A bus-road-cloud-field-station five-dimensional cooperative control system of intelligent public transport

Abstract

The present application relates to a kind of wisdom bus car-road-cloud-field-station five-dimensional collaborative control system, including vehicle end, road side equipment, cloud platform, field station system and platform system five subsystems, five subsystems are interconnected by pipeline layer, constitute complete end-pipe cloud collaborative architecture, the vehicle end uses distributed network architecture, is divided into power chassis domain, intelligent driving domain and car body control domain three functional domains according to function, three functional domains are communicated by three main CAN bus, and the communication rate of each network segment is 400-500kbps, and information transfer between each domain and network segment isolation are realized by central gateway.This application not only realizes the whole process unmanned closed loop from vehicle out of field, road operation, platform stop, passenger flow interaction to field station back to field charging, but also effectively solves the problem of incomplete system architecture dimension, information island.

Description

Technical Field

[0001] This invention relates to the field of smart public transportation technology, and in particular to a five-dimensional collaborative control system for smart public transportation, encompassing vehicle-road-cloud-site-station. Background Technology

[0002] With the rapid development of autonomous driving technology, driverless buses, as an important component of smart city transportation, are receiving increasing attention.

[0003] In the prior art, Chinese patent (publication number: CN119142365A) discloses an autonomous driving and control system for intelligent buses. This system includes a cloud subsystem and a vehicle-side subsystem. The vehicle-side subsystem comprises a collaborative control module, an in-vehicle monitoring module, and an autonomous driving control module. The cloud subsystem can determine the reference comfort threshold for the intelligent bus under various preset operating conditions based on vehicle parameters and road condition information. The collaborative control module, upon determining that the intelligent bus has entered any of these preset operating conditions, corrects the reference comfort threshold for the target operating condition based on passenger image data collected by the in-vehicle monitoring module, obtaining the target comfort threshold for that target condition. This allows the autonomous driving control module to control the intelligent bus's operation based on the target comfort threshold. However, this solution lacks a complete system architecture and a closed-loop design. Existing driverless bus systems typically only involve interaction between the vehicle and the cloud, or collaboration between the vehicle and the station, failing to integrate the five core elements—vehicle, roadside equipment, cloud platform, depot, and station—into a unified architecture for integrated design. The vehicles, roads, cloud computing, facilities, and stations operate independently, lacking standardized data interaction interfaces and collaborative control mechanisms. This prevents the entire process, from vehicle departure, road operation, station parking to station return for charging, from being truly unmanned and closed-loop.

[0004] The switching mechanism between autonomous driving and remote driving is imperfect and lacks safety. Existing systems lack a safe and reliable remote takeover mechanism when the autonomous driving system malfunctions or encounters unmanageable scenarios. The switching between autonomous driving mode and remote driving mode lacks strict interlocking and status confirmation mechanisms, posing a risk that both systems may simultaneously gain control or lose control.

[0005] There are major defects in the deployment and data transmission of the roadside perception system. Existing vehicle-road cooperative roadside devices generally have the following problems: (a) Unreasonable sensor arrangement: lidars adopt a diagonal arrangement scheme, resulting in large perception blind areas; (b) Lack of multi-source sensor fusion: some intersections are only deployed with millimeter-wave radars, which cannot effectively identify static objects and pedestrians; (c) Insufficient throughput for large-volume data transmission: current roadside devices adopt a standard framework for transmission, and the data volume of a single packet is only about 1900 bytes (10Hz), which is far from meeting the data volume required for cooperative perception (actual measurement shows that about 5406-101424 bytes is required); (d) Roadside devices do not use GPS timing to achieve time synchronization, which makes it impossible to effectively fuse vehicle-road perception data.

[0006] The station and platform system lacks in-depth collaboration with vehicles and cloud platforms. Existing station systems are mostly independently operated access control and charging management systems, which cannot link with the vehicle automatic driving system and cloud scheduling platform, making it difficult to achieve unmanned parking guidance, automatic charging and automatic station entry and exit. The platform system also only has a simple information release function, and lacks the capabilities of passenger flow monitoring, abnormal behavior identification, order access and other capabilities linked with vehicle scheduling.

