service traffic area subgraph range, the computing power upper limit and the data receiving port of each elastic cloud computing node are recorded, and stored to the elastic cloud computing node mapping library of the cloud platform.

[0127] The Kubernetes container orchestration technology is used to configure the elastic cloud computing node, and the Redis cache database is used to store the node information of the elastic cloud computing node mapping library. The time consumption of the elastic cloud computing node configuration is shortened from hours to minutes, the response speed of the elastic cloud computing node mapping library query is improved to milliseconds, and the subsequent data is quickly distributed.

[0128] Step S303, real-time monitoring of passenger flow changes and public opinion heat of each traffic area subgraph, specifically including:

[0129] If the traffic area subgraph load is greater than the third percentage of the computing power benchmark value, the corresponding elastic cloud computing node is expanded.

[0130] If the traffic area subgraph load is less than the fourth percentage of the computing power benchmark value, the elastic cloud computing node computing power is reduced.

[0131] Step S4 includes the following sub-steps:

[0132] Step S401, through the cloud computing processing model, simultaneously reading the global operation data of multi-period rail transit, using the isolated forest algorithm to mine the abnormal features in the historical data, building a pre-check database for parallel access of multiple elastic cloud computing nodes, and updating the threshold value of the pre-check database with seasonal passenger flow and public opinion hotspots, the abnormal features include passenger flow fluctuation threshold, device parameter threshold and negative emotion threshold of public opinion.

[0133] The cloud computing processing model includes a distributed preprocessing layer and a cloud conversion layer, the preprocessing layer integrates a data cleaning module, and the cloud conversion layer embeds a data clustering algorithm based on machine learning.

[0134] Step S402, distribute the global operation data to each elastic cloud computing node according to the traffic area subgraph correspondence relationship, call the pre-check database from the preprocessing layer, and perform feature correction on passenger flow data, equipment working condition data and social media public opinion data respectively through parallel processing of multiple elastic cloud computing nodes to screen out abnormal data of rail transit matching abnormal features.

[0135] Step S403, the screened abnormal data is transmitted into the cloud conversion layer, and the machine learning clustering algorithm is used for classification and parallel processing according to the main data mode, the data correlation conversion efficiency is improved, the intra-modal data correlation rules and cross-modal correlation rules are established, and the abnormal data distribution is determined, the main data mode includes passenger flow congestion, equipment failure and public opinion early warning.

[0136] Step S5 includes the following sub-steps:

[0137] Step S501, the cloud platform monitors the key events in the rail transit network in real time, and starts the data summary process as soon as a key event is detected. When starting the data summary process, the cloud platform will notify all elastic cloud computing nodes corresponding to the traffic area subgraph, so that the elastic cloud computing nodes upload the abnormal data processed by themselves to the cloud platform. The elastic cloud computing node preferentially uploads the abnormal data of the event correlation area. The cloud platform monitors the key events in real time, quickly focuses on the core area of the problem, and reduces the delay caused by irrelevant data transmission.

[0138] Key events include:

[0139] The third percentile of the real-time passenger flow of the station exceeds the preset passenger flow threshold, the train fails to stop, and the negative sentiment index of the rail transit related public opinion on the social media is greater than the second preset warning line.

[0140] Event correlation area includes:

[0141] When the train fails, the abnormal data of the station to be passed through by the subsequent train of the failed train is preferentially uploaded.

[0142] When the public opinion early warning, the abnormal data of the key station mentioned in the public opinion is preferentially uploaded.

[0143] Step S502, after the cloud platform summarizes the abnormal data uploaded by each elastic cloud computing node, the correlation strength between the elastic cloud computing nodes is analyzed according to the machine learning algorithm to analyze the common features of the abnormal data, and the correlation relationship diagram of the abnormal data and the key events is constructed.

[0144] The combination algorithm of graph neural network combined with spectral clustering and association rule learning is used as the machine learning algorithm, and the specific logic includes:

[0145] The elastic cloud computing nodes are treated as nodes in a graph neural network. The frequency of data interaction between nodes and the probability of abnormal data co-occurrence are regarded as the weights of the edges. The association strength between elastic cloud computing nodes is learned through the message passing mechanism of the graph neural network. The graph neural network model aggregates the features of neighboring nodes, dynamically updates the connection weights between nodes, and quantifies the abnormal propagation association of different traffic area subgraphs.

