Roadside perception sharing system based on cross-network integration

Abstract

The present invention relates to a roadside perception sharing system based on cross-network integration, which belongs to the field of vehicle-road-cloud collaborative technology, and includes: a traffic object information processing subsystem including a message receiving module and a preprocessing module, the message receiving module is used to parse the traffic objects perceived by the roadside perception equipment and the traffic objects in the received perception sharing message and incorporate them into an object pool for unified management; the preprocessing module screens out objects with large errors through a weighting method based on time and position; the perception sharing message generation subsystem is used to extract traffic objects from the object pool, and generate perception sharing messages through judgment rules based on time and space transformation; the perception information distribution subsystem is used to distribute the perception sharing messages of the roadside equipment, and when receiving the perception sharing message distribution request, the perception sharing message priority is set based on the fusion of multiple factors such as the source, type and destination of the message, and then the method and order of sending the perception sharing message are selected.

Description

Technical Field

[0001] The present invention belongs to the field of vehicle-road-cloud collaboration technology and relates to a roadside perception sharing system based on cross-network integration. Background Art

[0002] Vehicle-road-cloud collaborative autonomous driving is the result of the coordinated development of the transportation, automobile, information and communication industries. The three major industries of transportation, automobile, information and communication are cross-integrated, mutually premised, mutually promoted and supported, forming a "smart car + smart road + integrated cloud" basic architecture, supported by high-precision maps, navigation and positioning and other industries, integrating key technologies such as information security, big data, and artificial intelligence, providing safe and efficient travel services for transportation applications, and forming a vehicle-road-cloud collaborative autonomous driving system. It is the only way to achieve high-level autonomous driving.

[0003] The industry has proposed the system architecture and technical roadmap for cellular vehicle-to-everything (C-V2X) (including LTE-V2X and its evolved NR-V2X). With the development of standards spearheaded by the International Organization for Standardization (3GPP), C-V2X has gained industry recognition for its technological advantages and scalability, surpassing competing technologies to become the de facto sole global standard for vehicle-to-everything (V2X) communications. Based on C-V2X technology, a complete industry ecosystem has emerged in China, encompassing V2X communication chips, modules, terminal devices, roadside equipment, testing and certification, and security services.

[0004] However, existing vehicle-road cooperative systems mostly adopt a fragmented approach and lack unified technical standards and protocols, resulting in compatibility and interoperability difficulties between different manufacturers and systems. For example, the information transmission protocols between on-board equipment, roadside units (RSUs), and cloud platforms are inconsistent, and the processing efficiency of perception sharing messages on each end is low, further increasing the delay in large-scale and complex traffic scenarios, reducing data reliability, and directly affecting the stability and real-time performance of the vehicle-road cooperative system. At the same time, the lack of uniformity in data formats and interface standards has seriously restricted the large-scale application and promotion of vehicle-road cooperative systems. These are all important reasons for the failure to fully share information between the human end, the vehicle end, and the road end, which will make it difficult for the system to achieve global optimization. Information such as traffic events and vehicle status cannot be shared and coordinated in real time, reducing the accuracy of roadside equipment's prediction of traffic conditions at the current intersection and the efficiency of traffic management. Summary of the Invention

[0005] In view of this, the purpose of the present invention is to address the problem of poor universality caused by the vehicle-road cooperative system's failure to effectively solve the collaborative sharing of perception on each end side of the vehicle-road cooperative system, and propose a human-vehicle-road-cloud cross-network perception sharing system architecture and method to solve the service architecture that solves the pain points of current vehicle networks and vehicle-road cooperative applications.

