Multi-mode collaborative passenger flow acquisition and guidance method and system for honeycomb bus port in peak period

Abstract

The invention discloses a multi-mode collaborative passenger flow acquisition and guidance method and system for a honeycomb bus port in a peak period, and the method comprises the steps: obtaining the real-time passenger flow video data, collected by hardware equipment, of a target region, and calculating the risk score of the target region according to the real-time passenger flow video data; generating a corresponding passenger flow guiding strategy based on the risk score; the passenger flow guiding strategy comprises an unmanned aerial vehicle guiding strategy and a ground guiding strategy; controlling the unmanned aerial vehicle and the ground personnel to guide the passenger flow according to the passenger flow guide strategy; according to the invention, the accuracy, high efficiency and intelligentization of passenger flow guidance are realized, the guidance efficiency and emergency response speed are effectively improved, and the system toughness and safety in a large passenger flow scene are significantly enhanced.

Description

Technical Field

[0001] This invention relates to the field of passenger flow guidance technology, and in particular to a multimodal collaborative passenger flow collection and guidance method and system for a honeycomb bus port during peak hours. Background Technology

[0002] Against the backdrop of increasingly complex urban transportation systems, densely populated areas face severe station congestion and chaotic travel order during morning and evening rush hours. This is especially true in bus hub areas, where traditional stations, due to their rigid spatial organization and lack of dynamic scheduling mechanisms, struggle to cope with the demand for a large number of passengers boarding, alighting, or transferring within a short period. This can easily lead to severe congestion in localized areas, affecting operational safety and traffic efficiency. While the diversified development of intelligent station organization has improved the carrying capacity of station structures, how to achieve reasonable passenger dispersion and guidance within the same station during peak hours, and avoid regional congestion and service imbalances in a short period, remains a pressing challenge that needs to be overcome.

[0003] Current research on passenger flow guidance at bus stops often suffers from response delays and unstable judgments under conditions of sudden changes in passenger flow or severe obstruction, making it difficult to achieve refined, dynamic scheduling and continuous optimization of passenger flow within the station. Therefore, there is an urgent need to develop a collaborative guidance method for large-scale smart bus hubs during peak hours, possessing multimodal linkage capabilities and full-cycle control capabilities, to achieve orderly dispersion of high-density passenger flow, coordinated transfers, and stable operation. Summary of the Invention

[0004] This invention provides a multimodal collaborative passenger flow collection and guidance method for honeycomb bus ports during peak hours, which realizes the precision, efficiency and intelligence of passenger flow guidance, effectively improves the efficiency of traffic management and the speed of emergency response, and significantly enhances the system resilience and safety in large passenger flow scenarios.

[0005] In a first aspect, embodiments of the present invention provide a method for multimodal collaborative passenger flow collection and guidance at a honeycomb bus terminal during peak hours, including: Acquire real-time passenger flow video data of the target area collected by hardware devices, and calculate the risk score of the target area based on the real-time passenger flow video data; Based on the risk score, a corresponding passenger flow guidance strategy is generated; the passenger flow guidance strategy includes drone guidance strategy and ground guidance strategy. Control drones and ground personnel to guide passenger flow according to the aforementioned passenger flow guidance strategy.

[0006] Furthermore, the step of calculating the risk score of the target area based on the real-time passenger flow video data includes: Target detection is performed on the real-time passenger flow video data, and the real-time passenger flow density value of the target passenger in each frame of the passenger flow image is calculated; The average velocity amplitude and motion direction entropy of all the target passengers are calculated using a sparse optical flow algorithm. Obtain the historical passenger flow density sequence and input it into the temporal convolutional network, then output the predicted passenger flow density value for the next time step as predicted by the network. The risk score of the target area is calculated by weighting and summing the real-time passenger flow density value, the average movement speed amplitude, the movement direction entropy, and the predicted passenger flow density value.

[0007] Furthermore, the step of performing target detection on the real-time passenger flow video data and calculating the real-time passenger flow density value of the target passenger in each frame of the passenger flow image includes: For each frame of passenger flow image in the real-time passenger flow video data, the bounding box information of the target passenger is extracted using the YOLOv8s object detection algorithm; Based on the bounding box information, all target passengers are selected using a region of interest filtering algorithm, and the real-time passenger flow density value is calculated.

[0008] Furthermore, the method also includes: When the amount of data in the historical passenger flow density sequence is insufficient, the predicted passenger flow density for the next moment is calculated using an exponentially weighted moving average model.

