Traffic hub passenger flow prediction method, system and device based on big data

Abstract

This invention discloses a method, system, and device for predicting passenger flow at transportation hubs based on big data, belonging to the field of data processing technology. The method includes: determining the traffic transfer type of a first transportation hub; constructing a passenger flow prediction model for each traffic transfer type using single-traffic-type flow-direction prediction and multi-traffic-type interactive prediction, wherein the passenger flow prediction model is embedded in the traffic management system and includes a three-layer fully connected first prediction layer, a second prediction layer, and an integration output layer; receiving prediction tasks, combining them with real-time passenger flow monitoring data, and transferring them to the passenger flow prediction model for hierarchical cascading directional triggering and decision-making, determining the passenger flow prediction result, converting it into a passenger flow guidance map, and performing visualization and congestion early warning management on terminal equipment. This invention solves the technical problems of low accuracy in passenger flow prediction at transportation hubs and difficulty in handling multi-traffic-type interactive scenarios in existing technologies, achieving the technical effect of improving passenger flow prediction accuracy and system response efficiency.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and specifically to a method, system, and equipment for predicting passenger flow in transportation hubs based on big data. Background Technology

[0002] With the accelerating pace of urbanization, urban transportation hubs (such as high-speed rail stations, subway stations, airports, and passenger transport centers) play a crucial role in transferring and distributing traffic between various modes of transportation. The surge in passenger flow and the diversification of transportation types place higher demands on the real-time monitoring, scheduling, and early warning capabilities of traffic management systems. Most existing passenger flow forecasting methods rely on historical data from a single transportation mode for modeling, lacking the ability to deeply model the interactions between multiple transportation types. This results in low accuracy of forecasts in complex transportation hubs, failing to meet the demands for high timeliness and precision in practical applications. Furthermore, current systems generally lack response mechanisms for real-time dynamic data streams, making it difficult to achieve visualized applications of forecast results and coordinated control with congestion warnings. Summary of the Invention

[0003] This application provides a method, system, and equipment for predicting passenger flow in transportation hubs based on big data, which solves the technical problems of low accuracy in predicting passenger flow in transportation hubs and difficulty in dealing with interactive scenarios of multiple transportation types in the prior art.

[0004] The first aspect of this application provides a method for predicting passenger flow at transportation hubs based on big data, the method comprising:

[0005] The system determines the traffic transfer types of the first transportation hub, wherein the traffic transfer types include at least two; it constructs a passenger flow prediction model for the traffic transfer types using flow-direction prediction for a single traffic type and interactive prediction for multiple traffic types, wherein the passenger flow prediction model is embedded in the traffic management system and includes a three-layer fully connected first prediction layer, a second prediction layer, and an integrated output layer; it receives prediction tasks, combines them with real-time passenger flow monitoring data, and transfers them to the passenger flow prediction model for hierarchical cascading directional triggering and decision-making, determines the passenger flow prediction results, converts them into a passenger flow guidance map, and performs terminal equipment visualization and congestion early warning management.

[0006] A second aspect of this application provides a big data-based passenger flow prediction system for transportation hubs, the system comprising:

[0007] Traffic type determination module: determines the traffic transfer type of the first transportation hub, wherein the traffic transfer type includes at least two; Model building module: constructs a passenger flow prediction model for the traffic transfer type using single traffic type flow-direction prediction and multi-traffic type interactive prediction, wherein the passenger flow prediction model is embedded in the traffic management system and includes a three-layer fully connected first prediction layer, second prediction layer and integrated output layer; Management module: receives prediction tasks, combines them with real-time passenger flow monitoring data, transfers them to the passenger flow prediction model for hierarchical cascading directional triggering and decision-making, determines the passenger flow prediction results, converts them into a passenger flow guidance map, and performs terminal equipment visualization and congestion early warning management.

[0008] A third aspect of this application provides an electronic device, comprising: a memory for storing executable instructions; and a processor for executing the executable instructions stored in the memory to implement the big data-based transportation hub passenger flow prediction method provided in this application.