[0146] For the aggregated abnormal data, the spectral clustering algorithm is used to group the abnormal data with high similarity into one class.

[0147] Common features within a modality are clustered based on passenger flow density and propagation speed features to extract propagation path features that spread along transfer routes.

[0148] Cross-modal common features are obtained through feature fusion, specifically by mapping public opinion heat and passenger flow growth rate to the same feature space, and clustering to obtain the synergistic pattern that increased public opinion heat leads to increased passenger flow congestion.

[0149] The frequent pattern growth algorithm is used to mine the association rules between abnormal data and key events. Combined with the node association strength output by the graph neural network, an association graph is constructed with key events as the core, abnormal data as nodes, and association strength as edge weights, which intuitively presents the linkage logic between events, anomalies and affected areas.

[0150] Common features include intramodal commonalities and cross-modal commonalities.

[0151] Commonalities within a modality include the propagation path characteristics of passenger congestion and the cascading reaction types of equipment failures.

[0152] The commonalities across modalities include the coordinated changes in public opinion intensity and passenger congestion levels.

[0153] Step S503: Based on the correlation diagram and combined with the real-time passenger flow evolution trend, predict the abnormal distribution to obtain the predicted abnormal data. Integrate the real-time abnormal data and the predicted abnormal data to generate an abnormal distribution map. The real-time passenger flow evolution trend includes passenger flow direction and passenger flow growth rate.

[0154] The logic for predicting anomalous distributions includes:

[0155] An abnormal distribution is predicted using a spatiotemporal graph neural network and a multi-factor constrained regression prediction model. The specific prediction logic is as follows:

[0156] The system is based on a spatiotemporal graph neural network, integrating a time-series prediction module and a rule constraint layer. The spatiotemporal graph neural network is used to capture the spatial and temporal linkage characteristics of anomalies, the time-series prediction module handles the dynamic changes in passenger flow growth rate, and the rule constraint layer integrates prior knowledge of common features. The spatial linkage characteristics are the station and route association characteristics, and the temporal linkage characteristics are the passenger flow evolution characteristics.

[0157] The passenger flow direction is combined with the spatial characteristics, specifically including:

[0158] The passenger flow direction is converted into a spatial adjacency weight matrix, the weight of the station pointed by the preset flow direction is a first spatial connection weight, the weight of the station pointed by the opposite direction of the preset flow direction is a second spatial connection weight, the first spatial connection weight and the second spatial connection weight are taken as the spatial input of the space-time graph neural network, the first spatial connection weight is greater than the second spatial connection weight, guiding the prediction model to preferentially learn the propagation law of the anomaly along the flow direction, if the passenger flow of the fault station mainly flows to the east interchange station, the prediction model enhances the abnormal probability prediction weight of the east station.

[0159] The passenger flow growth rate is combined with the time characteristics, specifically including:

[0160] The passenger flow growth rate is taken as the time sequence input of the space-time graph neural network, when the passenger flow growth rate is greater than a third preset range, the time decay coefficient of the combined model is reduced, the time gradient calculation of the abnormal diffusion is accelerated, and it is predicted that the anomaly covers adjacent first number of stations in a first unit of time. When the passenger flow growth rate is less than the third preset range, the time decay coefficient of the combined model is increased, the abnormal diffusion speed is delayed, and it is predicted that the anomaly covers adjacent second number of stations in a second unit of time, the first number is greater than the second number.

[0161] If there is a modal intra-common feature of the propagation path feature of the passenger flow congestion in the association relationship graph, an additional weight is given to the connection between the stations of the propagation path in the spatial attention layer of the space-time graph neural network, forcing the prediction model to preferentially predict the abnormal diffusion along the propagation path.

[0162] If the interlocking reaction type of the equipment failure has a modal intra-common feature of the interlocking reaction type of the equipment failure, the third number of signal associated stations upstream and downstream of the fault point are marked as high-risk nodes, and the initial value of the abnormal probability is increased.

[0163] If there is a cross-modal commonness that the passenger flow congestion range expands by a fifth percentage every time the public opinion heat rises by a fourth percentage in the association relationship graph, the real-time public opinion heat is converted into a correction coefficient to magnify the prediction range output by the space-time graph neural network.