[0009] Furthermore, the step of generating a corresponding passenger flow guidance strategy based on the risk score includes: When the risk score is within the preset low-risk score range, the corresponding drone guidance strategy is to perform automated inspection tasks, and the ground guidance strategy is to perform passenger service and facility maintenance tasks. When the risk score is within the preset medium risk range, the corresponding drone guidance strategy is to perform an active guidance task, and the ground guidance strategy is to perform an on-site evacuation task. When the risk score is within the preset high-risk range, the corresponding drone guidance strategy is to perform an airborne command mission, and the ground guidance strategy is to perform an emergency response mission.

[0010] Furthermore, the method also includes: Calculate the risk score of the target area after passenger flow guidance and compare it with the risk score before passenger flow guidance; The passenger flow guidance strategy is evaluated and optimized based on the comparison results.

[0011] Furthermore, the control of drones and ground personnel to guide passenger flow according to the passenger flow guidance strategy includes: Send a mission execution command to the drone so that the drone can plan the optimal flight path based on the mission execution command and a pre-stored target area map; wherein, the mission execution command includes target area information and guidance strategy parameters; Control the drone to fly to the target area along the optimal flight path and execute the corresponding task based on the guidance strategy parameters.

[0012] Secondly, embodiments of the present invention provide a multimodal collaborative passenger flow collection and guidance system for a peak-hour honeycomb bus terminal, comprising: The data acquisition module is used to acquire real-time passenger flow video data of the target area collected by the hardware device, and calculate the risk score of the target area based on the real-time passenger flow video data; The strategy generation module is used to generate corresponding passenger flow guidance strategies based on the risk score; the passenger flow guidance strategies include drone guidance strategies and ground guidance strategies; The passenger flow guidance module is used to control drones and ground personnel to guide passenger flow according to the passenger flow guidance strategy.

[0013] Thirdly, embodiments of the present invention provide an electronic device, comprising: Memory, used to store computer programs; A processor for executing the computer program; When the processor executes the computer program, it implements the multimodal collaborative passenger flow collection and guidance method for the peak-hour honeycomb bus port as described in any of the first aspects above.

[0014] Fourthly, embodiments of the present invention provide a computer-readable storage medium storing a computer program, which, when executed, implements the multimodal collaborative passenger flow collection and guidance method for the peak-hour honeycomb bus port described in any of the first aspects above.

[0015] Compared with existing technologies, the multimodal collaborative passenger flow collection and guidance method for a honeycomb bus port during peak hours provided by this invention has the following advantages: It acquires real-time passenger flow video data of the target area collected by hardware devices; calculates a risk score for the target area based on the real-time passenger flow video data; generates a corresponding passenger flow guidance strategy based on the risk score; the passenger flow guidance strategy includes a drone guidance strategy and a ground guidance strategy; and controls drones and ground personnel to guide passenger flow according to the passenger flow guidance strategy. This invention achieves precise, efficient, and intelligent passenger flow guidance, effectively improving traffic management efficiency and emergency response speed, and significantly enhancing system resilience and safety in high-passenger-flow scenarios. Attached Figure Description

[0016] To more clearly illustrate the technical features of the embodiments of the present invention, the drawings used in the embodiments of the present invention will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 This is a flowchart illustrating a multimodal collaborative passenger flow collection and guidance method for a honeycomb bus port during peak hours, provided by an embodiment of the present invention. Figure 2 This is a schematic diagram of the structure of a multimodal collaborative passenger flow collection and guidance system for a honeycomb bus port during peak hours, provided in an embodiment of the present invention. Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0018] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0019] It should be noted that although functional modules are divided in the device schematic diagram and a logical order is shown in the flowchart, in some cases, the steps shown or described may be performed in a different order than the module division in the device or the order in the flowchart. The terms "first," "second," etc., in the specification, claims, and the aforementioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence.

[0020] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used herein is for the purpose of describing embodiments of the invention only and is not intended to limit the invention.

[0021] In a first aspect, embodiments of the present invention provide a method for multimodal collaborative passenger flow collection and guidance at a honeycomb bus terminal during peak hours, see [link to relevant documentation]. Figure 1 This is a flowchart illustrating an embodiment of a multimodal collaborative passenger flow collection and guidance method for a honeycomb bus port during peak hours provided by the present invention.

[0022] like Figure 1 As shown, the method includes the following steps: S1: Acquire real-time passenger flow video data of the target area collected by the hardware device, and calculate the risk score of the target area based on the real-time passenger flow video data; S2: Generate a corresponding passenger flow guidance strategy based on the risk score; the passenger flow guidance strategy includes a drone guidance strategy and a ground guidance strategy; S3: Control the drones and ground personnel to guide passenger flow according to the passenger flow guidance strategy.