[0009] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0010] First, the traffic transfer types of the first transportation hub are determined, with at least two types included. Then, based on single-traffic-type flow-direction prediction and multi-traffic-type interactive prediction, a passenger flow prediction model is constructed for each traffic transfer type. This model is embedded in the traffic management system and includes a three-layer fully connected first prediction layer, a second prediction layer, and an integrated output layer. Finally, prediction tasks are received, combined with real-time passenger flow monitoring data, and fed into the passenger flow prediction model for hierarchical cascading targeted triggering and decision-making. The passenger flow prediction results are then converted into a passenger flow guidance map for visualization on terminal devices and congestion warning management. This approach solves the technical problems of low passenger flow prediction accuracy and difficulty in handling multi-traffic-type interactive scenarios in existing technologies, achieving the technical effect of improving passenger flow prediction accuracy and system response efficiency. Attached Figure Description

[0011] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0012] Figure 1 A schematic diagram of the big data-based passenger flow prediction method for transportation hubs provided in this application embodiment;

[0013] Figure 2A schematic diagram of the structure of a big data-based transportation hub passenger flow prediction system provided in this application embodiment;

[0014] Figure 3 This is a schematic diagram of the structure of an exemplary electronic device of this application.

[0015] Explanation of reference numerals in the attached diagram: Traffic type determination module 11, model building module 12, management module 13, processor 21, memory 22, input device 23, output device 24. Detailed Implementation

[0016] This application provides a method, system, and equipment for predicting passenger flow in transportation hubs based on big data, which solves the technical problems of low accuracy in passenger flow prediction and difficulty in dealing with interactive scenarios of multiple transportation types in the prior art.

[0017] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0018] It should be noted that the terms "comprising" and "having" are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or server that includes a series of steps or units is not necessarily limited to those steps or units that are explicitly listed, but may include other steps or modules that are not explicitly listed or that are inherent to these processes, methods, products, or devices.

[0019] Example 1, as Figure 1 As shown in the embodiments of this application, a method for predicting passenger flow in transportation hubs based on big data is provided, wherein the method includes:

[0020] The traffic transfer type of the first transportation hub is determined, wherein the traffic transfer type includes at least two.

[0021] Transportation transfer types refer to the transfer paths or patterns between different modes of transportation within the same transportation hub. Transportation transfer types include at least two or more combinations of transportation modes, such as: transfer between subway and high-speed rail (subway-high-speed rail), transfer between bus and subway (bus-subway), transfer between taxi and high-speed rail (taxi-high-speed rail), and transfer between walking shuttle and subway (walking-subway), etc.

[0022] Optionally, by accessing the infrastructure data, station layout information, and historical traffic records of the transportation hub, the set of transportation modes covered by the hub can be obtained. Then, based on the existing transportation connections, a transfer map is constructed. Using the entry and exit records of transportation vehicles, passenger flow transfer paths, and timestamps as a basis, combinations of transportation modes that exhibit frequent transfer behavior within a certain time window can be identified and confirmed as valid transfer types.

[0023] For example, within a high-speed rail station, if a high-frequency passenger flow migration path is detected at the subway entrance points near the high-speed rail exit, and the subway entry time is highly correlated with the high-speed rail exit time, then "high-speed rail-subway" can be determined as a type of transportation transfer. Similarly, if there is a stable and continuous passenger flow conversion relationship between bus stops and subway station entrances / exits, it can also be confirmed as a "bus-subway" transportation transfer type.

[0024] Based on flow-direction prediction for a single traffic type and interactive prediction for multiple traffic types, a passenger flow prediction model is constructed for the traffic transfer type. The passenger flow prediction model is embedded in the traffic management system and includes a three-layer fully connected first prediction layer, a second prediction layer, and an integrated output layer.