[0006] In order to achieve the above object, the present invention provides the following technical solutions:

[0007] A roadside perception sharing system based on cross-network integration, including a traffic object information processing subsystem, a perception sharing message generation subsystem, and a perception information distribution subsystem;

[0008] The traffic object information processing subsystem includes a message receiving module and a pre-processing module. The message receiving module is used to parse the traffic objects sensed by the roadside sensing device and the traffic objects in the received perception sharing message and incorporate them into the object pool for unified management; the pre-processing module uses a weighting method based on time and location to screen out objects with large errors;

[0009] The perception sharing message generation subsystem is used to extract traffic objects from the object pool and generate perception sharing messages through judgment rules based on time-space transformation;

[0010] The perception information distribution subsystem is used to distribute perception sharing messages from roadside equipment. When receiving a perception sharing message distribution request, it sets the perception sharing message priority based on a fusion of multiple factors such as the source, type, and destination of the message, and then selects the method and order for sending the perception sharing message.

[0011] Furthermore, the message receiving module uses dynamic and static message separation technology based on the object pool to manage traffic objects in the perception sharing message using different queues, scanning strategies, and priorities, distinguishing their dynamic and static attributes and designing different processing flows, specifically including:

[0012] When the roadside equipment receives the perception sharing message from different end sides, it scans and extracts the traffic object information in the object pool one by one, and analyzes their spatiotemporal information; first, it extracts their dynamic and static attributes. If they are static objects, they are added to the static object queue and sent to the message preprocessing module for processing after a certain period of time; if they are dynamic objects, they are directly sent to the message preprocessing module for processing.

[0013] Furthermore, after the object pool receives the perception results from the roadside perception device and the perception sharing messages from other end sides, the preprocessing module determines whether the difference between the timestamp of the roadside perception device and the timestamp of the perception sharing messages from other end sides is within the time deviation threshold range. If it is not within the time deviation threshold range, the perception sharing messages from the other end sides are discarded; if it is within the time deviation threshold range, the location attributes of the perception sharing messages from the other end sides continue to be extracted.

[0014] Furthermore, when the object pool receives the perception sharing message, the pre-processing module records the location of the source of the perception sharing message distribution request;

[0015] The perception sharing message generation subsystem determines the confidence of the perception sharing message by perceiving the GPS location of the sharing message source, combined with the source, location, current environment, publishing object, and perception accuracy of the perception sharing message request information; other perception sharing messages with higher confidence than the current perception sharing message are used as supplements to the current perception sharing message and do not participate in the generation of the perception sharing message.

[0016] Furthermore, the information generation method of the perception sharing message generation subsystem specifically includes the following steps:

[0017] Determine whether the perception shared message object in the object pool is a newly included object, and at the same time, based on the redundancy mitigation strategy of dynamic rules, pay attention to whether the change of its spatiotemporal information exceeds the set spatiotemporal threshold;

[0018] After being determined to be included in the generation of the perception sharing message, the dynamic objects therein are compensated and positioned.

[0019] Furthermore, the redundancy mitigation strategy based on dynamic rules is as follows: the object pool sets a threshold according to the busyness of the current traffic intersection; for historical objects already included in the object pool, their historical spatiotemporal information since the last time they were included in the perception sharing message generation is compared, and when their displacement change exceeds the displacement change threshold, or their speed change exceeds the speed change threshold, or they have exceeded the time change threshold since the last time they were included in the information generation, they are included in the generation of the perception sharing message; for data not included in the object pool, they are directly included in the generation of the perception sharing message.

[0020] Furthermore, dynamic objects can be compensated and positioned, including:

[0021] The difference between the timestamp of the perception sharing message after classification and the timestamp of the current moment is calculated and recorded as the offset compensation time Δt. The historical data of the traffic object in the object pool is then used as the prediction of the three-dimensional information after Δt. The position change vectors at adjacent moments are extracted using the recorded historical data (λ1, φ1, h1) and (λ2, φ2, h2):

[0022]

[0023] Normalize the position change vector to get the unit direction vector:

[0024]

[0025] Combined with the offset compensation time, the position change vector at the next moment is obtained:

[0026]

[0027] The position change vector at the next moment is used as the spatial compensation of the traffic object, added to the spatiotemporal information of the traffic object, and the message parameters are updated.

