A vehicle-road cooperative data processing method and related equipment

Abstract

The present invention discloses a vehicle-road cooperative data processing method and related equipment, which relate to the field of communication technology, in order to solve the problem in related technologies that the security and reliability of roadside perception data are poor, thereby affecting the security and reliability of vehicle-road cooperative information services. The method includes: obtaining N groups of roadside perception data of the target road section; obtaining M groups of basic safety message BSM data of the target road section; making a confidence judgment on the i-th group of roadside perception data based on the i-th group of roadside perception data and the M groups of BSM data, where i takes values from 1 to N in sequence; outputting K groups of roadside perception data whose confidence meets the requirements among the N groups of roadside perception data, where K is a positive integer less than or equal to N; providing vehicle-road cooperative information services to traffic objects on the target road section based on the traffic data on the target road section, where the traffic data includes K groups of roadside perception data. The embodiments of the present invention can improve the security and reliability of roadside perception data, thereby ensuring the security and reliability of vehicle-road cooperative information services.

Description

Technical Field

[0001] The present invention relates to the field of communication technologies, and in particular to a vehicle-road collaborative data processing method and related equipment. Background Art

[0002] In related technologies, most vehicle-infrastructure cooperative systems use cameras, lidar, millimeter-wave radar, and other roadside sensing devices to obtain roadside perception data. These data can also be combined with vehicle-side perception data to obtain more accurate road traffic information for analysis and use. In real-world traffic scenarios, the perception data from these devices is unstable and, in some cases, can differ significantly from the actual data. Related technologies currently lack a solution for assessing the confidence of roadside or vehicle-side perception data, which in turn affects the safety and reliability of vehicle-infrastructure cooperative information services. Summary of the Invention

[0003] An embodiment of the present invention provides a vehicle-road cooperative data processing method and related equipment to solve the problem in related technologies that the security and reliability of roadside perception data are poor, which in turn affects the security and reliability of vehicle-road cooperative information services.

[0004] In a first aspect, an embodiment of the present invention provides a method for processing vehicle-road cooperative data, including:

[0005] Obtain N sets of roadside sensing data for the target road section, where N is a positive integer;

[0006] Obtaining M groups of BSM data (Basic Safety Message) for the target road section, where M is a positive integer;

[0007] Performing a confidence judgment on the i-th group of roadside sensing data based on the i-th group of roadside sensing data and the M groups of BSM data, where i takes values from 1 to N in sequence;

[0008] Outputting K groups of roadside perception data whose confidence meets the requirement among the N groups of roadside perception data, where K is a positive integer less than or equal to N;

[0009] Based on the traffic data on the target road section, a vehicle-road collaborative information service is provided to traffic objects on the target road section, wherein the traffic data includes the K groups of roadside perception data.

[0010] Optionally, the performing confidence judgment on the i-th group of roadside perception data based on the i-th group of roadside perception data and the M groups of BSM data includes:

[0011] Calculating the confidence level of the i-th group of roadside perception data based on the i-th group of roadside perception data and the M groups of BSM data;

[0012] When the confidence level of the i-th group of roadside perception data is within a preset range, it is determined that the confidence level of the i-th group of roadside perception data meets the requirement.

[0013] Optionally, the preset range is greater than or equal to the confidence threshold and less than or equal to 1;

[0014] There are multiple groups of roadside sensing devices arranged according to their positions on the target road section, each group of roadside sensing data is collected by a group of roadside sensing devices on the target road section, and the confidence threshold is calculated based on the multiple groups of roadside sensing data collected by the multiple groups of roadside sensing devices within a historical time period, and the multiple groups of BSM data on the target road section within the historical time period.

[0015] Optionally, calculating the confidence level of the i-th group of roadside perception data based on the i-th group of roadside perception data and the M groups of BSM data includes:

[0016] Calculating M confidence levels of the i-th group of roadside perception data based on the i-th group of roadside perception data and each group of BSM data in the M groups of BSM data;

[0017] When the confidence level of the i-th group of roadside perception data is within a preset range, determining that the confidence level of the i-th group of roadside perception data meets the requirement includes:

[0018] If less than L confidence levels among the M confidence levels are not within the preset range, determining that the confidence level of the i-th group of roadside sensing data meets the requirement, where L is a preset value and L is a positive integer;

[0019] After calculating the M confidence levels of the i-th group of roadside perception data, the method further includes:

[0020] If more than L of the M confidence levels are outside the preset range, it is determined that the confidence level of the i-th group of roadside perception data does not meet the requirement, and output of the i-th group of roadside perception data is stopped.