[0025] In this embodiment, based on the determined traffic transfer types, a passenger flow prediction model suitable for transportation hub scenarios is constructed by combining flow-direction prediction for a single traffic type with interactive prediction for multiple traffic types. The passenger flow prediction model fully utilizes the passenger flow evolution patterns and interactive coupling characteristics among various transportation modes, employing deep learning modeling methods to achieve fusion analysis and accurate prediction of multi-source real-time data. Specifically, the passenger flow prediction model has a multi-layered fully connected structure, including a first prediction layer, a second prediction layer, and an integrated output layer. It is deployed as a core prediction unit within the traffic management system, operating in an embedded manner to support real-time decision-making and management of transportation hubs. The first prediction layer constructs an independent time-series prediction channel for each mode of transportation. Using time series data as input, it combines flow and direction data features to generate preliminary prediction results for a single transportation type. The second prediction layer further introduces spatial location relationships, transfer behavior patterns, and historical interaction path data among multiple modes of transportation to explore their potential correlations. Through deep neural networks, it models the interaction effects between multiple types to form an interaction-enhanced prediction output. The integration output layer receives and merges the aforementioned prediction results, and based on the spatial topology and traffic rules within the transportation hub, it completes the reconstruction and optimization of the prediction data, outputting the final passenger flow prediction results.

[0026] Furthermore, the construction of the first prediction layer includes:

[0027] A first traffic type is determined, wherein the first traffic type is any one of the traffic transfer types; for the first traffic type, a first prediction channel is constructed using time series as the first judgment node and flow-direction as the second prediction node; the traffic transfer types are traversed until the construction of the Nth prediction channel is completed, and the first prediction channel is constructed in parallel until the Nth prediction channel is constructed, thereby generating the first prediction layer.

[0028] The first prediction layer establishes an independent prediction channel for each mode of transportation within the transportation transfer type. Specifically, it first determines the first transportation type, which is any mode of transportation in the identified transportation transfer type set, such as subway, high-speed rail, bus, or taxi. Then, based on the passenger flow data of this mode of transportation over a historical period, it constructs its corresponding time series input. The time series serves as the first judgment node, reflecting the timeliness and periodicity of passenger flow evolution. Simultaneously, it extracts the flow changes and direction distribution under this mode of transportation as the second prediction node, used to capture spatial behavior patterns and transfer tendencies. Based on this, a first prediction channel corresponding to this mode of transportation is established. By traversing all modes of transportation in the transportation transfer type set, corresponding prediction channels are constructed sequentially until the prediction channel for the Nth mode of transportation is completed. Finally, the N prediction channels are structurally combined in parallel to form the first prediction layer, enabling parallel extraction and prediction processing of passenger flow characteristics for multiple modes of transportation within the same time frame.

[0029] The system receives prediction tasks, combines them with real-time passenger flow monitoring data, and transfers them to the passenger flow prediction model for hierarchical cascading targeted triggering and decision-making. It then determines the passenger flow prediction results, converts them into a passenger flow guidance map, and performs visualization and congestion early warning management on terminal devices.

[0030] When the traffic management system receives a passenger flow forecasting task, it will automatically trigger the embedded passenger flow forecasting model to execute the forecasting process. After receiving the task, the system first retrieves real-time passenger flow monitoring data related to the task, including but not limited to subway / high-speed rail station entry and exit traffic, bus card swipe records, infrared / video recognition of the number of people, mobile terminal trajectories, etc. After preprocessing, this data is used as input features for the model and transferred to the passenger flow forecasting model for inference calculation.

[0031] The passenger flow prediction model employs a hierarchical, cascading structure for targeted triggering. Based on the traffic type and prediction area included in the prediction task, it prioritizes activating relevant prediction channels in the first prediction layer to perform preliminary predictions based on time series and flow-direction characteristics. If the model determines that the output of a single channel is insufficient to meet the decision-making accuracy requirements of the task, it automatically triggers the second prediction layer, invoking interactive prediction units related to other modes of transportation. This involves combining spatial location relationships and historical collaborative behavior to perform cross-channel prediction completion, and the prediction results are then integrated into the output layer for fusion and reconstruction. The passenger flow prediction results generated by the integrated output layer include information such as passenger flow intensity distribution, flow trends, and transfer hotspots for various regions and modes of transportation within a specific future time period.