[0028] Furthermore, the perception information distribution subsystem adds a message header identifier to the perception sharing message according to the source, destination, and priority of the received perception sharing message distribution request, and classifies the perception sharing message distribution request into messages of different priorities according to its source; then, the perception sharing messages are filled into different network sending buffers according to the message header identifiers of different priorities, and then sent in sequence; the priority judgment rules include:

[0029] Special vehicles have higher priority than ordinary vehicles; moving objects have higher priority than stationary objects; request objects with a shorter distance from the current roadside equipment have higher priority than those with a longer distance; perception sharing message distribution requests sent by V2X communication have higher priority than perception sharing message distribution requests sent by cellular communication and wired communication.

[0030] The beneficial effects of the present invention are as follows: the present invention proposes a perception sharing architecture for the cross-network perception sharing demand, realizes the perception sharing and collaboration of people, vehicles, roads and clouds in a cross-network environment, and proposes a roadside perception fusion and sharing distribution strategy. The cross-network collaborative architecture of people, vehicles, roads and clouds can make full use of the currently laid infrastructure, can fully take care of intelligent vehicles, non-intelligent vehicles and pedestrians, and make full use of the hardware resources of mobile terminals, intelligent networked vehicles and roadside equipment. At the same time, the perception sharing collaboration method can send perception sharing messages to different users based on the request information of each end side. Users receive the latest information about traffic congestion, accidents, road conditions and navigation instructions, and can prepare in advance to improve traffic safety and reduce accident risks.

[0031] Other advantages, objects, and features of the present invention will be described in part in the following description and, in part, will be apparent to those skilled in the art upon examination of the following description or may be learned from practice of the present invention. The objects and other advantages of the present invention may be realized and obtained through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] In order to make the purpose, technical solutions and advantages of the present invention more clear, the present invention will be described in detail below with reference to the accompanying drawings, in which:

[0033] Figure 1 Design a diagram for the vehicle-road-cloud infrastructure;

[0034] Figure 2 This is a scene diagram for traffic intersection application;

[0035] Figure 3It is the structure diagram of traffic objects in the object pool;

[0036] Figure 4 This is the message preprocessing flow chart;

[0037] Figure 5 To add to the collaborative sensing information flow diagram. DETAILED DESCRIPTION

[0038] The following describes the embodiments of the present invention by means of specific examples, and those skilled in the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that the illustrations provided in the following embodiments are only schematic illustrations of the basic concept of the present invention, and the following embodiments and features in the embodiments can be combined with each other without conflict.

[0039] Among them, the accompanying drawings are only for illustrative purposes and represent only schematic diagrams rather than actual pictures, and should not be understood as limiting the present invention. In order to better illustrate the embodiments of the present invention, some parts of the accompanying drawings may be omitted, enlarged or reduced, and do not represent the dimensions of actual products. For those skilled in the art, it is understandable that some well-known structures and their descriptions may be omitted in the accompanying drawings.

[0040] The same or similar numbers in the drawings of the embodiments of the present invention correspond to the same or similar parts; in the description of the present invention, it should be understood that if there are terms such as "upper", "lower", "left", "right", "front", "back", etc. indicating directions or positional relationships, they are based on the directions or positional relationships shown in the drawings. They are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific direction, be constructed and operate in a specific direction. Therefore, the terms describing the positional relationship in the drawings are only used for illustrative purposes and cannot be understood as limiting the present invention. For ordinary technicians in this field, the specific meanings of the above terms can be understood according to specific circumstances.