[0021] Optionally, after calculating M confidence levels of the i-th group of roadside perception data, the method further includes:

[0022] When it is calculated based on the i-th group of roadside perception data and the j-th group of BSM data that the j-th confidence level of the i-th group of roadside perception data is not within the preset range, and no more than L confidence levels among the M confidence levels are not within the preset range, it is determined that the confidence level of the j-th group of BSM data does not meet the requirements, and a warning message is sent to the target vehicle-mounted terminal corresponding to the j-th group of BSM data, where j is an integer between 1 and M.

[0023] Optionally, after determining that the confidence level of the i-th group of roadside perception data does not meet the requirement and stopping outputting the i-th group of roadside perception data, the method further includes:

[0024] The i-th group of roadside perception data is input into a pre-established confidence correction model to obtain the i-th group of roadside perception data after correction output by the confidence correction model.

[0025] Optionally, after inputting the i-th group of roadside perception data into a pre-established confidence correction model to obtain the corrected i-th group of roadside perception data output by the confidence correction model, the method further includes:

[0026] Obtaining M′ groups of BSM data for the target road section, where M′ is a positive integer;

[0027] Calculating M′ confidence levels of the i-th group of roadside perception data based on the corrected i-th group of roadside perception data and each group of BSM data in the M′ groups of BSM data;

[0028] When more than L confidence levels among the M′ confidence levels are within the preset range, it is determined that the confidence level of the corrected i-th group of roadside perception data meets the requirement, and the corrected i-th group of roadside perception data is output.

[0029] Optionally, the confidence correction model is Y=c0X 5 +d0X 4 +e0X 3 +f0X 2 +g0X+h0, where X is the roadside perception data before correction, Y is the roadside perception data after correction, and c0, d0, e0, f0, g0, and h0 are model parameters.

[0030] Optionally, after determining that the confidence level of the i-th group of roadside perception data does not meet the requirement and stopping outputting the i-th group of roadside perception data, the method further includes:

[0031] Performing a confidence judgment on the target roadside sensing device corresponding to the i-th group of roadside sensing data;

[0032] When the confidence level of the target roadside perception device does not meet the requirement, the roadside perception data collected by the target roadside perception device is corrected.

[0033] Optionally, multiple groups of roadside sensing devices are arranged on the target road section according to their positions, each group of roadside sensing data is collected by a group of roadside sensing devices on the target road section, a group of roadside sensing devices includes multiple roadside sensors, and each group of roadside sensing data includes multiple sensor sensing data;

[0034] After acquiring N groups of roadside perception data of the target road section, and before performing confidence judgment on the i-th group of roadside perception data based on the i-th group of roadside perception data and the M groups of BSM data, the method further includes:

[0035] The plurality of sensor perception data in the i-th group of roadside perception data among the N groups of roadside perception data are fused to obtain the i-th group of fused roadside perception data.

[0036] Optionally, after outputting K groups of roadside perception data whose confidences meet the requirements among the N groups of roadside perception data, the method further includes:

[0037] When K is greater than 1, the K groups of roadside perception data are fused to obtain global roadside fused perception data;

[0038] The traffic data includes the global roadside fusion perception data.

[0039] Optionally, after fusing the K groups of roadside sensing data, the method further includes:

[0040] Fusing the M groups of BSM data with the global roadside fusion perception data to obtain first fused traffic data;

[0041] The traffic data includes the first fused traffic data.

[0042] Optionally, the method further includes:

[0043] Obtain G groups of SSM data (Sensor Sharing Message, perception sharing message) of the target road section, where G is a positive integer;

[0044] After fusing the M groups of BSM data with the global roadside fusion perception data, the method further includes:

[0045] fusing the G group of SSM data with the first fused traffic data to obtain second fused traffic data;

[0046] The traffic data includes the second fused traffic data.

[0047] Optionally, providing a vehicle-road cooperative information service to traffic objects on the target road section based on the traffic data on the target road section includes:

[0048] determining object information and motion data of traffic objects on the target road segment based on traffic data on the target road segment;

[0049] Based on the object information and motion data of the traffic object, the motion trajectory of the traffic object is predicted, or the risk rating of the traffic object is performed.