[0032] After the prediction results are generated, the system converts them into a visualized passenger flow guidance map. Specifically, based on the spatial walkway topology within the transportation hub, the system constructs a node-path hub topology map; it uses flow direction data from the prediction results to mark path directions and uses flow intensity to indicate path thickness or color, dynamically drawing passenger flow migration trends to form an intuitive graphical guidance map. This guidance map is displayed in real time through terminal devices (such as traffic control screens, duty terminals, mobile dispatch terminals, etc.) to help dispatchers understand the evolution of passenger flow. Congestion identification and early warning management are based on the passenger flow guidance map; by analyzing the density of flow identifiers in the hub topology map, the system automatically determines the location and level of potential congestion and generates corresponding alarm information.

[0033] Furthermore, receiving prediction tasks and transferring them to the passenger flow prediction model for hierarchical cascading targeted triggering and decision-making includes:

[0034] The system receives the prediction task, interprets the task, and determines the directional prediction information. Based on the directional prediction information, it executes prediction channel triggering based on the first prediction layer, performs time node judgment and traffic-direction prediction, and determines single-channel prediction data, wherein the triggering judgment is based on the associated traffic type.

[0035] The system parses targeted prediction information based on parameters such as the target area, time period, and mode of transportation provided in the prediction task. Targeted prediction information includes, but is not limited to, the traffic type to be predicted, the specific time points for prediction, the relevant transportation hub areas, and the predicted passenger flow attributes (such as inbound and outbound flow, flow direction, and transfer flow). By interpreting this information, the system can clearly identify the specific prediction task scope and objectives required by the task.

[0036] Based on the aforementioned targeted prediction information, the system then triggers prediction channels within the first prediction layer. In this process, the system first selectively activates the corresponding prediction channels in the first prediction layer according to the determined traffic type. Specifically, each traffic transfer type corresponds to one prediction channel, and each channel is trained based on historical data for that mode of transportation to reflect its flow and direction change characteristics. When triggering a channel, the system determines the specific prediction period by judging the time node, thereby performing flow-direction prediction. At this stage, the system compares and analyzes the current time node with historical data to generate passenger flow prediction data for that mode of transportation within that time period, thus deriving the prediction result for a single channel.

[0037] To ensure the accuracy of the prediction results, the system will trigger judgments based on the "related transportation types" in the task. That is, based on the first prediction layer, if there are interactions or transfer relationships between multiple transportation types, the system will trigger predictions for the corresponding related transportation modes in parallel. For example, in the interaction scenario involving subway and bus, the system will trigger the prediction channels for both transportation modes simultaneously, and obtain more accurate passenger flow prediction data by calculating their interaction impact.

[0038] Furthermore, it is determined whether the single-channel prediction data satisfies the directional prediction information. If it does, the single-channel prediction data is used as the passenger flow prediction result.

[0039] After receiving the single-channel prediction data generated based on the first prediction layer, the system will verify the prediction data to determine whether it conforms to the directional prediction information required by the task. Specifically, the system will determine consistency by comparing the prediction target in the directional prediction information with the generated single-channel prediction data. The directional prediction information typically includes conditions such as the passenger flow intensity range, time period, mode of transportation, and regional restrictions of the prediction target. The system will verify the single-channel prediction data according to these conditions to ensure that the prediction results meet the task requirements.

[0040] For example, if a task requires forecasting subway traffic at a transportation hub and specifies a particular time window (such as peak hours), the system will verify whether the single-channel forecast data can reflect the actual passenger flow fluctuation trend of the subway during that time period and whether the traffic flow is within a reasonable range. If the single-channel forecast data meets all the constraints of the directional forecast information (such as time consistency, traffic range, and flow direction accuracy), then the data can be used as the final passenger flow forecast result.

[0041] If the single-channel prediction data meets the requirements of the directional prediction information, the system will output the data as the final result of this passenger flow prediction for use in subsequent stages.