[0164] The event associated area is marked as a seed node, a higher initial abnormal probability is given in the initial feature vector of the space-time graph neural network, and the influence of the neighborhood of the seed node is preferentially calculated in the spatial propagation, ensuring that the prediction result is inclined to the core associated area.

[0165] Step S6 includes the following sub-steps:

[0166] Step S601, a coordinate mapping relationship between the abnormal distribution map and the dynamic panoramic topology map is established, and real-time abnormal areas and predicted abnormal areas in the abnormal distribution map are respectively matched with corresponding station and line positions in the dynamic panoramic topology map.

[0167] The matching mode includes: using differentiated visual identification according to the abnormal type, marking passenger flow congestion with red, marking equipment failure with yellow, and marking public opinion early warning with orange.

[0168] Step S602, global early warning information is generated based on the mapping result of the coordinate mapping relationship between the abnormal distribution map and the dynamic panoramic topology map, and the early warning information includes the abnormal type, the abnormal occurrence position, the abnormal influence range and the economic impact assessment.

[0169] The economic impact assessment includes the increase of per capita commuting cost caused by train delay and the estimation of surrounding business revenue loss caused by station congestion.

[0170] According to the abnormal influence range, the early warning levels are divided, and the early warning levels are divided into first-level early warning, second-level early warning and third-level early warning in descending order.

[0171] The first-level early warning is a network-related abnormality, the second-level early warning is a cross-line abnormality, and the third-level early warning is a single-station abnormality.

[0172] Step S603, the global early warning information is synchronized to the cloud platform visual display interface, supporting the operation personnel to check the abnormal details and evolution trend, and at the same time, the interface is pushed to the rail transit operation management system, the station broadcasting system and the passenger service APP, realizing the multi-end synchronization of the early warning information, and according to the update of the dynamic panoramic topology map, the early warning information is refreshed in real time, ensuring that the early warning information is consistent with the actual running state.

[0173] Embodiment 2, referring to Figure 2 , provides a rail transit data analysis system based on cloud computing, which comprises a construction module, a segmentation module, a configuration module, a processing module, a monitoring module and an early warning module.

[0174] The construction module is used for collecting global running data of a target rail transit network in real time, constructing an initial panoramic topology map based on the global running data, and dynamically adjusting the initial panoramic topology map according to real-time update data to obtain a dynamic panoramic topology map.

[0175] The segmentation module is used for segmenting the dynamic panoramic topology map into N traffic area subgraphs under the constraint of a preset principle tolerance.

[0176] The configuration module is used for configuring elastic cloud computing nodes mapped to the N traffic area subgraphs.

[0177] The processing module is configured to distribute the global operation data to corresponding elastic cloud computing nodes, and to determine abnormal data distribution by monitoring data initialization screening and cloud association conversion through a data clustering algorithm in combination with a cloud computing processing model.

[0178] The monitoring module is configured to monitor key events in real time, aggregate abnormal data of each elastic cloud computing node, mine commonalities of abnormal data, determine a predicted abnormal distribution in combination with a real-time passenger flow evolution trend, and obtain an abnormal distribution map.

[0179] The early warning module is configured to map the abnormal distribution map in a dynamic panoramic topology map in real time, and generate global early warning information.

[0180] The present application improves the overall efficiency of rail transit operation, data processing capability, abnormal response time and passenger experience. By integrating passenger flow, train operation status, equipment working condition and social media public opinion data, and combining the time and space synchronization characteristics of the dynamic panoramic topology map, the traditional multi-source data island problem is effectively solved, the data integrity is improved, the cross-modal abnormal recognition rate is improved, the chain correlation from the rise of public opinion heat to the intensification of passenger flow congestion to the overload of equipment is accurately captured, and the one-sidedness of single data analysis is avoided. At the same time, based on the elastic cloud computing node scheduling mechanism of the traffic area subgraph load, the computing power is dynamically adapted, the computing power resource utilization rate is improved, the invalid computing power waste is reduced, and the computing power expansion response time is shortened during peak hours, which avoids data processing delay in high-load areas and reduces energy consumption during off-peak hours, balancing efficiency and economy. The prediction algorithm using spatiotemporal graph neural network combined with common constraints upgrades the abnormal processing from passive response to active prediction, improves the accuracy of abnormal prediction, and can predict the abnormal diffusion range and time in advance, shortens the emergency response time, and the correlation graph can intuitively present the linkage of events, abnormalities and affected areas, helping the operator to quickly locate the core problem. The multi-end collaborative early warning system can synchronize the early warning information to the operation management system, the station broadcast and the passenger service APP, shorten the multi-end synchronization delay, speed up the passengers to obtain abnormal information, improve the passenger satisfaction, and the operator can also use the economic impact evaluation to quantify the loss, make more accurate decisions, and provide core support for fine operation of large-scale rail transit network.