[0041] like Figure 1As shown, the present invention first provides a perception sharing system architecture based on cross-network integration, including: perception information generation, routing, forwarding equipment and systems such as human-side, road-side, vehicle-side, and cloud-side. Among them, the human-side can upload traffic object information through a handheld smart device; the road-side device obtains the current traffic conditions at the intersection through the road-side perception device, and identifies and tracks the interested traffic object participants in real time, and generates perception sharing messages through the built-in algorithm; the cloud-side device can receive messages sent by the road-side, vehicle-side, and human-side for fusion processing, which can be used for front-end display and distribution of perception sharing messages to related systems and devices; a typical traffic intersection application scenario is as follows Figure 2 As shown in the figure, vehicles are divided into intelligent vehicles with V2X communication capabilities and ordinary vehicles. Intelligent vehicles obtain GPS location information and surrounding environmental information through a sensor network. When a vehicle detects an abnormal traffic environment, such as traffic congestion or a large number of pedestrians crossing an intersection, it first generates a collaborative information request message based on its GPS location, vehicle type, timestamp, and other information. This request message is then sent via V2X. Non-intelligent vehicles, on the other hand, can access the cloud server through their onboard smart devices to obtain perception sharing information about the current intersection.

[0042] When surrounding roadside devices receive this perception sharing message, they leverage the superior field of view and performance of the edge roadside devices to process it. First, the roadside device may receive more than one perception sharing message and prioritizes the corresponding one based on the priority of the request message. Since the roadside device receives messages from multiple traffic types simultaneously, it employs a filtering algorithm to filter out valid messages, eliminating outdated traffic objects outside of its perception area. The roadside device then generates a perception sharing message for the current area based on its built-in algorithm and finally transmits it based on the identified message header.

[0043] At the same time, people can access the 5G-V2X network in their current traffic area through their smart devices, receiving traffic data for the area. The smart devices can display the current traffic conditions, traffic incidents, and abnormal traffic objects in real time. They can also connect to the cloud to obtain traffic information from more distant locations, helping to plan their routes. Pedestrians can also upload information about their surroundings through their handheld devices, filling in blind spots that are not visible to roadside sensing equipment.

[0044] Secondly, the present invention provides a roadside perception sharing system based on cross-network integration, including a traffic object information processing subsystem, a perception sharing message generation subsystem and a perception information distribution subsystem.

[0045] The traffic object information processing subsystem includes a message receiving module and a preprocessing module. The message receiving module parses traffic objects sensed by roadside devices and received perception sharing messages and integrates them into an object pool for unified management. The preprocessing module preprocesses traffic objects in the object pool, selecting those within the same time threshold and filtering out objects with large errors based on time and location weights. The perception sharing message generation subsystem generates perception sharing messages. The perception information distribution subsystem distributes perception sharing messages from roadside devices.

[0046] Perception messages from different sources will be included in the object pool as traffic objects, where the structure of each object is as follows: Figure 3 As shown in the figure, the object ID, object location, and object status jointly mark a traffic object. The timestamp records the time of detection. The speed, acceleration, direction, and whether it is a historical traffic object are used as the basis for judging its spatiotemporal changes. Finally, the source of the perception result and the weight value are recorded. When the object pool uniformly manages traffic objects, a dynamic and static separation mechanism is adopted. This strategy uses different queues, scanning strategies, and priorities to manage traffic objects in the perception sharing message, distinguish their dynamic and static attributes, and design different processing flows to improve the efficiency of the roadside message processing system. Taking the traffic objects in the object pool as an example, according to the object status, if the object is a dynamic object, it will directly enter the next module; if it is a static object, the next step is determined by checking whether it is a historical object: if it is not a historical object, it is determined to be a new object and has a high priority for adding to the perception sharing message, and needs to be directly added to the generation of the perception sharing message; if it is a historical object, it is determined to be a low-priority traffic object and will be added to the static queue with other low-priority static objects. After a certain time threshold, the perception sharing message will be generated uniformly. The flow chart of its message processing is shown in the figure. Figure 4 shown.

[0047] The message pre-processing module will first use the timestamp of its own device as the main Then set it according to your own perception of speed Consider the received message as set M, the classification set as C, the time threshold as Δt, and extract the traffic object m in the set. i Timestamp t i , when the timestamp is in a new category c k When the message is within the range of , it is determined that the message has time validity; the formula is as follows:

[0048]

[0049] in

[0050] Then further extract the message set c k The spatiotemporal information of traffic objects in the data is divided into regions based on the location of the roadside equipment. Traffic objects in this region are considered to have spatial validity. If a message does not meet spatiotemporal validity requirements, it is discarded. If it does, the location attributes of the message are extracted. To ensure the accuracy of the combined results, data with excessive errors must be removed.