[0023] It should be noted that, to clearly define the specific research object targeted by this invention, the term "Honeycomb Bus Port" is proposed. This term specifically refers to a collection of hub bus stations possessing comprehensive public service functions and diverse intelligent facilities (including drone cabins), distinct from traditional bus stations that only serve as vehicle stops and passenger pick-up / drop-off points. It is an integrated transportation hub combining rich public services and advanced intelligent technologies. This invention's multimodal collaborative passenger flow collection and guidance method for honeycomb bus ports during peak hours aims to accurately and effectively disperse peak passenger flow pressure within the station, alleviate platform congestion risks, and establish an efficient passenger flow guidance mechanism with spatial resolution and behavioral responsiveness, thereby improving the overall operational efficiency and passenger experience of the bus port during peak hours. This provides a new solution for intelligent bus station design and planning that balances efficiency, flexibility, and robustness.

[0024] In practice, real-time passenger flow video data of the target area is acquired by hardware devices, and the risk score of the target area is calculated based on the real-time passenger flow video data. The score ranges from (0,1). This score directly reflects the congestion risk level of the target platform area. When the value is close to 1, it indicates that the passenger density is high, the mobility is poor, and there is potential congestion or safety hazards.

[0025] Based on the risk score, corresponding passenger flow guidance strategies are generated, including drone guidance strategies and ground guidance strategies. Drones and ground personnel are controlled to guide passenger flow according to the passenger flow guidance strategies, thereby realizing the linkage control of drones and ground guidance equipment.

[0026] In summary, this invention, through a closed-loop design encompassing "real-time passenger flow data collection, multi-dimensional risk quantification, and air-ground collaborative guidance execution," not only relies on hardware devices to accurately acquire passenger flow video data and achieves scientific quantification of risk scores through multi-indicator fusion calculation, but also matches differentiated drone and ground guidance strategies based on risk levels. This enables precise, efficient, and intelligent passenger flow guidance, effectively improving traffic management efficiency and emergency response speed, and significantly enhancing system resilience, operational economy, and passenger safety in high-traffic scenarios.

[0027] In one optional implementation, calculating the risk score of the target area based on the real-time passenger flow video data includes: Target detection is performed on the real-time passenger flow video data, and the real-time passenger flow density value of the target passenger in each frame of the passenger flow image is calculated; The average velocity amplitude and motion direction entropy of all the target passengers are calculated using a sparse optical flow algorithm. Obtain the historical passenger flow density sequence and input it into the temporal convolutional network, then output the predicted passenger flow density value for the next time step as predicted by the network. The risk score of the target area is calculated by weighting and summing the real-time passenger flow density value, the average movement speed amplitude, the movement direction entropy, and the predicted passenger flow density value.

[0028] Specifically, the video images are first converted into quantifiable density values ​​to reflect the current level of congestion in the area. Then, a target detection algorithm is used to generate a density estimation matrix of the target passengers for each frame of the real-time passenger flow video data. Based on the density estimation matrix, the real-time passenger flow density value is calculated.

[0029] Furthermore, a Temporal Convolutional Network (TCN) is used to predict future passenger flow changes. Historical passenger flow density sequences are extracted in units of sliding time windows. Passenger flow density is defined as the ratio of the number of people to the maximum number of people or a specified upper limit, as shown below: ; in, The value represents the passenger flow density, ranging from [0,1]. Let t be the real-time number of passengers in the target area. Specify the maximum capacity or upper limit for the area.

[0030] Construct a sliding window with a length of 32, and the historical passenger flow density sequence is as follows: The data is then fed into a temporal convolutional network (TCN) to output the predicted passenger flow density for the next time step. The TCN model consists of three dilated convolutional layers with dilation coefficients of 1, 2, and 4, respectively. It employs residual connections and ReLU activation, and uses mean squared error as the loss function. The training objective MSE of the TCN model is: ; The training objective is to minimize the squared difference between the predicted and the true values.

[0031] After the model is trained, a rolling inference mode is adopted. For each new real-time density, the sliding window is scrolled forward one position and input into the TCN model to obtain the predicted value at the next time step. Finally, the prediction curve with the same length as the original sequence is output to judge the density increase trend in advance.

[0032] Furthermore, a sparse optical flow algorithm is used to calculate the average velocity amplitude and direction entropy of all target passengers. The average velocity amplitude reflects the overall flow intensity of the crowd from the speed dimension. The higher the speed, the better the flow and the lower the risk of congestion. The calculation formula is as follows: ; in, Let be the average velocity amplitude of the group at time t. Let be the effective target number of passengers at time t, and let be the velocity amplitude of the kth passenger.