[0042] Furthermore, if the conditions are not met, the second prediction layer is triggered, and the single-channel prediction data is imported into the second prediction layer to perform cluster prediction based on spatial location interaction and determine interactive prediction data; the single-channel prediction data and the interactive prediction data are imported into the integration output layer to perform passenger flow cycle integration based on spatial distribution and determine the passenger flow prediction result.

[0043] When the system determines that the single-channel prediction data does not meet the requirements for directional prediction information, it automatically triggers a second prediction layer for further prediction supplementation and correction. Specifically, if the single-channel prediction data has a bias or is insufficient to provide adequate prediction accuracy, the system imports the data into the second prediction layer and performs clustered prediction based on spatial location interaction. The second prediction layer is used to model the interaction effects between multiple modes of transportation to supplement the deficiencies in the single-mode prediction data.

[0044] In the second prediction layer, the system employs a spatial location interaction modeling method to aggregate and predict the spatial correlations and transfer flows between multiple modes of transportation. This includes constructing an interaction relationship graph between different modes of transportation based on historical data and real-time monitoring information, and identifying transfer hotspots and flow convergence points between each mode of transportation, thereby obtaining more comprehensive and accurate passenger flow prediction data. Through the predictions of the second prediction layer, the system can better capture complex traffic interaction patterns and potential passenger flow aggregation phenomena, obtaining interactive prediction data.

[0045] After importing single-channel and interactive prediction data into the integrated output layer, the system fuses and optimizes the two. In the integrated output layer, the system performs passenger flow circulation integration based on the spatial distribution characteristics and traffic intensity of the transportation hub; that is, it makes global adjustments to traffic flow and direction based on passenger flow patterns in different areas within the transportation hub. The integrated output layer uses an ensemble algorithm to weight and integrate the two prediction data, and adjusts and optimizes the final passenger flow prediction results according to the actual structure and traffic patterns of the transportation hub.

[0046] Furthermore, converting to a passenger flow guidance map includes:

[0047] For the first transportation hub, a hub topology map is constructed based on the hierarchical spatial street topology; a first identifier is defined by flow direction and a second identifier is defined by flow rate; the topology drawing of the passenger flow prediction results is performed on the hub topology map to determine the passenger flow guidance map.

[0048] After obtaining the final passenger flow forecast, the system converts the forecast into a passenger flow guidance map to achieve visualized management and decision support for the transportation hub. Specifically, firstly, the system constructs a hub topology map based on the hierarchical spatial street topology structure of the first transportation hub's geographic spatial layout. The hub topology map not only reflects various traffic paths and nodes within the hub but also incorporates the transfer relationships and flow patterns between different modes of transportation (such as subway, bus, taxi, etc.), forming the overall structural framework of the hub.

[0049] The system utilizes two core predictive data points—flow direction and flow rate—to define two identifiers in the topology map. Specifically, flow direction is defined as the direction of traffic flow along each path within the hub, representing the path taken by passengers from one node to another; while flow rate represents the passenger flow intensity on each path, measuring the passenger volume passing through that path at a specific time point. Based on this, the system matches and plots the flow direction and flow rate data from the passenger flow prediction results with the corresponding nodes and paths in the hub's topology map. Through the topology plotting process, the system dynamically correlates each traffic flow direction and flow rate value with the paths within the hub, visually displaying the congestion status, flow rate fluctuations, and predicted changes of each traffic route, forming a passenger flow guidance map with spatial topological characteristics.

[0050] Furthermore, congestion early warning management includes:

[0051] Identify the passenger flow guidance map, locate congestion points based on flow identifiers, and generate alarm information according to the congestion level; based on the passenger flow guidance map, make passenger flow evacuation guidance decisions according to preset rules of congestion mode-evacuation mode, and determine the evacuation plan; perform congestion early warning management based on the alarm information and the evacuation plan.