[0181] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present application and not to limit it. Although the present application has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present application can be modified or replaced by equivalents without departing from the spirit and scope of the present application, which should be covered by the claims of the present application.

Claims

1. A cloud computing-based method for analyzing rail transit data, characterized in that, Includes the following steps: Step S1: Collect full-domain operation data of the target rail transit network, and construct a dynamic panoramic topology map based on the full-domain operation data; Step S2: Divide the dynamic panoramic topology map into N traffic area sub-maps; Based on real-time changes in passenger flow and updates to the dynamic panoramic topology map, the coverage of the initial traffic area sub-map is adaptively optimized to obtain the traffic area sub-map. Real-time monitoring of passenger flow density in each initial traffic area sub-map; if the passenger flow in one of the initial traffic area sub-maps exceeds a preset load threshold, the coverage area of ​​that initial traffic area sub-map is reduced and the surrounding area is assigned to an adjacent initial traffic area sub-map. If the load difference between the initial traffic area sub-maps is greater than the second percentage due to passenger flow tides, then the boundaries of the initial traffic area sub-maps will be adjusted. If the public opinion heat associated with one of the initial traffic area sub-maps is greater than the first preset warning line, then priority will be given to ensuring that the coverage of the one initial traffic area sub-map is completely matched with the sites involved in the public opinion. Step S3: Configure elastic cloud computing nodes mapped to the N traffic area sub-maps, record the service traffic area sub-map range, computing power limit and data receiving port of each elastic cloud computing node, and monitor passenger flow changes and public opinion heat of each traffic area sub-map in real time. Assess the service capacity requirements for each traffic area submap, specifically including: By combining real-time passenger flow load, data processing volume, and public opinion correlation of traffic area sub-maps, the baseline computing power value of cloud computing nodes required for each traffic area sub-map is determined. Real-time monitoring of passenger flow changes and public opinion intensity in each traffic area sub-map, specifically including: If the load of a traffic area submap exceeds the third percent of the computing power baseline, then the corresponding elastic cloud computing node will be expanded. If the load of the traffic area submap is less than 400% of the computing power baseline, then reduce the computing power of the elastic cloud computing nodes. Step S4: Distribute the full-domain operation data to the corresponding elastic cloud computing nodes. Combine the cloud computing processing model and use the data clustering algorithm to perform initial screening and cloud association transformation of the full-domain operation data to determine the distribution of abnormal data. Step S5: After the cloud platform detects a critical event, it aggregates the abnormal data from each elastic cloud computing node, mines the commonalities of the abnormal data, and determines the predicted distribution of abnormal data by combining it with the real-time passenger flow evolution trend, and obtains an abnormal distribution map. Step S6: Map the anomaly distribution map in real time onto the dynamic panoramic topology map to generate global early warning information.

2. The cloud computing-based rail transit data analysis method as described in claim 1, characterized in that, Step S1 includes the following sub-steps: Step S101: Real-time collection of full-domain operation data of the target rail transit network, including passenger flow data, train operation status data, equipment operating status data, and social media sentiment data; Step S102: Based on the global operation data, construct an initial panoramic topology map of the rail transit network; Step S103: Based on the real-time stream of the global operation data, dynamically adjust the initial panoramic topology map to obtain a dynamic panoramic topology map.