[0050] Optionally, after fusing the G group SSM data with the first fused traffic data to obtain second fused traffic data, the method further includes:

[0051] Uploading the second fused traffic data to the central cloud;

[0052] Receiving traffic information of the target road section sent by the central cloud;

[0053] The traffic information is sent to traffic objects on the target road section.

[0054] In a second aspect, an embodiment of the present invention further provides another method for processing vehicle-road cooperative data, including:

[0055] Receive second fused traffic data sent by the edge cloud, where the second fused traffic data is obtained by fusing K groups of roadside perception data that meet confidence requirements, M groups of BSM data, and G groups of SSM data for the target road section, where K is an integer greater than 1, and M and G are both positive integers;

[0056] determining traffic information of the target road section based on the second fused traffic data;

[0057] Send the traffic information to the edge cloud.

[0058] In a third aspect, an embodiment of the present invention further provides an edge cloud, including:

[0059] A first acquisition module is used to acquire N groups of roadside sensing data of a target road section, where N is a positive integer;

[0060] A second acquisition module is configured to acquire M groups of basic safety message (BSM) data of the target road section, where M is a positive integer;

[0061] A first processing module is configured to perform a confidence judgment on the i-th group of roadside sensing data based on the i-th group of roadside sensing data and the M groups of BSM data, where i is 1 to N in sequence;

[0062] an output module, configured to output K groups of roadside perception data whose confidences meet the requirements among the N groups of roadside perception data, where K is a positive integer less than or equal to N;

[0063] An execution module is used to provide vehicle-road collaborative information services to traffic objects on the target road section based on the traffic data on the target road section, wherein the traffic data includes the K groups of roadside perception data.

[0064] Optionally, the first processing module includes:

[0065] a calculation submodule, configured to calculate the confidence level of the i-th group of roadside perception data based on the i-th group of roadside perception data and the M groups of BSM data;

[0066] The first determination submodule is configured to determine whether the confidence level of the i-th group of roadside perception data meets the requirement when the confidence level of the i-th group of roadside perception data is within a preset range.

[0067] Optionally, the preset range is greater than or equal to the confidence threshold and less than or equal to 1;

[0068] There are multiple groups of roadside sensing devices arranged according to their positions on the target road section, each group of roadside sensing data is collected by a group of roadside sensing devices on the target road section, and the confidence threshold is calculated based on the multiple groups of roadside sensing data collected by the multiple groups of roadside sensing devices within a historical time period, and the multiple groups of BSM data on the target road section within the historical time period.

[0069] Optionally, the calculation submodule includes:

[0070] a calculation unit, configured to calculate M confidence levels of the i-th group of roadside perception data based on the i-th group of roadside perception data and each group of BSM data in the M groups of BSM data;

[0071] The first determination submodule is configured to determine that the confidence level of the i-th group of roadside sensing data meets the requirement when less than L confidence levels among the M confidence levels are not within the preset range, where L is a preset value and is a positive integer;

[0072] The edge cloud further includes:

[0073] The second processing module is used to determine that the confidence level of the i-th group of roadside perception data does not meet the requirements and stop outputting the i-th group of roadside perception data when more than L confidence levels among the M confidence levels are not within the preset range.

[0074] Optionally, the edge cloud further includes:

[0075] The third processing module is used to determine that the confidence level of the j-th group of BSM data does not meet the requirements when calculating, based on the i-th group of roadside perception data and the j-th group of BSM data, that the j-th confidence level of the i-th group of roadside perception data is not within the preset range, and no more than L confidence levels among the M confidence levels are not within the preset range, and send a warning message to the target vehicle-mounted terminal corresponding to the j-th group of BSM data, where j is an integer between 1 and M.

[0076] Optionally, the edge cloud further includes:

[0077] The first correction module is used to input the i-th group of roadside perception data into a pre-established confidence correction model to obtain the corrected i-th group of roadside perception data output by the confidence correction model.

[0078] Optionally, the edge cloud further includes:

[0079] A third acquisition module is configured to acquire M′ groups of BSM data of the target road section, where M′ is a positive integer;

[0080] a calculation module, configured to calculate M′ confidence levels of the i-th group of roadside perception data based on the corrected i-th group of roadside perception data and each group of BSM data in the M′ groups of BSM data;

[0081] The fourth processing module is used to determine that the confidence level of the corrected i-th group of roadside perception data meets the requirements when more than L confidence levels among the M′ confidence levels are within the preset range, and output the corrected i-th group of roadside perception data.