[0007] The cloud scheduling system lacks adaptation to special scenarios of driverless buses. The existing bus scheduling system is mainly designed for manned buses, fails to fully consider the special requirements of driverless buses such as remote monitoring, fault diagnosis and abnormal event handling, and lacks native support for unmanned operation scenarios such as automatic return of vehicle fault codes and automatic low-power alarm. Summary of the Invention

[0008] In view of the above problems, the present invention provides a vehicle-road-cloud-field-station five-dimensional cooperative control system for smart public transport, which not only realizes the whole-process unmanned closed-loop from vehicle departure, road operation, platform berthing, passenger flow interaction to charging after returning to the station, but also effectively solves the problems of incomplete system architecture dimension and information islands in existing systems.

[0009] In order to achieve the above objectives and other related purposes, the technical solution provided by the present invention is as follows: A vehicle-road-cloud-field-station five-dimensional cooperative control system for smart public transport, comprising five subsystems: a vehicle end, a roadside device, a cloud platform, a station system and a platform system, wherein the five subsystems achieve interconnection and intercommunication through a pipeline layer to form a complete end-pipeline-cloud collaborative architecture, The vehicle end adopts a distributed network architecture, and is divided into three functional domains according to functions: a power chassis domain, an intelligent driving domain and a vehicle body control domain. The three functional domains communicate through three main CAN buses, the communication rate of each network segment is 400-500kbps, and information forwarding and network segment isolation between the domains are realized through a central gateway; The roadside equipment includes sensors, edge computing units, traffic signals, RSU devices, and timing devices, which are used to provide vehicles with a top-down perspective perception capability and solve problems related to blind spots, medium and long-distance perception, and traffic light perception. The cloud platform is the core of the system's scheduling and management, including basic service modules, scheduling service modules, diagnostic service modules, operation and maintenance service modules, and large-screen visualization modules; The station system interacts with the cloud platform in real time, receives dispatch instructions and provides feedback on vehicle status, and realizes fully automated management of the entire process of vehicle entry, charging, parking and exit. The platform system includes an information publishing module, a passenger flow monitoring module, an abnormal behavior monitoring module, an order access module, and a video storage module, serving as a crucial hub for interaction between people, stations, vehicles, and the platform.

[0010] Furthermore, the powertrain chassis domain is responsible for vehicle chassis control, including the VCU and chassis actuators; the intelligent driving domain is responsible for the functions of the autonomous driving subsystem, remote driving subsystem, and V2X subsystem. The sensors in the intelligent driving domain are connected to the HAD through a private network segment. The OBU, RCU, and HAD are all connected to this network segment. When controlling the vehicle, the output is sent to the powertrain chassis network segment and the body network segment after comprehensive decision-making by the GW, ensuring the uniqueness of the control interface of the vehicle's related components; the body control domain is responsible for the control of body accessories and / or human-machine interaction display comfort functions, with low real-time requirements. The BCM is responsible for low-voltage power supply control, sliding door control, lighting control, horn control, and rear defrosting control. The air conditioning controller is responsible for air conditioning heating and ventilation, and the HMI controller is responsible for information display and interaction.

[0011] Furthermore, the autonomous driving subsystem includes a perception module, a localization module, a map engine module, a decision planning module, and a control module. The perception module adopts a BEV+Transformer-based algorithm architecture, converting multi-channel visual sensor data into a bird's-eye view representation, and combining it with target detection and ROI point cloud segmentation based on point cloud deep learning. The localization module adopts a multi-source fusion localization scheme, combining GPS / IMU integrated navigation localization, laser point cloud feature localization, and visual feature localization. The decision planning module adopts a vehicle model-based integrated planning and control solution algorithm.

[0012] Furthermore, the remote driving subsystem includes an on-board unit and a remote driving platform unit, which reuses the video data collected by the cameras of the autonomous driving subsystem and distributes the video data through the autonomous driving controller for use by the remote driving application and the autonomous driving application, respectively.

[0013] Furthermore, the roadside equipment features a multi-sensor fusion layout: employing a multi-sensor fusion scheme combining lidar, millimeter-wave radar, and cameras, arranged at four corners or multiple points at intersections to eliminate blind spots and achieve full coverage perception at intersections; the roadside equipment also features high-precision time synchronization: achieving time synchronization through GPS / BeiDou, using the same time source as the vehicle, ensuring time consistency of vehicle-road perception data.