3. The cloud computing-based rail transit data analysis method as described in claim 2, characterized in that, Step S2 includes the following sub-steps: Step S201: Under the constraint of the preset principle tolerance, and combined with the real-time full-domain operation data, the dynamic panoramic topology map is initially segmented to obtain N initial traffic area sub-maps. The preset principle tolerance includes the passenger flow peak balance principle and the weak dependence principle; The principle of balancing peak passenger flow includes: The initial traffic area submap passenger flow load fluctuation threshold is within the first preset range; During peak passenger flow periods or periods of public opinion warning, the weight of the principle of balancing peak passenger flow is adjusted to the first percentile. The weak dependency principle includes: the data interaction frequency between initial traffic area sub-maps is within a second preset range, and the boundary division of the initial traffic area sub-maps prioritizes maintaining the integrity of continuous stations on the same line.

4. The cloud computing-based rail transit data analysis method as described in claim 3, characterized in that, Step S3 includes the following sub-steps: Configure corresponding elastic cloud computing nodes according to the computing power benchmark value, establish a dynamic mapping relationship between elastic cloud computing nodes and traffic area sub-maps, record the service traffic area sub-map range, computing power limit and data receiving port of each elastic cloud computing node, and store them in the elastic cloud computing node mapping library of the cloud platform.

5. The cloud computing-based rail transit data analysis method as described in claim 4, characterized in that, Step S4 includes the following sub-steps: Step S401: Using a cloud computing processing model, simultaneously read the full-domain operation data of rail transit across multiple time periods, mine abnormal features, and build a pre-inspection database that can be accessed in parallel by multiple elastic cloud computing nodes. The pre-inspection database updates thresholds based on seasonal passenger flow and public opinion hotspots. The abnormal features include passenger flow fluctuation thresholds, equipment parameter thresholds, and negative public opinion thresholds. The cloud computing processing model includes a distributed preprocessing layer and a cloud conversion layer. The preprocessing layer integrates a data cleaning module, and the cloud conversion layer embeds a data clustering algorithm based on machine learning. Step S402: Distribute the overall operation data to each elastic cloud computing node according to the correspondence of traffic area sub-maps. Call the pre-inspection database from the preprocessing layer. Through multiple elastic cloud computing nodes in parallel, perform feature verification on passenger flow data, equipment condition data and social media sentiment data respectively, and filter out abnormal data of rail transit that match abnormal features. Step S403: The filtered abnormal data is transmitted to the cloud conversion layer, and classified according to the master data modality by the machine learning clustering algorithm. Intramodal data association rules and cross-modal association rules are established to determine the distribution of abnormal data. The master data modality includes passenger flow congestion, equipment failure and public opinion warning.

6. The cloud computing-based rail transit data analysis method as described in claim 5, characterized in that, Step S5 includes the following sub-steps: Step S501: The cloud platform monitors key events in the rail transit network in real time. Once a key event is detected, the data aggregation process is initiated. When the data aggregation process is initiated, the cloud platform will notify all elastic cloud computing nodes of the corresponding traffic area sub-map, so that the elastic cloud computing nodes can upload the abnormal data they have processed to the cloud platform. The elastic cloud computing nodes will prioritize uploading abnormal data of the event-related area. The key events include: The following conditions must be met: the real-time passenger flow at the station exceeds the third percent of the preset passenger flow threshold; the train malfunctions and stops; and the negative sentiment index related to rail transit on social media exceeds the second preset warning line. The event-related area includes: When a train malfunctions, priority is given to uploading abnormal data from stations the malfunctioning train will subsequently pass through; When issuing a public opinion warning, prioritize uploading abnormal data from key websites mentioned in the public opinion. Step S502: After the cloud platform aggregates the abnormal data uploaded by each elastic cloud computing node, it learns the correlation strength between elastic cloud computing nodes based on machine learning algorithms, analyzes the common characteristics of abnormal data, and constructs a correlation graph between abnormal data and key events. The common features include intramodal common features and cross-modal common features; The common features within the modality include the propagation path characteristics of passenger flow congestion and the chain reaction types of equipment failure; The common characteristics across modalities include the coordinated change patterns of public opinion intensity and passenger flow congestion. Step S503: Based on the correlation diagram and combined with the real-time passenger flow evolution trend, predict the abnormal distribution to obtain predicted abnormal data. Integrate the real-time abnormal data and the predicted abnormal data to generate an abnormal distribution map. The real-time passenger flow evolution trend includes passenger flow direction and passenger flow growth rate.