[0033] Motion direction entropy reflects the consistency of crowd movement in terms of direction. It is used to identify local congestion or disorder trends. The lower the entropy value, the more uniform the direction, the better the order, and the lower the risk of congestion. First, the motion direction angle of each target passenger is calculated: ; in, Let the motion direction angle of the k-th target passenger be , This refers to the passenger's horizontal displacement. This refers to the passenger's vertical displacement.

[0034] Divide the 0°~360° omnidirectional circle into 8 sections and calculate the percentage of passengers in each section. And further calculate the entropy of the motion direction: ; in, The entropy is the direction of motion. For direction number, It is a local minimum.

[0035] The risk score for the target area is calculated by weighted summation after Minmax normalization of real-time passenger flow density, average movement speed amplitude, movement direction entropy, and predicted passenger flow density. The calculation formula is as follows: ; in, The risk score for the target region at time t. This is the real-time passenger flow density value. This is the predicted value for passenger flow density. The average velocity amplitude, The entropy is the direction of motion. These are the corresponding weight coefficients, and Exemplary It can take the value 0.45. It is 0.35. It is 0.1. It is 0.1.

[0036] It should be noted that passenger flow density is easily affected by instantaneous factors (such as a sudden increase or decrease in density due to rapid movement of people, sudden changes in lighting, or camera shake in a certain frame). Directly determining the risk level based on a single frame of data can lead to false triggers, such as misjudging a low-risk situation as a medium-risk situation, initiating unnecessary guidance, affecting operational efficiency and wasting resources. This embodiment introduces a hysteresis judgment mechanism, which uses continuous sampling period verification to filter instantaneous noise, ensuring that the risk level determination is based on a stable passenger flow trend rather than a single fluctuation, thereby improving the reliability of the system's decision-making.

[0037] This embodiment achieves comprehensive perception of passenger flow density, future density trends, directional consistency, and short-term evolution trends within the platform through multi-dimensional data fusion and intelligent algorithms. It maps static congestion characteristics and dynamic behavioral characteristics into a continuous risk score, thereby realizing the transformation from single threshold judgment to multi-dimensional dynamic evaluation. It can accurately depict the instantaneous state and changing trend of passenger flow, significantly improving the recognition stability and timeliness in complex scenarios, and providing a reliable decision-making basis for differentiated guidance within the station.

[0038] In one optional implementation, the step of performing target detection on the real-time passenger flow video data and calculating the real-time passenger flow density value of the target passenger in each frame of the passenger flow image includes: For each frame of passenger flow image in the real-time passenger flow video data, the bounding box information of the target passenger is extracted using the YOLOv8s object detection algorithm; Based on the bounding box information, all target passengers are selected using a region of interest filtering algorithm, and the real-time passenger flow density value is calculated.

[0039] Specifically, lightweight YOLOv8s is used to perform object detection on real-time passenger flow video data, retaining only the set of bounding boxes for the category "person". : ; ; in, This is the i-th valid bounding box (the confidence level must be greater than the preset threshold). The coordinates of the top left corner The coordinates are the bottom right corner.

[0040] For each bounding box, calculate its overlap ratio with the ROI mask. : ; in, This represents the total number of pixels in the bounding box. This represents the number of pixels within the bounding box that belong to the ROI.

[0041] like ( If the overlap threshold is set to 0.5 or the bounding box center is located within the ROI, then the target passenger is determined to be a passenger within the target area. The number of bounding boxes that meet the conditions is the number of target passengers in the current frame. The real-time passenger flow density value of the target area is calculated based on the number of target passengers.

[0042] This embodiment combines YOLOv8s' efficient target detection with dynamic region of interest filtering to achieve high accuracy, low latency, and strong adaptability in real-time passenger flow density calculation, providing a reliable data foundation for subsequent risk assessment.

[0043] In an optional implementation, the method further includes: When the amount of data in the historical passenger flow density sequence is insufficient, the predicted passenger flow density for the next moment is calculated using an exponentially weighted moving average model.

[0044] Specifically, when the TCN model fails to function due to insufficient sample length (e.g., when the system has just started and historical data is scarce) or the deep learning module is unavailable (e.g., due to computational resource overload or model failure), the Exponentially Weighted Moving Average (EWMA) model can be used as an alternative to ensure uninterrupted prediction. The recursive formula for EWMA is: ; ; in, This is the predicted value for passenger flow density. Here is the EWMA prediction value at time t. Let be the real-time relative density at time t. The EWMA prediction value at time t-1. This is the weighting coefficient, which can be 0.5.