[0051] At a traffic intersection, multiple smart devices are detecting the current environment. Devices may receive repeated perception sharing messages for the same object. To avoid repeated perception of the same object, in the preprocessing module, the roadside device records the location of the source of the requested information upon receiving the perception sharing message. During the perception sharing message generation process, when receiving perception sharing data from other devices, the device determines the confidence level of the information based on the GPS location of the data source, the source of the perception sharing message request, its location, the current environment, the issuing object, and the perception accuracy. The calculated perception accuracy is then added to the perception sharing message distribution request as the basis for determining the accuracy of the traffic object. The perception confidence level is calculated as follows:

[0052]

[0053] Among them, w represents the weight of the message, d is the distance from the sensing device to the sensing target, and w d represents the weight of the distance influencing factor, a represents the accuracy of the perception device in identifying the object, and w ac Represents the weight of recognition accuracy, r represents the influence of the current environment, w r is the weight of environmental impact, w o If the traffic object perception data comes from an authoritative organization, its confidence level will be relatively higher. After obtaining the weight value for each traffic object, when generating a perception sharing message, if the confidence level of a traffic object in another perception sharing message is higher than the confidence level of the roadside device's own perception results, the object will not be included in the perception sharing message generation, thus avoiding the duplication of multiple perception sharing messages.

[0054] The specific steps of the preprocessing module are as follows:

[0055] A1: The roadside receives perception sharing messages from itself and other end-sides.

[0056] A2: Determine whether the message is expired based on the message time attribute and the valid time value. If the message is expired, the message event is considered an invalid event.

[0057] A3: Perform spatial validity judgment on the perception sharing message, including judging the effective distance, message flow direction, and message forwarding of the perception sharing message, and deleting the perception sharing message that does not meet the processing requirements.

[0058] A4: Extract the traffic object from the message, initialize it according to the traffic object format, and preliminarily determine whether it is a dynamic object.

[0059] A5: Whether to proceed to the next generation step is determined based on the confidence level of the traffic object. Data that is not perceived by the roadside itself is added to the message generation queue based on the confidence level. If the confidence level of the perceived traffic object is relatively higher, the next step is entered. If the confidence level is lower than that of other messages, the high-confidence message is used to supplement the details of the traffic object itself, but will not enter the next generation step.

[0060] A6: Dynamic objects use a clustering algorithm to classify information about the same object at different time thresholds, remove data with large errors, and calculate the average. Static objects enter a static queue and generate perception sharing messages uniformly after a certain time threshold.

[0061] A7: The dynamic object enters the queue and is available for use by the next process.

[0062] In order to reduce the data size of the perception sharing message, the pre-processing module sets a judgment standard for the generation of the perception sharing message, which is called the collaborative information generation rule, and builds a redundancy mitigation strategy based on the dynamic rule: if the device wants to add the traffic object to the new perception sharing message, it must check the position and speed change of the traffic object. If the absolute position change of the participant exceeds the displacement threshold compared to the last time it was added to the collaborative information, or its speed change exceeds the speed change threshold, or its time difference exceeds the time change threshold compared to the last time it was added to the collaborative information, then the traffic object will be included in the generation of the perception sharing message; and data not included in the object pool will be directly included in the generation of the perception sharing message. The flowchart for adding perception objects is as follows: Figure 5 shown.

[0063] The perception sharing message generation subsystem is used to extract traffic objects from the object pool. Its purpose is to aggregate the perception results of the same traffic object on different ends in the cross-network perception sharing message. The subsystem classifies messages based on whether the spatiotemporal transformation of the perception sharing message exceeds the threshold set by the system, thereby filtering out data that is not within the current spatiotemporal scope. The subsystem analyzes the quality of the perception results by checking the confidence of the perception sharing messages on different ends, avoiding repeated broadcasts of the same perception sharing message.