[0014] Furthermore, the roadside equipment features high-throughput data transmission: it adopts an enhanced network transmission architecture to transmit more than 5700 bytes of data within 6ms, meeting the transmission requirements of dynamic environment and road boundary data in collaborative perception scenarios; the roadside equipment also features V2X collaborative perception: the vehicle-mounted C-V2X OBU connects the roadside and vehicle sides to realize V2I / V2V / V2P / V2N functions, supports customized warning information and perception data sharing, and achieves L4 level beyond-line-of-sight collaborative perception.

[0015] Furthermore, the basic service module is used for area management, vehicle registration, account management, role management, and system login; the dispatch service module is used for operation monitoring, operation analysis, order management, road condition monitoring, and operation scheduling; the diagnostic service module is used for fault analysis, abnormal event analysis, real-time vehicle video viewing, and risk prediction; the operation and maintenance service module is used for inspection management and work order scheduling; and the large-screen visualization module is used for 2D standard-size large-screen display to intuitively display core data such as operation status, road condition information, and fault warnings.

[0016] Furthermore, the station system includes an unmanned access control subsystem, an intelligent charging subsystem, an unmanned parking guidance subsystem, a station monitoring subsystem, and an IoT device monitoring subsystem.

[0017] Furthermore, the unmanned access control subsystem is used for vehicles to automatically identify and enter / exit the station access control system via autonomous driving; the intelligent charging subsystem supports DC fast charging, with a charging time of less than 6 hours from 20% to 100% SOC, and automatically charges in conjunction with the vehicle's low battery alarm; the unmanned parking guidance subsystem is used for vehicles to automatically drive into designated parking spaces, realizing parking space planning; the station monitoring subsystem is used for real-time monitoring of vehicles and the environment within the station; and the IoT device monitoring subsystem is used for unified management of various IoT devices within the station.

[0018] Furthermore, the information publishing module receives data from the cloud-based dispatching platform and displays vehicle trajectory, arrival time, and distance information in real time; the passenger flow monitoring module accesses passenger flow monitoring data within the platform, identifies human targets, tracks their movement, and supports forward, reverse, and bidirectional crossing counting and cumulative statistics; the abnormal behavior monitoring module monitors abnormal passenger behavior within the platform, automatically broadcasts abnormalities via voice, and returns the information to the background; the order access module receives passenger ride-hailing order information, feeds it back to the vehicle platform, and completes on-demand, real-time, and flexible dispatching; and the video storage module stores platform video for at least 720 hours with a resolution of at least 1920×1080.

[0019] The present invention has the following positive effects: 1. This invention integrates vehicle-side equipment, roadside equipment, cloud platform, depot system, and platform system into a unified "end-pipe-cloud" architecture, realizing a fully unmanned closed loop from vehicle departure, road operation, platform parking, passenger flow interaction to depot return charging, effectively solving the problems of incomplete dimensions and information silos in the existing system architecture.

[0020] 2. This invention significantly improves the coverage and data transmission capabilities of roadside sensing. By employing multi-sensor fusion deployment, GPS time synchronization, and an enhanced network transmission architecture, it solves the problems of blind spots, missing multi-source sensors, and insufficient data transmission throughput in existing roadside equipment, achieving L4-level beyond-line-of-sight collaborative sensing.

[0021] 3. This invention improves the system's safety redundancy. Through a dual-system backup mechanism for autonomous driving and remote driving, strict interlocking switching logic, and bidirectional real-time monitoring of the chassis system and intelligent driving system, it ensures that the vehicle remains controllable even in the event of a single system failure.

[0022] 4. This invention realizes intelligent and unmanned operation of stations and platforms. Through deep collaboration between the station system, vehicles, and cloud platform, it achieves full-process automation of unmanned access control, automatic parking, and intelligent charging; through passenger flow monitoring, abnormal behavior recognition, and order access of the platform system, it achieves flexible on-demand scheduling.

[0023] 5. This invention achieves accurate collection and storage of critical data. Through an event-triggered emergency data storage and uploading mechanism, data recording is triggered only in the event of manual intervention or system failure, ensuring that critical event data is not overwritten and providing complete data support for post-event analysis and system optimization.