[0045] Understandably, EWMA only requires the current data and the smoothed value from the previous time step to be calculated, without the need for complex model training or a large amount of historical data, and can be used as an alternative to the TCN model.

[0046] This embodiment uses an exponentially weighted average method to complement the TCN temporal convolutional network, achieving adaptive prediction backoff under conditions of sparse or abrupt sample changes.

[0047] In one optional implementation, generating a corresponding passenger flow guidance strategy based on the risk score includes: When the risk score is within the preset low-risk score range, the corresponding drone guidance strategy is to perform automated inspection tasks, and the ground guidance strategy is to perform passenger service and facility maintenance tasks. When the risk score is within the preset medium risk range, the corresponding drone guidance strategy is to perform an active guidance task, and the ground guidance strategy is to perform an on-site evacuation task. When the risk score is within the preset high-risk range, the corresponding drone guidance strategy is to perform an airborne command mission, and the ground guidance strategy is to perform an emergency response mission.

[0048] Specifically, when the risk score is within the preset low-risk range, the corresponding drone guidance strategy is to perform automated inspection tasks, conduct low-altitude cruises along preset routes, collect global video data through onboard wide-angle lenses, and transmit it back to the intelligent hub to monitor whether passenger flow is abnormally increasing and whether facilities are in good condition. The onboard speakers will play vehicle schedule information and friendly reminders, playing a role in environmental awareness and basic services. At this time, ground personnel are freed from monitoring tasks and can shift their focus to passenger service and facility maintenance. They can proactively provide passengers with inquiries and assistance, help special groups board the train, and conduct daily inspections and maintenance of station facilities, making full use of the low passenger flow period to improve service quality.

[0049] When the risk score is within the preset medium-risk range, the corresponding drone guidance strategy is to perform proactive guidance. The drone automatically identifies passenger gathering points through onboard computer vision and flies to the target area for dynamic intervention. It projects light arrows on the ground through the onboard laser projection module to indicate the diversion direction. At the same time, it broadcasts diversion suggestions through directional broadcasts. Ground personnel go to the designated area according to the warning information sent by the drone, and, in coordination with the drone's aerial guidance, provide on-site explanations and gesture guidance to help passengers understand the diversion plan and prepare emergency supplies to prevent further increase in passenger flow, forming a flexible diversion system that coordinates air and ground.

[0050] When the risk score falls within the preset high-risk range, the corresponding drone guidance strategy is to execute an aerial command mission. The drone continuously hovers over key nodes, transmitting panoramic video back to the command center in real time via its onboard high-definition camera. It also uses a strong light projection to delineate the boundaries of emergency passages in the crowd and continuously broadcasts evacuation instructions through a high-power loudspeaker. Ground personnel form an emergency response team to quickly handle abnormal events based on the accurate intelligence provided by the drone. They reinforce and guide key nodes in the projection passage with human resources and strictly implement the emergency plan formulated by the command center based on the drone's perspective. Ultimately, this achieves human-machine collaborative emergency response and safety control.

[0051] Based on the quantitative results of risk scores, this embodiment divides the guidance strategy into three levels: low-risk inspection service, medium-risk proactive guidance, and high-risk emergency response. Through the collaborative mode of drone aerial execution and ground resource linkage, it not only ensures operational safety during peak periods but also improves service quality during off-peak periods.

[0052] In an optional implementation, the method further includes: Calculate the risk score of the target area after passenger flow guidance and compare it with the risk score before passenger flow guidance; The passenger flow guidance strategy is evaluated and optimized based on the comparison results.

[0053] Specifically, during passenger flow guidance, the drone continuously collects passenger flow video streams in the target area through its onboard sensor module and transmits them back to the command center in real time. Using the risk score calculation method from step S1, the risk score after guidance is calculated in real time. The real-time risk score after guidance is directly compared with the initial risk score before guidance. If the score after guidance meets the standard, a return command is issued, and the drone plans the optimal return path through its control module and accurately returns to the airport, allowing ground personnel to resume normal work. If the score after guidance does not meet the standard, a guidance upgrade command is issued, and ground personnel upgrade their actions simultaneously. If the risk increases instead of decreasing after guidance, a strategy switching command is issued, and the system triggers an alarm to notify the command center for manual intervention.