[0064] Due to the time offset caused by the computational effort in the generation of perception sharing messages, a perception sharing location and time offset compensation positioning method needs to be designed based on the spatiotemporal information estimation of traffic objects when the information is generated. Offset compensation is performed on each dynamic object in the perception sharing message generation list. The difference between the timestamp of the vehicle in the set after vehicle perception fusion and the timestamp of the current moment is calculated and recorded as the offset compensation time Δt. Then, the position change vector at the adjacent moment is extracted through the recorded historical data. The position change vector is normalized to obtain a unit vector and then combined with the offset compensation time to obtain the position change vector at the next moment. This position change vector will be used as the spatial compensation of the traffic object. Its specific description equation is:

[0065] The historical location data is:

[0066] (λ1,φ1,h1),(λ2,φ2,h2)

[0067] Calculate the position change vector at adjacent moments:

[0068]

[0069] Normalize the position change vector to get the unit direction vector:

[0070]

[0071] Use the current speed and unit direction vector to predict the position change vector at the next moment

[0072]

[0073] Add the position change vector to the current position to get the position at the next moment:

[0074]

[0075] After the calculation is completed, the position change vector is added to the spatiotemporal information of the traffic object as spatial compensation, and then the next step is entered.

[0076] The steps of the perception sharing message generation subsystem are as follows:

[0077] B1: When receiving the shared perception message dispatch request, it starts to read the object from the dynamic queue;

[0078] B2: Based on the historical information of the dynamic object, determine whether its spatiotemporal information changes exceed the range set by the algorithm. If so, add it to the message generation queue;

[0079] B3: Compensate for dynamic objects in terms of time and space;

[0080] B4: Read the time information of the static data and determine whether the set time threshold has been exceeded. If it has been exceeded, all static objects are added to the message generation queue. If not, continue to remain silent.

[0081] The perception sharing message distribution subsystem adds a message header identifier to the perception sharing message based on the source, destination, and priority of the received perception sharing message distribution request. First, the source of the message is determined, with special vehicles taking precedence over ordinary vehicles; moving objects taking precedence over stationary objects; and requesting objects that are closer to the current roadside equipment taking precedence over those farther away. Secondly, the message source is analyzed, with the perception sharing message distribution request sent via V2X communication set to a higher priority to meet V2X's real-time requirements. Request messages sent via cellular or wired communication will be set to a lower priority, and the sending subsystem will prioritize high-priority messages. When the sending subsystem begins operation, it determines the availability of all network resources, then establishes a resource manager to manage all network resources, stores the corresponding identification files, and, based on the message header identifier, populates the perception sharing message body into different network sending buffers before sending it out.

[0082] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not limiting. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention can be modified or replaced by equivalents without departing from the purpose and scope of the technical solutions, which should all be included in the scope of the claims of the present invention.

Claims

1. A roadside perception sharing system based on cross-network integration, characterized by: It includes traffic object information processing subsystem, perception sharing message generation subsystem and perception information distribution subsystem; The traffic object information processing subsystem includes a message receiving module and a pre-processing module. The message receiving module is used to parse the traffic objects sensed by the roadside sensing device and the traffic objects in the received perception sharing message and incorporate them into the object pool for unified management; the pre-processing module uses a weighting method based on time and location to screen out objects with large errors; The perception sharing message generation subsystem is used to extract traffic objects from the object pool and generate perception sharing messages through judgment rules based on time-space transformation; The perception information distribution subsystem is used to distribute perception sharing messages from roadside equipment. When receiving a perception sharing message distribution request, it sets the perception sharing message priority based on a fusion of multiple factors such as the source, type, and destination of the message, and then selects the method and order for sending the perception sharing message.