[0024] 6. This invention constructs a hierarchical cloud-based scheduling and management system. Through a modular design of basic services, scheduling services, diagnostic services, operation and maintenance services, and large-screen visualization, it achieves comprehensive control and intelligent scheduling of driverless bus operations. Attached Figure Description

[0025] Figure 1 This is a diagram of the architecture of the driverless intelligent bus system of the present invention; Figure 2 This is a diagram of the vehicle-side architecture of the driverless intelligent bus system of the present invention; Figure 3 This is a roadside architecture diagram of the driverless intelligent bus system of the present invention; Figure 4 This is a system architecture diagram of the cloud platform for the driverless intelligent bus system of the present invention; Figure 5 This is a diagram of the station system architecture of the driverless intelligent bus system of the present invention; Figure 6 This is a diagram of the station system architecture of the driverless intelligent bus system of the present invention. Detailed Implementation

[0026] The exemplary embodiments of this disclosure are described below with reference to the accompanying drawings, including various details of the embodiments to aid understanding, and should be considered merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of this disclosure. Similarly, for clarity and brevity, descriptions of well-known functions and structures are omitted in the following description.

[0027] Example: Figure 1 or Figure 2 As shown, a five-dimensional collaborative control system for smart buses, encompassing vehicle-road-cloud-depot-station, includes five subsystems: vehicle-side equipment, roadside devices, cloud platform, depot system, and platform system. These five subsystems are interconnected through a pipeline layer, forming a complete end-to-end cloud collaborative architecture. The vehicle-side adopts a distributed network architecture, which is divided into three functional domains: power chassis domain, intelligent driving domain, and body control domain. The three functional domains communicate through three main CAN buses, with each network segment having a communication rate of 400-500kbps. Information forwarding and network segment isolation between the domains are achieved through a central gateway. The roadside equipment includes sensors, edge computing units, traffic signals, RSU devices, and timing devices, which are used to provide vehicles with a top-down perspective perception capability and solve problems related to blind spots, medium and long-distance perception, and traffic light perception. The cloud platform is the core of the system's scheduling and management, including basic service modules, scheduling service modules, diagnostic service modules, operation and maintenance service modules, and large-screen visualization modules; The station system interacts with the cloud platform in real time, receives dispatch instructions and provides feedback on vehicle status, and realizes fully automated management of the entire process of vehicle entry, charging, parking and exit. The platform system includes an information publishing module, a passenger flow monitoring module, an abnormal behavior monitoring module, an order access module, and a video storage module, serving as a crucial hub for interaction between people, stations, vehicles, and the platform.

[0028] In this embodiment, the powertrain chassis domain is responsible for vehicle chassis control, including the VCU and chassis actuators; the intelligent driving domain is responsible for the functions of the autonomous driving subsystem, remote driving subsystem, and V2X subsystem. The sensors of the intelligent driving domain are connected to the HAD through a private network segment. The OBU, RCU, and HAD are all connected to this network segment. When controlling the vehicle, the GW makes a comprehensive decision and outputs the results to the powertrain chassis network segment and the body network segment to ensure the uniqueness of the control interface of the vehicle's related components; the body control domain is responsible for the control of body accessories and / or human-machine interaction display comfort functions, which have low real-time requirements. The BCM is responsible for low-voltage power supply control, sliding door control, lighting control, horn control, and rear defrosting control. The air conditioning controller is responsible for air conditioning heating and ventilation, and the HMI controller is responsible for information display and interaction.

[0029] In this embodiment, the autonomous driving subsystem includes a perception module, a localization module, a map engine module, a decision planning module, and a control module. The perception module adopts a BEV+Transformer-based algorithm architecture, converting multi-channel visual sensor data into a bird's-eye view representation, and combining it with target detection and ROI point cloud segmentation based on point cloud deep learning. The localization module adopts a multi-source fusion localization scheme, combining GPS / IMU integrated navigation localization, laser point cloud feature localization, and visual feature localization. The decision planning module adopts a vehicle model-based integrated planning and control solution algorithm.

[0030] In this embodiment, the remote driving subsystem includes an on-board unit and a remote driving platform unit. It reuses the video data collected by the camera of the autonomous driving subsystem and distributes the video data through the autonomous driving controller for use by the remote driving application and the autonomous driving application, respectively.

[0031] In this embodiment, as Figure 3 As shown, the roadside equipment features a multi-sensor fusion layout: employing a multi-sensor fusion scheme combining lidar, millimeter-wave radar, and cameras, arranged at four corners or multiple points at intersections to eliminate blind spots and achieve full coverage perception at intersections; the roadside equipment also features high-precision time synchronization: the roadside equipment achieves time synchronization through GPS / BeiDou, using the same time source as the vehicle to ensure time consistency of vehicle-road perception data.