[0054] This embodiment uses the logic of "real-time evaluation - dynamic adjustment - long-term iteration" to upgrade the passenger flow guidance strategy from a fixed template to an adaptive intelligent solution, providing core technical support for the long-term, efficient and safe operation of the smart bus port.

[0055] In one optional implementation, the control of the drone and ground personnel to guide passenger flow according to the passenger flow guidance strategy includes: Send a mission execution command to the drone so that the drone can plan the optimal flight path based on the mission execution command and a pre-stored target area map; wherein, the mission execution command includes target area information and guidance strategy parameters; Control the drone to fly to the target area along the optimal flight path and execute the corresponding task based on the guidance strategy parameters.

[0056] Specifically, the drone consists of five main modules and specific execution units. The power module includes a battery unit and a power management unit; the communication module includes a data transmission unit and a signal processing unit; the sensor module includes an image acquisition unit, a positioning unit, and an environmental perception unit; the drive module includes a motor drive unit and a power execution unit; and the control module includes a main control unit and a bus unit, etc.

[0057] When controlling the drone to guide passenger flow according to the passenger flow guidance strategy, the system sends a task execution command to the drone, including target area information (GPS coordinates of the target area, ROI boundary coordinates, area area, current risk level, etc.) and guidance strategy parameters (cruising altitude, cruising cycle, laser projection mode, gathering point identification threshold, emergency passage width, etc.). After the drone receives the command, the control module automatically verifies the integrity of the parameters. If the verification is successful, the drone returns a command reception confirmation signal to the decision-making unit through the communication module. If the verification fails, a parameter error message is returned, requesting the re-issuance of the command to avoid guidance failure due to parameter errors.

[0058] Furthermore, the drone undergoes a comprehensive hardware self-check to ensure that the hardware is fault-free. The power module checks the battery capacity to ensure that it meets the energy consumption requirements for mission round trips. The communication module verifies the link connection with the ground control station and field edge nodes, confirms that the data transmission channel is unobstructed through signal handshake. The sensor module checks whether it can normally acquire passenger flow images, location information, and surrounding obstacle data. All test results are summarized to form a self-check report. The deployment phase only begins when all indicators are normal.

[0059] After self-checking, the drone plans the optimal path from the cabin to the target area based on the pre-stored regional map data. During flight, the environmental perception unit avoids obstacles, and the positioning unit corrects the position to ensure accurate arrival at the designated area. After arrival, the drone maintains stable hovering, and the image acquisition unit transmits passenger flow images to ensure timely intervention. It also executes corresponding tasks based on the guidance strategy parameters.

[0060] This embodiment ensures that drones and ground personnel can work together efficiently according to the guidance strategy through standardized command issuance, collaborative hardware activation, optimized path planning, and precise air-to-ground deployment. This achieves a precise match between risk level, guidance action, and execution effect, providing reliable support for passenger flow management.

[0061] Secondly, embodiments of the present invention provide a multimodal collaborative passenger flow collection and guidance device for a honeycomb bus terminal during peak hours, see [link to relevant documentation]. Figure 2 This is a schematic diagram of an embodiment of a multimodal collaborative passenger flow collection and guidance device for a honeycomb bus port during peak hours provided by the present invention.

[0062] like Figure 2 As shown, the device includes: Data acquisition module 21 is used to acquire real-time passenger flow video data of the target area collected by hardware devices, and calculate the risk score of the target area based on the real-time passenger flow video data; Strategy generation module 22 is used to generate corresponding passenger flow guidance strategies based on the risk score; the passenger flow guidance strategies include drone guidance strategies and ground guidance strategies; The passenger flow guidance module 23 is used to control the drone and ground personnel to guide passenger flow according to the passenger flow guidance strategy.

[0063] In one optional implementation, calculating the risk score of the target area based on the real-time passenger flow video data includes: Target detection is performed on the real-time passenger flow video data, and the real-time passenger flow density value of the target passenger in each frame of the passenger flow image is calculated; The average velocity amplitude and motion direction entropy of all the target passengers are calculated using a sparse optical flow algorithm. Obtain the historical passenger flow density sequence and input it into the temporal convolutional network, then output the predicted passenger flow density value for the next time step as predicted by the network. The risk score of the target area is calculated by weighting and summing the real-time passenger flow density value, the average movement speed amplitude, the movement direction entropy, and the predicted passenger flow density value.

[0064] In one optional implementation, the step of performing target detection on the real-time passenger flow video data and calculating the real-time passenger flow density value of the target passenger in each frame of the passenger flow image includes: For each frame of passenger flow image in the real-time passenger flow video data, the bounding box information of the target passenger is extracted using the YOLOv8s object detection algorithm; Based on the bounding box information, all target passengers are selected using a region of interest filtering algorithm, and the real-time passenger flow density value is calculated.