2. The roadside perception sharing system based on cross-network integration according to claim 1 is characterized by: The message receiving module manages traffic objects in the perception sharing message based on the dynamic and static message isolation management strategy, adopts different queues, scanning strategies and priorities, distinguishes their dynamic and static attributes, and designs different processing processes, including: When the roadside equipment receives the perception sharing message from different end sides, it scans and extracts the traffic object information in the object pool one by one, and analyzes their spatiotemporal information; first, it extracts their dynamic and static attributes. If they are static objects, they are added to the static object queue and sent to the message preprocessing module for processing after a certain period of time; if they are dynamic objects, they are directly sent to the message preprocessing module for processing.

3. The roadside perception sharing system based on cross-network integration according to claim 1 is characterized by: After the object pool receives the perception results from the roadside perception device and the perception sharing messages from other end sides, the preprocessing module determines whether the difference between the timestamp of the roadside perception device and the timestamp of the perception sharing messages from other end sides is within a time deviation threshold range. If it is not within the time deviation threshold range, the perception sharing messages from the other end sides are discarded; if it is within the time deviation threshold range, the location attributes of the perception sharing messages from the other end sides continue to be extracted.

4. The roadside perception sharing system based on cross-network integration according to claim 1 is characterized by: The pre-processing module records the location of the source of the perception sharing message distribution request when the object pool receives the perception sharing message; The perception sharing message generation subsystem determines the confidence level of the perception sharing message by sensing the GPS location of the sharing message source, combining the source, location, current environment, publishing object, and perception accuracy of the perception sharing message request information; Other perception sharing messages with higher confidence than the current perception sharing message are used as supplements to the current perception sharing message and do not participate in the generation of the perception sharing message.

5. The roadside perception sharing system based on cross-network integration according to claim 1 is characterized by: The information generation method of the perception sharing message generation subsystem specifically includes the following steps: Determine whether the perception shared message object in the object pool is a newly included object, and at the same time, based on the redundancy mitigation strategy of dynamic rules, pay attention to whether the change of its spatiotemporal information exceeds the set spatiotemporal threshold; After being determined to be included in the generation of the perception sharing message, the dynamic objects therein are compensated and positioned.

6. The roadside perception sharing system based on cross-network integration according to claim 5 is characterized by: The dynamic rule-based redundancy mitigation strategy is as follows: the object pool sets a threshold based on the busyness of the current traffic intersection; for historical objects already included in the object pool, their historical spatiotemporal information since the last time they were included in the perception sharing message generation is compared, and when their displacement change exceeds the displacement change threshold, or their speed change exceeds the speed change threshold, or they have exceeded the time change threshold since the last time they were included in the information generation, they are included in the generation of the perception sharing message; for data not included in the object pool, they are directly included in the generation of the perception sharing message.

7. The roadside perception sharing system based on cross-network integration according to claim 5 is characterized by: Compensate and position dynamic objects, including: The difference between the timestamp of the perception sharing message after classification and the timestamp of the current moment is calculated and recorded as the offset compensation time Δt. The historical data of the traffic object in the object pool is then used as the prediction of the three-dimensional information after Δt. The position change vectors at adjacent moments are extracted using the recorded historical data (λ1, φ1, h1) and (λ2, φ2, h2): Normalize the position change vector to get the unit direction vector: Combined with the offset compensation time, the position change vector at the next moment is obtained: The position change vector at the next moment is used as the spatial compensation of the traffic object, added to the spatiotemporal information of the traffic object, and the message parameters are updated.

8. The roadside perception sharing system based on cross-network integration according to claim 1 is characterized by: The perception information distribution subsystem adds a message header identifier to the perception sharing message according to the source, purpose and priority of the received perception sharing message distribution request, and classifies the perception sharing message distribution request into messages of different priorities according to its source; Then, the perception sharing messages are filled into different network sending buffers according to the message header identifiers of different priorities, and then sent in sequence; the priority judgment rules include: Special vehicles have higher priority than ordinary vehicles; moving objects have higher priority than stationary objects; request objects with a shorter distance from the current roadside equipment have higher priority than those with a longer distance; perception sharing message distribution requests sent by V2X communication have higher priority than perception sharing message distribution requests sent by cellular communication and wired communication.

Images