[0032] In this embodiment, the roadside equipment achieves high-throughput data transmission by employing an enhanced network transmission architecture, enabling the transmission of over 5700 bytes of data within 6ms, thus meeting the transmission requirements of dynamic environment and road boundary data in collaborative perception scenarios. The roadside equipment also features V2X collaborative perception: the vehicle-mounted C-V2X OBU connects the roadside and vehicle ends, enabling V2I / V2V / V2P / V2N functions, supporting customized warning information and perception data sharing, and achieving L4-level beyond-line-of-sight collaborative perception.

[0033] In this embodiment, as Figure 4 As shown, the basic service module is used for area management, vehicle registration, account management, role management, and system login; the dispatch service module is used for operation monitoring, operation analysis, order management, road condition monitoring, and operation scheduling; the diagnostic service module is used for fault analysis, abnormal event analysis, real-time vehicle video viewing, and risk prediction; the operation and maintenance service module is used for inspection management and work order scheduling; and the large-screen visualization module is used for 2D standard-size large-screen display to intuitively display core data such as operation status, road condition information, and fault warnings.

[0034] In this embodiment, as Figure 5 As shown, the station system includes an unmanned access control subsystem, an intelligent charging subsystem, an unmanned parking guidance subsystem, a station monitoring subsystem, and an IoT device monitoring subsystem.

[0035] In this embodiment, the unmanned access control subsystem is used for vehicles to automatically identify and enter / exit the station access control system via autonomous driving; the intelligent charging subsystem is used to support DC fast charging, with a charging time of less than 6 hours from 20% to 100% SOC, and to achieve automatic charging in conjunction with the vehicle's low battery alarm; the unmanned parking guidance subsystem is used for vehicles to automatically drive into designated parking spaces, realizing parking space planning; the station monitoring subsystem is used to monitor vehicles and the environment within the station in real time; and the IoT device monitoring subsystem is used for the unified management of various IoT devices within the station.

[0036] In this embodiment, as Figure 6 As shown, the information publishing module receives data from the cloud-based dispatching platform and displays vehicle trajectory, arrival time, and distance information in real time; the passenger flow monitoring module accesses passenger flow monitoring data within the platform, identifies human targets, tracks their movement trajectories, and supports forward, reverse, and bidirectional crossing counting and cumulative statistics; the abnormal behavior monitoring module monitors abnormal passenger behavior within the platform, automatically broadcasts abnormalities via voice, and returns the information to the background; the order access module receives passenger ride-hailing order information, feeds it back to the vehicle platform, and completes on-demand, real-time, and flexible dispatching; and the video storage module stores platform video for at least 720 hours with a resolution of at least 1920×1080.

[0037] In this embodiment, the vehicle terminal also includes a driving mode switching and safety redundancy mechanism: (1) Driving mode switching mechanism: The switching between autonomous driving mode and remote driving mode is realized through HMI and / or cloud backend. During the switching process, each system needs to detect its own status and obtain confirmation from the other before it can obtain control, ensuring that only one system has control at the same time. (2) Chassis system status monitoring: The automatic driving controller and / or remote driving controller monitor the VCU, EPS, EPB, ESC and / or Ebooster status of the chassis system in real time. When an abnormality is detected, the fault code is set and the corresponding fault handling mechanism is executed. (3) Intelligent driving system status monitoring: The chassis system monitors the operating status of the autonomous driving controller and / or remote driving controller in real time. When the perception and decision system is detected to be offline, the signal is abnormal (the counter and checksum do not change for ten consecutive cycles) or the fault is set (for ten consecutive cycles), each system enters the corresponding safety processing procedure.

[0038] In this embodiment, the activation of the autonomous driving system requires the following conditions to be met: within the coverage area of ​​the high-precision map, the driving route is selected manually on the in-vehicle HMI screen, the autonomous driving function is activated when the vehicle is static, and the autonomous driving system completes the response and starts within 2 seconds; after the system is powered on, it needs to complete a real-time self-check within 150 seconds, and the function is prohibited if the self-check is not completed.

[0039] In this embodiment, the station stopping function of the autonomous driving system is configured as follows: when the vehicle stops, the maximum distance from the inner side of the lane is no more than 0.5m, the maximum longitudinal deviation from the parking station is no more than 10m, the vehicle's tilt angle is no more than 15°, and the vehicle door opens within 3 seconds after the vehicle comes to a stop.