[0065] In an optional embodiment, the device is further configured to: When the amount of data in the historical passenger flow density sequence is insufficient, the predicted passenger flow density for the next moment is calculated using an exponentially weighted moving average model.

[0066] In one optional implementation, generating a corresponding passenger flow guidance strategy based on the risk score includes: When the risk score is within the preset low-risk score range, the corresponding drone guidance strategy is to perform automated inspection tasks, and the ground guidance strategy is to perform passenger service and facility maintenance tasks. When the risk score is within the preset medium risk range, the corresponding drone guidance strategy is to perform an active guidance task, and the ground guidance strategy is to perform an on-site evacuation task. When the risk score is within the preset high-risk range, the corresponding drone guidance strategy is to perform an airborne command mission, and the ground guidance strategy is to perform an emergency response mission.

[0067] In an optional embodiment, the device is further configured to: Calculate the risk score of the target area after passenger flow guidance and compare it with the risk score before passenger flow guidance; The passenger flow guidance strategy is evaluated and optimized based on the comparison results.

[0068] In one optional implementation, the control of the drone and ground personnel to guide passenger flow according to the passenger flow guidance strategy includes: Send a mission execution command to the drone so that the drone can plan the optimal flight path based on the mission execution command and a pre-stored target area map; wherein, the mission execution command includes target area information and guidance strategy parameters; Control the drone to fly to the target area along the optimal flight path and execute the corresponding task based on the guidance strategy parameters.

[0069] Thirdly, embodiments of the present invention provide an electronic device, see [link to previous document]. Figure 3 The diagram shown is a structural schematic of an electronic device provided in an embodiment of the present invention.

[0070] like Figure 3 As shown, the device includes: Memory 31 is used to store computer programs; Processor 32 is used to execute the computer program; When the processor 32 executes the computer program, it implements the multimodal collaborative passenger flow collection and guidance method for the peak-hour honeycomb bus port as described in any of the above embodiments.

[0071] For example, the computer program may be divided into one or more modules / units, which are stored in the memory 31 and executed by the processor 32 to complete the present invention. The one or more modules / units may be a series of computer program instruction segments capable of performing a specific function, which describe the execution process of the computer program in the electronic device.

[0072] The processor 32 can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or any conventional processor.

[0073] The memory 31 can be used to store the computer programs and / or modules. The processor 32 implements various functions of the electronic device by running or executing the computer programs and / or modules stored in the memory 31 and calling the data stored in the memory 31. The memory 31 may mainly include a program storage area and a data storage area. The program storage area may store the operating system, at least one application program required for a function (such as sound playback function, image playback function, etc.), etc.; the data storage area may store data created according to the use of the mobile phone (such as audio data, phonebook, etc.). In addition, the memory 31 may include high-speed random access memory, and may also include non-volatile memory, such as hard disk, memory, plug-in hard disk, smart media card (SMC), secure digital card (SD) card, flash card, at least one disk storage device, flash memory device, or other volatile solid-state storage device.

[0074] It should be noted that the aforementioned electronic devices include, but are not limited to, processors and memory, as will be understood by those skilled in the art. Figure 3 The structural diagram is merely an example of the electronic device described above and does not constitute a limitation on the electronic device. It may include more components than shown in the diagram, or combine certain components, or use different components.

[0075] Fourthly, embodiments of the present invention also provide a computer-readable storage medium storing a computer program, which, when executed, implements the multimodal collaborative passenger flow collection and guidance method for the peak-hour honeycomb bus port described in any of the above embodiments.

[0076] It should be understood that the present invention can implement all or part of the processes in the above-mentioned multimodal collaborative passenger flow collection and guidance method for the Hive Bus Terminal during peak hours, which can also be accomplished by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium. When the computer program is executed by a processor, it can implement the steps of the above-mentioned multimodal collaborative passenger flow collection and guidance method for the Hive Bus Terminal during peak hours. The computer program includes computer program code, which can be in the form of source code, object code, executable file, or some intermediate form. The computer-readable medium can include: any entity or device capable of carrying the computer program code, recording media, USB flash drive, portable hard drive, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc.

[0077] The above description is merely a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. It should be noted that, for those skilled in the art, several equivalent obvious modifications and / or equivalent substitutions can be made without departing from the technical principles of the present invention, and these obvious modifications and / or equivalent substitutions should also be considered within the scope of protection of the present invention.