[0082] Optionally, the confidence correction model is Y=c0X 5 +d0X 4 +e0X 3 +f0X 2 +g0X+h0, where X is the roadside perception data before correction, Y is the roadside perception data after correction, and c0, d0, e0, f0, g0, and h0 are model parameters.

[0083] Optionally, the edge cloud further includes:

[0084] a fifth processing module, configured to perform a confidence judgment on the target roadside sensing device corresponding to the i-th group of roadside sensing data;

[0085] The second correction module is used to correct the roadside perception data collected by the target roadside perception device when the confidence level of the target roadside perception device does not meet the requirements.

[0086] Optionally, multiple groups of roadside sensing devices are arranged on the target road section according to their positions, each group of roadside sensing data is collected by a group of roadside sensing devices on the target road section, a group of roadside sensing devices includes multiple roadside sensors, and each group of roadside sensing data includes multiple sensor sensing data;

[0087] The edge cloud further includes:

[0088] The first fusion module is used to fuse multiple sensor perception data in the i-th group of roadside perception data among the N groups of roadside perception data to obtain the fused i-th group of roadside perception data.

[0089] Optionally, the edge cloud further includes:

[0090] A second fusion module is configured to fuse the K groups of roadside perception data to obtain global roadside fusion perception data when K is greater than 1;

[0091] The traffic data includes the global roadside fusion perception data.

[0092] Optionally, the edge cloud further includes:

[0093] a third fusion module, configured to fuse the M groups of BSM data with the global roadside fusion perception data to obtain first fused traffic data;

[0094] The traffic data includes the first fused traffic data.

[0095] Optionally, the edge cloud further includes:

[0096] A fourth acquisition module is used to obtain G groups of SSM data of the target road section, where G is a positive integer;

[0097] a fourth fusion module, configured to fuse the G group SSM data with the first fused traffic data to obtain second fused traffic data;

[0098] The traffic data includes the second fused traffic data.

[0099] Optionally, the execution module includes:

[0100] a second determining submodule, configured to determine object information and motion data of traffic objects on the target road section based on the traffic data on the target road section;

[0101] The processing submodule is used to predict the motion trajectory of the traffic object based on the object information and motion data of the traffic object, or to perform risk rating on the traffic object.

[0102] Optionally, the edge cloud further includes:

[0103] A second sending module, configured to upload the second fused traffic data to a central cloud;

[0104] A second receiving module is used to receive the traffic information of the target road section sent by the central cloud;

[0105] The third sending module is used to send the traffic information to the traffic objects on the target road section.

[0106] In a fourth aspect, an embodiment of the present invention further provides a central cloud, including:

[0107] A first receiving module is configured to receive second fused traffic data sent by the edge cloud, wherein the second fused traffic data is obtained by fusing K groups of roadside perception data, M groups of BSM data, and G groups of SSM data of the target road section, whose confidence levels meet requirements, where K is an integer greater than 1, and M and G are both positive integers;

[0108] a determination module, configured to determine traffic information of the target road section based on the second fused traffic data;

[0109] The first sending module is used to send the traffic information to the edge cloud.

[0110] In the fifth aspect, an embodiment of the present invention also provides a network side device, including: a transceiver, a memory, a processor, and a computer program stored in the memory and runnable on the processor. When the processor executes the computer program, it implements the steps in the vehicle-road collaborative data processing method described in the first aspect above; or implements the steps in the vehicle-road collaborative data processing method described in the second aspect above.

[0111] In the sixth aspect, an embodiment of the present invention also provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, it implements the steps in the vehicle-road collaborative data processing method as described in the first aspect above; or implements the steps in the vehicle-road collaborative data processing method as described in the second aspect above.