[0052] After generating a passenger flow guidance map, the system uses this map for congestion early warning management, enabling real-time monitoring and early warning of potential congestion within transportation hubs. Specifically, the system first identifies various flow identifiers in the passenger flow guidance map and automatically locates congestion points based on the spatial distribution and trend of flow data. During this process, the system analyzes the flow values ​​on each path to determine which paths have passenger flow exceeding predetermined thresholds, thereby identifying potential congestion areas, including key nodes such as transportation stations, transfer corridors, and lanes.

[0053] Once the system identifies a congested area, it generates corresponding alarm information based on the defined congestion level (e.g., light congestion, moderate congestion, severe congestion). The alarm information includes not only the specific location of the congestion, the congestion level, and the potential impact area, but also the warning time window and passenger flow trends, helping managers quickly assess the urgency and expected development of the congestion.

[0054] After generating an alarm, the system automatically makes passenger flow evacuation guidance decisions based on the provided passenger flow guidance map and preset congestion mode-evacuation rules. These evacuation mode rules are preset schemes developed based on historical data and traffic management experience, covering evacuation methods, priority evacuation routes, and the capacity of each evacuation route under different types of congestion. For example, during peak hours or emergencies, the system will recommend diverting passenger flow to alternative routes or transfer stations with lower traffic volume to effectively alleviate passenger flow pressure in congested areas. Based on the above evacuation guidance decisions, the system further determines the evacuation plan, including the selection of evacuation routes, adjustments to traffic control, and the issuance of guidance signs, to ensure the efficiency and safety of the evacuation process. During execution, the system dynamically adjusts the evacuation plan based on changes in real-time data to ensure optimal passenger flow guidance. Finally, the system combines the alarm information and the evacuation plan to activate congestion early warning management. By continuously monitoring changes in the passenger flow guidance map, the system updates the congestion status and evacuation execution in real time, and provides real-time information pushes and reminders to management personnel, dispatchers, and the public inside and outside the transportation hub. For example, managers can view congestion maps through terminal devices and take appropriate scheduling measures, while the public can obtain the latest evacuation guidance information through a real-time announcement system.

[0055] In summary, the embodiments of this application have at least the following technical effects:

[0056] First, the traffic transfer types of the first transportation hub are determined, with at least two types included. Then, based on single-traffic-type flow-direction prediction and multi-traffic-type interactive prediction, a passenger flow prediction model is constructed for each traffic transfer type. This model is embedded in the traffic management system and includes a three-layer fully connected first prediction layer, a second prediction layer, and an integrated output layer. Finally, prediction tasks are received, combined with real-time passenger flow monitoring data, and fed into the passenger flow prediction model for hierarchical cascading targeted triggering and decision-making. The passenger flow prediction results are then converted into a passenger flow guidance map for visualization on terminal devices and congestion warning management. This approach solves the technical problems of low passenger flow prediction accuracy and difficulty in handling multi-traffic-type interactive scenarios in existing technologies, achieving the technical effect of improving passenger flow prediction accuracy and system response efficiency.

[0057] Example 2 is based on the same inventive concept as the big data-based transportation hub passenger flow prediction method in the previous examples, such as... Figure 2 As shown, this application provides a big data-based passenger flow prediction system for transportation hubs, wherein the system includes:

[0058] Traffic type determination module 11: Determines the traffic transfer type of the first transportation hub, wherein the traffic transfer type includes at least two; Model construction module 12: Constructs a passenger flow prediction model for the traffic transfer type using flow-direction prediction for a single traffic type and interactive prediction for multiple traffic types, wherein the passenger flow prediction model is embedded in the traffic management system and includes a three-layer fully connected first prediction layer, second prediction layer and integrated output layer; Management module 13: Receives prediction tasks, combines them with real-time passenger flow monitoring data, transfers them to the passenger flow prediction model for hierarchical cascading directional triggering and decision-making, determines the passenger flow prediction results, converts them into a passenger flow guidance map, and performs terminal equipment visualization and congestion early warning management.