7. The cloud computing-based rail transit data analysis method as described in claim 6, characterized in that, The logic for predicting abnormal distributions includes: Integrating passenger flow patterns with spatial characteristics, specifically including: The passenger flow direction is transformed into a spatial adjacency weight matrix. The weight of the station to which the flow direction points is the first spatial connection weight, and the weight of the station in the opposite direction to the flow direction is the second spatial connection weight. The first spatial connection weight and the second spatial connection weight are used as the spatial input of the prediction model, and the first spatial connection weight is greater than the second spatial connection weight. Integrating passenger flow growth rate with time characteristics, specifically including: The passenger flow growth rate is used as the time-series input of the prediction model. When the passenger flow growth rate is greater than the third preset range, it is predicted that the first number of adjacent stations will be abnormally covered within the first unit time. If the passenger flow growth rate is less than the third preset range, it is predicted that the second number of adjacent stations will be abnormally covered within the second unit time. The first number is greater than the second number. The spatial features are the station and route association features, and the temporal features are the passenger flow evolution features; Common features in the relationship graph are introduced as constraints for the prediction model. If they are common features within a mode, the diffusion range is corrected according to the propagation path characteristics of passenger flow congestion and the chain reaction type of equipment failure. If they are common features across modes, the prediction results are adjusted by combining real-time public opinion data. When making predictions using predictive models, sites in areas associated with the event should be prioritized for coverage. The system predicts the sites and lines with abnormal coverage in stages according to preset time periods, and generates predicted abnormal data for each time period.

8. The cloud computing-based rail transit data analysis method as described in claim 7, characterized in that, Step S6 includes the following sub-steps: Step S601: Establish the coordinate mapping relationship between the anomaly distribution map and the dynamic panoramic topology map, and match the real-time anomaly area and the predicted anomaly area in the anomaly distribution map with the corresponding station and line locations in the dynamic panoramic topology map. The matching method includes: using differentiated visual identifiers according to the type of anomaly, using red to mark passenger congestion, yellow to mark equipment failure, and orange to mark public opinion warnings; Step S602: Generate global early warning information based on the mapping result of the coordinate mapping relationship between the anomaly distribution map and the dynamic panoramic topology map. The early warning information includes anomaly type, anomaly location, anomaly impact range, and economic impact assessment. The economic impact assessment includes estimates of increased per capita commuting costs due to train delays and lost revenue for surrounding businesses caused by station congestion. The warning levels are divided according to the scope of the abnormal impact, and are classified into Level 1, Level 2 and Level 3 warnings in descending order; The first-level warning is a network-wide related anomaly, the second-level warning is a cross-line anomaly, and the third-level warning is a single-site anomaly. Step S603: The global early warning information is synchronized to the cloud platform's visual display interface, and simultaneously pushed to the rail transit operation management system, station broadcasting system, and passenger service APP through the interface. The early warning information is also refreshed in real time according to the update of the dynamic panoramic topology map.

9. A cloud-based rail transit data analysis system, applied in any one of the cloud-based rail transit data analysis methods as described in claims 1-8, characterized in that, It includes a building module, a segmentation module, a configuration module, a processing module, a monitoring module, and an early warning module; The construction module is used to collect real-time full-domain operation data of the target rail transit network, construct an initial panoramic topology map based on the full-domain operation data, and dynamically adjust the initial panoramic topology map according to the real-time updated data to obtain a dynamic panoramic topology map. The segmentation module is used to segment the dynamic panoramic topology map into N traffic area sub-maps under the constraints of a preset principle tolerance. The configuration module is used to configure elastic cloud computing nodes mapped to the N traffic area sub-maps; The processing module is used to distribute the full-domain operation data to the corresponding elastic cloud computing nodes, and, in conjunction with the cloud computing processing model, use data clustering algorithms to perform monitoring data initialization screening and cloud association transformation to determine the distribution of abnormal data. The monitoring module is used to monitor key events in real time, summarize abnormal data from each elastic cloud computing node, mine commonalities in abnormal data, determine the predicted abnormal distribution by combining real-time passenger flow evolution trends, and obtain an abnormal distribution map. The early warning module is used to map the abnormal distribution map onto the dynamic panoramic topology map in real time and generate global early warning information.

Images