[0112] In an embodiment of the present invention, N groups of roadside perception data are obtained for a target road section, where N is a positive integer; M groups of basic safety message (BSM) data are obtained for the target road section, where M is a positive integer; a confidence judgment is performed on the i-th group of roadside perception data based on the i-th group of roadside perception data and the M groups of BSM data, where i sequentially takes values from 1 to N; K groups of roadside perception data with confidence levels that meet requirements are output from the N groups of roadside perception data, where K is a positive integer less than or equal to N; and a vehicle-road cooperative information service is provided to traffic objects on the target road section based on traffic data on the target road section, where the traffic data includes the K groups of roadside perception data. In this way, by performing confidence judgment on the multiple groups of roadside perception data obtained for the target road section, and providing the vehicle-road cooperative information service to traffic objects on the target road section based on the roadside perception data with confidence levels that meet requirements, the security and reliability of the roadside perception data can be improved, thereby ensuring the security and reliability of the vehicle-road cooperative information service. BRIEF DESCRIPTION OF THE DRAWINGS

[0113] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in describing the embodiments of the present invention. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0114] Figure 1 This is one of the flow charts of the vehicle-road cooperative data processing method provided by an embodiment of the present invention;

[0115] Figure 2 This is an example flow chart of a method for processing vehicle-road cooperative data according to an embodiment of the present invention;

[0116] Figure 3 Schematic diagram of a vehicle-infrastructure cooperative system architecture to which the vehicle-infrastructure cooperative data processing method provided in an embodiment of the present invention can be applied;

[0117] Figure 4 This is the second flowchart of the vehicle-road cooperative data processing method provided by an embodiment of the present invention;

[0118] Figure 5 This is a structural diagram of the edge cloud provided by an embodiment of the present invention;

[0119] Figure 6 This is a structural diagram of the central cloud provided by an embodiment of the present invention;

[0120] Figure 7 It is a structural diagram of the network side device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0121] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of them. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0122] See also Figure 1 , Figure 1 is a flow chart of the vehicle-road cooperative data processing method provided by an embodiment of the present invention, such as Figure 1 As shown, the following steps are included:

[0123] Step 101: Obtain N sets of roadside sensing data of a target road section, where N is a positive integer.

[0124] The above-mentioned target road section can be any road section where traffic data needs to be collected to provide vehicle-road collaborative information services in actual applications. For example, when it is necessary to provide dynamic traffic information services to vehicles, road sections, pedestrians and other traffic participants in a certain area, and provide vehicle-road collaborative information services to vehicles, pedestrians and other traffic objects on the road section, this road section is the target road section.

[0125] The method of this embodiment can be executed by an edge cloud arranged at the edge side of the network, and it is an edge cloud corresponding to the target road section, that is, the coverage of the edge cloud includes the target road section, so that the roadside perception data of the target road section can be obtained.

[0126] In an embodiment of the present invention, the target road segment may be equipped with roadside sensing devices for collecting roadside sensing data. This roadside sensing data is information collected about traffic objects on the target road segment, such as vehicles, pedestrians, and obstacles. Based on this roadside sensing data, relevant data about the traffic objects on the target road segment, such as their location, size, speed, and heading angle, can be determined. A single roadside sensing device can be, for example, a camera, a lidar, or a millimeter-wave radar. The roadside sensing devices on the target road segment can be arranged in groups. A group of roadside sensing devices can consist of several roadside sensing devices, for example, one camera and one lidar, two cameras and one lidar, one camera and one millimeter-wave radar, and so on. A group of roadside sensing devices can cover a portion of the target road segment. That is, multiple groups of roadside sensing devices can be distributed along the target road segment, and these multiple groups can cover the entire target road segment.

[0127] The aforementioned acquisition of N sets of roadside perception data for the target road section may be performed by acquiring roadside perception data collected by several sets of roadside perception devices on the target road section, thereby obtaining the N sets of roadside perception data, i.e., N is related to the number of roadside perception devices on the target road section, and one set of roadside perception data can be collected by one set of roadside perception devices, and one set of roadside perception data includes the perception data of several roadside perception devices. Specifically, the several sets of roadside perception devices on the target road section may upload their collected roadside perception data to an RSU (Road Side Unit) or a base station, and the RSU or base station then uploads the roadside perception data of each roadside perception device to the edge cloud, such as via the PC5 port of a cellular vehicle-to-everything (C-V2X) network or the Uu port of a 5G network, so that the edge cloud can receive the N sets of roadside perception data corresponding to the target road section. i.e., the RSU can establish a highly dynamic communication link with the roadside perception devices via the Uu interface and the PC5 interface and transmit the roadside perception data to the edge cloud.

[0128] Step 102: Obtain M groups of BSM data (Basic Safety Message) of the target road section, where M is a positive integer.

[0129] The above-mentioned BSM data may be vehicle driving data collected by the vehicle terminal, such as driving speed, driving position, heading angle, etc.