[0059] Furthermore, the model building module 12 is used to perform the following methods:

[0060] A first traffic type is determined, wherein the first traffic type is any one of the traffic transfer types; for the first traffic type, a first prediction channel is constructed using time series as the first judgment node and flow-direction as the second prediction node; the traffic transfer types are traversed until the construction of the Nth prediction channel is completed, and the first prediction channel is constructed in parallel until the Nth prediction channel is constructed, thereby generating the first prediction layer.

[0061] Furthermore, the management module 13 is used to perform the following methods:

[0062] The system receives the prediction task, interprets the task, and determines the directional prediction information. Based on the directional prediction information, it executes prediction channel triggering based on the first prediction layer, performs time node judgment and traffic-direction prediction, and determines single-channel prediction data, wherein the triggering judgment is based on the associated traffic type.

[0063] Furthermore, the management module 13 is used to perform the following methods:

[0064] Determine whether the single-channel prediction data satisfies the directional prediction information. If it does, use the single-channel prediction data as the passenger flow prediction result.

[0065] Furthermore, the management module 13 is used to perform the following methods:

[0066] If the conditions are not met, the second prediction layer is triggered, and the single-channel prediction data is imported into the second prediction layer to perform cluster prediction based on spatial location interaction and determine the interactive prediction data; the single-channel prediction data and the interactive prediction data are imported into the integration output layer to perform passenger flow cycle integration based on spatial distribution and determine the passenger flow prediction result.

[0067] Furthermore, the management module 13 is used to perform the following methods:

[0068] For the first transportation hub, a hub topology map is constructed based on the hierarchical spatial street topology; a first identifier is defined by flow direction and a second identifier is defined by flow rate; the topology drawing of the passenger flow prediction results is performed on the hub topology map to determine the passenger flow guidance map.

[0069] Furthermore, the management module 13 is used to perform the following methods:

[0070] Identify the passenger flow guidance map, locate congestion points based on flow identifiers, and generate alarm information according to the congestion level; based on the passenger flow guidance map, make passenger flow evacuation guidance decisions according to preset rules of congestion mode-evacuation mode, and determine the evacuation plan; perform congestion early warning management based on the alarm information and the evacuation plan.

[0071] Example 3, Figure 3 This is a schematic diagram of the structure of an electronic device provided in Embodiment 3 of the present invention, showing a block diagram of an exemplary electronic device suitable for implementing the embodiments of the present invention. Figure 3 The electronic device shown is merely an example and should not be construed as limiting the functionality or scope of the embodiments of the present invention. Figure 3 As shown, the electronic device includes a processor 21, a memory 22, an input device 23, and an output device 24; the number of processors 21 in the electronic device can be one or more. Figure 3 Taking a processor 21 as an example, the processor 21, memory 22, input device 23, and output device 24 in an electronic device can be connected via a bus or other means. Figure 3 Taking the example of a connection between China and Israel via a bus.

[0072] The memory 22, as a computer-readable storage medium, can be used to store software programs, computer-executable programs, and modules, such as the program instructions / modules corresponding to the big data-based transportation hub passenger flow prediction method in this embodiment of the invention. The processor 21 executes various functional applications and data processing of the electronic device by running the software programs, instructions, and modules stored in the memory 22, thereby realizing the aforementioned big data-based transportation hub passenger flow prediction method.

[0073] It should be noted that the order of the embodiments described above is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. Furthermore, the above description focuses on specific embodiments of this specification. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some implementations, multitasking and parallel processing are possible or may be advantageous.

[0074] The above description is only a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

[0075] This specification and accompanying drawings are merely illustrative examples of this application and are intended to cover any and all modifications, variations, combinations, or equivalents within the scope of this application. Clearly, those skilled in the art can make various alterations and modifications to this application without departing from its scope. Therefore, if such modifications and modifications fall within the scope of this application and its equivalents, this application intends to include such modifications and modifications.