[0040] In this embodiment, the present invention also provides a control method for an unmanned intelligent bus, which uses the above-mentioned system and includes the following steps: S1. Vehicle departs from the depot: The depot system issues a depot command to the vehicle through the cloud platform, and the vehicle drives out of the depot automatically and exits the depot through an unmanned access control system. S2. Vehicle Road Operation: Vehicles drive autonomously on the road using an automated driving system. Roadside equipment provides beyond-line-of-sight perception data to the vehicles via V2X, and the cloud platform monitors the vehicle status in real time. S3. Vehicle entry and stop: When a vehicle arrives at the platform area, it will automatically stop and open and close the doors based on the passenger flow and order information fed back by the platform system. The platform system displays the vehicle arrival information in real time. S4. Vehicle return to depot: After completing its operational tasks, the vehicle will automatically return to the depot, enter through the unmanned access gate, and automatically drive into the designated parking space; S5. Intelligent charging: When the vehicle's battery level is below 20%, it will automatically send an alarm to the cloud. The cloud dispatch system will then arrange a charging task, and the vehicle will automatically drive into the charging position to complete the charging process. S6. Abnormal Handling: When the autonomous driving system malfunctions, it can be taken over by the remote driving system through the driving mode switching mechanism, or by the safety operator.

[0041] In summary, this invention not only achieves a fully unmanned closed loop from vehicle departure, road operation, platform parking, passenger flow interaction to depot return charging, but also effectively solves the problems of incomplete dimensions and information silos in the existing system architecture.

[0042] The specific embodiments described above do not constitute a limitation on the scope of protection of this disclosure. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this disclosure should be included within the scope of protection of this disclosure.

Claims

1. A five-dimensional collaborative control system for intelligent public transportation, comprising vehicle-road-cloud-site-station, characterized in that: It comprises five subsystems: vehicle-side equipment, roadside equipment, cloud platform, depot system, and platform system. These five subsystems are interconnected through a pipeline layer, forming a complete end-to-end cloud collaborative architecture. The vehicle-side adopts a distributed network architecture, which is divided into three functional domains: power chassis domain, intelligent driving domain, and body control domain. The three functional domains communicate through three main CAN buses, with each network segment having a communication rate of 400-500kbps. Information forwarding and network segment isolation between the domains are achieved through a central gateway. The roadside equipment includes sensors, edge computing units, traffic signals, RSU devices, and timing devices, which are used to provide vehicles with a top-down perspective perception capability and solve problems related to blind spots, medium and long-distance perception, and traffic light perception. The cloud platform is the core of the system's scheduling and management, including basic service modules, scheduling service modules, diagnostic service modules, operation and maintenance service modules, and large-screen visualization modules; The station system interacts with the cloud platform in real time, receives dispatch instructions and provides feedback on vehicle status, and realizes fully automated management of the entire process of vehicle entry, charging, parking and exit. The platform system includes an information publishing module, a passenger flow monitoring module, an abnormal behavior monitoring module, an order access module, and a video storage module, serving as a crucial hub for interaction between people, stations, vehicles, and the platform.

2. The five-dimensional collaborative control system for intelligent public transportation based on vehicle-road-cloud-site-station as described in claim 1, characterized in that: The powertrain chassis domain is responsible for vehicle chassis control, including the VCU and chassis actuators. The intelligent driving domain is responsible for the functions of the autonomous driving subsystem, remote driving subsystem, and V2X subsystem. The sensors in the intelligent driving domain are connected to the HAD through a private network segment. The OBU, RCU, and HAD are all connected to this network segment. When controlling the vehicle, the GW makes a comprehensive decision and outputs the results to the powertrain chassis network segment and the body network segment to ensure the uniqueness of the control interface of the vehicle's related components. The body control domain is responsible for the control of body accessories and / or human-machine interaction display comfort functions, which have low real-time requirements. The BCM is responsible for low-voltage power supply control, sliding door control, lighting control, horn control, and rear defrosting control. The air conditioning controller is responsible for air conditioning heating and ventilation, and the HMI controller is responsible for information display and interaction.