Claims

1. A method for multimodal collaborative passenger flow collection and guidance at a honeycomb bus terminal during peak hours, characterized in that, include: Acquire real-time passenger flow video data of the target area collected by hardware devices, and calculate the risk score of the target area based on the real-time passenger flow video data; Based on the risk score, a corresponding passenger flow guidance strategy is generated; the passenger flow guidance strategy includes drone guidance strategy and ground guidance strategy. Control drones and ground personnel to guide passenger flow according to the aforementioned passenger flow guidance strategy.

2. The multimodal collaborative passenger flow collection and guidance method for peak-hour honeycomb bus terminals as described in claim 1, characterized in that, The calculation of the risk score for the target area based on the real-time passenger flow video data includes: Target detection is performed on the real-time passenger flow video data, and the real-time passenger flow density value of the target passenger in each frame of the passenger flow image is calculated; The average velocity amplitude and motion direction entropy of all the target passengers are calculated using a sparse optical flow algorithm. Obtain the historical passenger flow density sequence and input it into the temporal convolutional network, then output the predicted passenger flow density value for the next time step as predicted by the network. The risk score of the target area is calculated by weighting and summing the real-time passenger flow density value, the average movement speed amplitude, the movement direction entropy, and the predicted passenger flow density value.

3. The multimodal collaborative passenger flow collection and guidance method for peak-hour honeycomb bus terminals as described in claim 2, characterized in that, The step of performing target detection on the real-time passenger flow video data and calculating the real-time passenger flow density value of the target passenger in each frame of the passenger flow image includes: For each frame of passenger flow image in the real-time passenger flow video data, the bounding box information of the target passenger is extracted using the YOLOv8s object detection algorithm; Based on the bounding box information, all target passengers are selected using a region of interest filtering algorithm, and the real-time passenger flow density value is calculated.

4. The multimodal collaborative passenger flow collection and guidance method for peak-hour honeycomb bus terminals as described in claim 2, characterized in that, The method further includes: When the amount of data in the historical passenger flow density sequence is insufficient, the predicted passenger flow density for the next moment is calculated using an exponentially weighted moving average model.

5. The multimodal collaborative passenger flow collection and guidance method for peak-hour honeycomb bus terminals as described in claim 1, characterized in that, The process of generating corresponding passenger flow guidance strategies based on the risk score includes: When the risk score is within the preset low-risk score range, the corresponding drone guidance strategy is to perform automated inspection tasks, and the ground guidance strategy is to perform passenger service and facility maintenance tasks. When the risk score is within the preset medium risk range, the corresponding drone guidance strategy is to perform an active guidance task, and the ground guidance strategy is to perform an on-site evacuation task. When the risk score is within the preset high-risk range, the corresponding drone guidance strategy is to perform an airborne command mission, and the ground guidance strategy is to perform an emergency response mission.

6. The multimodal collaborative passenger flow collection and guidance method for peak-hour honeycomb bus terminals as described in claim 1, characterized in that, The method further includes: Calculate the risk score of the target area after passenger flow guidance and compare it with the risk score before passenger flow guidance; The passenger flow guidance strategy is evaluated and optimized based on the comparison results.

7. The multimodal collaborative passenger flow collection and guidance method for peak-hour honeycomb bus terminals as described in claim 1, characterized in that, The control of drones and ground personnel to guide passenger flow according to the passenger flow guidance strategy includes: Send a mission execution command to the drone so that the drone can plan the optimal flight path based on the mission execution command and a pre-stored target area map; wherein, the mission execution command includes target area information and guidance strategy parameters; Control the drone to fly to the target area along the optimal flight path and execute the corresponding task based on the guidance strategy parameters.

8. A multimodal collaborative passenger flow collection and guidance device for a honeycomb bus port during peak hours, characterized in that, include: The data acquisition module is used to acquire real-time passenger flow video data of the target area collected by the hardware device, and calculate the risk score of the target area based on the real-time passenger flow video data; The strategy generation module is used to generate corresponding passenger flow guidance strategies based on the risk score; the passenger flow guidance strategies include drone guidance strategies and ground guidance strategies; The passenger flow guidance module is used to control drones and ground personnel to guide passenger flow according to the passenger flow guidance strategy.

9. An electronic device, characterized in that, include: Memory, used to store computer programs; A processor for executing the computer program; When the processor executes the computer program, it implements the multimodal collaborative passenger flow collection and guidance method for the peak-hour honeycomb bus port as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed, implements the multimodal collaborative passenger flow collection and guidance method for the peak-hour honeycomb bus port as described in any one of claims 1 to 7.

Images