Claims

1. A method for predicting passenger flow in transportation hubs based on big data, characterized in that, The method includes: Determine the traffic transfer type of the first transportation hub, wherein the traffic transfer type includes at least two; Based on flow-direction prediction for a single traffic type and interactive prediction for multiple traffic types, a passenger flow prediction model is constructed for the traffic transfer type. The passenger flow prediction model is embedded in the traffic management system and includes a three-layer fully connected first prediction layer, second prediction layer and integrated output layer. The system receives prediction tasks, combines them with real-time passenger flow monitoring data, and transfers them to the passenger flow prediction model for hierarchical cascading directional triggering and decision-making. It then determines the passenger flow prediction results, converts them into a passenger flow guidance map, and performs visualization and congestion early warning management on terminal devices. The system receives prediction tasks and transfers them to the passenger flow prediction model for hierarchical cascading targeted triggering and decision-making, including: Receive the prediction task, interpret the task, and determine the directional prediction information; Based on the directional prediction information, the prediction channel triggering based on the first prediction layer is executed, time node judgment and traffic-direction prediction are performed, and single-channel prediction data is determined, wherein the triggering judgment is based on the associated traffic type; Determine whether the single-channel prediction data satisfies the directional prediction information. If it does, use the single-channel prediction data as the passenger flow prediction result. If the conditions are not met, the second prediction layer is triggered, and the single-channel prediction data is imported into the second prediction layer to perform cluster prediction based on spatial location interaction and determine the interactive prediction data. The single-channel prediction data and the interactive prediction data are imported into the integrated output layer, and passenger flow cycle integration based on spatial distribution is performed to determine the passenger flow prediction result.

2. The method for predicting passenger flow in transportation hubs based on big data as described in claim 1, characterized in that, The construction of the first prediction layer includes: Determine a first traffic type, wherein the first traffic type is any one of the traffic transfer types; For the first traffic type, a first prediction channel is constructed using time series as the first judgment node and flow-direction as the second prediction node; The traffic transfer types are traversed until the Nth prediction channel is constructed. The first prediction channel is constructed in parallel until the Nth prediction channel is constructed, thus generating the first prediction layer.

3. The method for predicting passenger flow in transportation hubs based on big data as described in claim 1, characterized in that, Convert to a passenger flow guidance map, including: For the first transportation hub, a hub topology map is constructed based on the hierarchical spatial street topology; Define a first identifier by flow direction and a second identifier by flow rate, perform topology drawing of the passenger flow prediction results in the hub topology map, and determine the passenger flow guidance map.

4. The method for predicting passenger flow in transportation hubs based on big data as described in claim 1, characterized in that, Implement congestion early warning management, including: Identify the passenger flow guidance map, locate congestion points based on flow identifiers, and generate alarm information according to the congestion level; Based on the passenger flow guidance map, and using the preset rules of congestion mode-evacuation mode, passenger flow evacuation guidance decisions are made to determine the evacuation plan; Based on the alarm information and the evacuation plan, congestion early warning management is carried out.

5. A big data-based passenger flow prediction system for transportation hubs, characterized in that: The system is used to implement the big data-based passenger flow prediction method for transportation hubs according to any one of claims 1-4, the system comprising: Traffic type determination module: determines the traffic transfer type of the first traffic hub, wherein the traffic transfer type includes at least two; Model building module: Based on flow-direction prediction for a single traffic type and interactive prediction for multiple traffic types, a passenger flow prediction model is built for the traffic transfer type. The passenger flow prediction model is embedded in the traffic management system and includes a three-layer fully connected first prediction layer, second prediction layer and integrated output layer. Management module: Receives prediction tasks, combines them with real-time passenger flow monitoring data, and transfers them to the passenger flow prediction model for hierarchical cascading targeted triggering and decision-making. It determines the passenger flow prediction results, converts them into passenger flow guidance maps, and performs visualization and congestion early warning management for terminal devices.

6. An electronic device, characterized in that, The electronic device includes: Memory, used to store executable instructions; The processor, when executing executable instructions stored in the memory, implements the big data-based passenger flow prediction method for transportation hubs as described in any one of claims 1-4.

Images