3. The five-dimensional collaborative control system for intelligent public transportation based on vehicle-road-cloud-site-station as described in claim 2, characterized in that: The autonomous driving subsystem includes a perception module, a localization module, a map engine module, a decision planning module, and a control module. The perception module adopts a BEV+Transformer-based algorithm architecture, which transforms multi-channel visual sensor data into a bird's-eye view representation and combines it with target detection and ROI point cloud segmentation based on point cloud deep learning. The localization module adopts a multi-source fusion localization scheme, which combines GPS / IMU integrated navigation localization, laser point cloud feature localization, and visual feature localization. The decision planning module adopts a vehicle model-based integrated planning and control solution algorithm.

4. The five-dimensional collaborative control system for intelligent public transportation based on vehicle-road-cloud-site-station as described in claim 2, characterized in that: The remote driving subsystem includes an on-board unit and a remote driving platform unit. It reuses video data collected by the cameras of the autonomous driving subsystem and distributes the video data through the autonomous driving controller for use by the remote driving application and the autonomous driving application, respectively.

5. The five-dimensional collaborative control system for intelligent public transportation based on vehicle-road-cloud-site-station as described in claim 1, characterized in that, The roadside equipment features a multi-sensor fusion layout: employing a fusion scheme of lidar, millimeter-wave radar, and cameras, arranged at four corners or multiple points at intersections to eliminate blind spots and achieve full coverage perception at intersections; the roadside equipment also features high-precision time synchronization: the roadside equipment achieves time synchronization through GPS / BeiDou, using the same time source as the vehicle to ensure time consistency of vehicle-road perception data.

6. The five-dimensional collaborative control system for intelligent public transportation based on vehicle-road-cloud-site-station as described in claim 1, characterized in that, The roadside equipment features high-throughput data transmission: it adopts an enhanced network transmission architecture to transmit more than 5700 bytes of data within 6ms, meeting the transmission requirements of dynamic environment and road boundary data in collaborative perception scenarios; the roadside equipment also features V2X collaborative perception: the vehicle-mounted C-V2X OBU connects the roadside and vehicle sides to realize V2I / V2V / V2P / V2N functions, supports customized warning information and perception data sharing, and achieves L4 level beyond-line-of-sight collaborative perception.

7. The five-dimensional collaborative control system for intelligent public transportation based on vehicle-road-cloud-site-station as described in claim 1, characterized in that, The basic service module is used for area management, vehicle registration, account management, role management and system login; the dispatch service module is used for operation monitoring, operation analysis, order management, road condition monitoring and operation dispatch scheduling. The diagnostic service module is used for fault analysis, abnormal event analysis, real-time vehicle video viewing, and risk prediction. The operation and maintenance service module is used for inspection management and work order scheduling; the large screen visualization module is used for 2D standard-size large screen display, which intuitively displays the core data of operation status, road condition information, and fault warning.

8. The five-dimensional collaborative control system for intelligent public transportation based on vehicle-road-cloud-site-station as described in claim 1, characterized in that: The station system includes an unmanned access control subsystem, an intelligent charging subsystem, an unmanned parking guidance subsystem, a station monitoring subsystem, and an IoT device monitoring subsystem.

9. The five-dimensional collaborative control system for intelligent public transportation based on vehicle-road-cloud-site-station as described in claim 8, characterized in that: The unmanned access control subsystem is used for vehicles to automatically identify and enter / exit the station access control system via autonomous driving; the intelligent charging subsystem supports DC fast charging, with a charging time of less than 6 hours from 20% to 100% SOC, and automatically charges in conjunction with the vehicle's low battery alarm; the unmanned parking guidance subsystem is used for vehicles to automatically drive into designated parking spaces, realizing parking space planning; the station monitoring subsystem is used for real-time monitoring of vehicles and the environment within the station; and the IoT device monitoring subsystem is used for unified management of various IoT devices within the station.

10. The five-dimensional collaborative control system for intelligent public transportation based on vehicle-road-cloud-site-station as described in claim 1, characterized in that: The information publishing module receives data from the cloud-based dispatching platform and displays vehicle trajectory, arrival time, and distance information in real time. The passenger flow monitoring module accesses passenger flow monitoring data within the platform, identifies human targets, tracks their movement, and supports forward, reverse, and bidirectional crossing counting and cumulative statistics. The abnormal behavior monitoring module monitors abnormal passenger behavior within the platform, automatically broadcasts an anomaly via voice, and returns the information to the backend. The order access module receives passenger booking order information, feeds it back to the vehicle platform, and enables on-demand, real-time, and flexible dispatching. The video storage module stores platform video for at least 720 hours with a resolution of at least 1920×1080.

Images