[0130] The above-mentioned acquisition of M groups of BSM data for the target road section can be obtained by acquiring roadside perception data collected by the OBU (On board Unit) of several vehicles traveling on the target road section, thereby obtaining the M groups of roadside perception data, that is, M is related to the number of vehicles traveling on the target road section, and the OBU of one vehicle can collect a group of BSM data. Specifically, the OBUs of several vehicles traveling on the target road section can upload the BSM data collected by each of them to the RSU, and then the RSU uploads the BSM data of the OBU of each vehicle to the edge cloud, so that the edge cloud can receive the M groups of BSM data corresponding to the target road section.

[0131] It should be noted that the collection time of the N groups of roadside perception data and the M groups of BSM data is corresponding, that is, the edge cloud can obtain roadside perception data and BSM data of the same road section and the same time period.

[0132] Step 103: Perform confidence judgment on the i-th group of roadside perception data based on the i-th group of roadside perception data and the M groups of BSM data, where i takes values from 1 to N in sequence.

[0133] In an embodiment of the present invention, to ensure the security and reliability of roadside sensing data, the N sets of roadside sensing data can be combined with the M sets of acquired BSM data to determine the credibility of each set of roadside sensing data. The M sets of BSM data can be highly reliable BSM data collected by each vehicle's OBU in stable mode. It should be noted that stable mode is a specific operating mode of the OBU. BSM data collected in this mode possesses a high degree of credibility and can be directly adopted.

[0134] Specifically, for each group of the N groups of roadside perception data, the group of roadside perception data can be compared with the M groups of BSM data, and whether the group of roadside perception data is credible can be determined based on the comparison result. For example, the confidence of each group of roadside perception data can be compared and analyzed to see whether it meets expectations. The roadside perception data whose confidence meets the expected level can be determined to be credible, that is, it is determined that the confidence of the group of roadside perception data meets the requirements.

[0135] For example, the i-th group of roadside perception data is compared with the M groups of BSM data respectively, and whether the i-th group of roadside perception data is credible is determined by analyzing the comparison results between the i-th group of roadside perception data and each group of BSM data. If the comparison results show that the traffic object information reflected by the i-th group of roadside perception data is basically consistent with that reflected by each group of BSM data, it can be determined that the i-th group of roadside perception data is credible.

[0136] It should be noted that in order to ensure the accuracy of confidence judgment, the edge cloud can also filter the N groups of roadside perception data before making confidence judgment on the N groups of roadside perception data. For example, each group of the N groups of roadside perception data can be smoothed, denoised and filtered based on multi-frame data to reduce the error of each group of roadside perception data.

[0137] Preferably, a plurality of groups of roadside sensing devices are arranged according to positions on the target road section, each group of roadside sensing data is collected by a group of roadside sensing devices on the target road section, a group of roadside sensing devices includes a plurality of roadside sensors, and each group of roadside sensing data includes a plurality of sensor sensing data;

[0138] After step 101 and before step 103, the method further includes:

[0139] The plurality of sensor perception data in the i-th group of roadside perception data among the N groups of roadside perception data are fused to obtain the i-th group of fused roadside perception data.

[0140] That is, in one embodiment, each group of roadside perception data includes perception data collected by multiple roadside perception devices in the corresponding group of roadside perception devices. On the edge cloud side, the multiple sensor perception data in each group of roadside perception data received can be fused to obtain each group of fused roadside perception data.

[0141] Taking the fusion of perception data from a group of roadside sensing devices, including lidar and cameras, as an example, the roadside lidar devices can be used to collect and generate structured point cloud data. During the data collection process, after selecting a data collection route, the roadside lidar devices can be used to collect various point cloud data. The collected point cloud data can then be calibrated, fused, and spliced in sequence to generate ground point cloud data. After reading the raw data stream from the roadside lidar device, time synchronization can be performed based on the received timestamp matching, and the raw data can then be converted into structured data using the CenterPoint algorithm. LiDAR-based perception and recognition data can include information such as the ID, location, size, speed, and heading angle of the traffic object.

[0142] The core step of image recognition is to use the CenterNet algorithm to transform the target detection problem into a standard key point estimation problem. The key point estimation can be used to find the center point of each traffic object in the image. The center point calculation formula can be (x1, y1) and (x2, y2) can be the coordinates of two edge points of the area where the traffic object is located. Ultimately, the center point coordinates of the object can be used to associate with other target attributes of the traffic object, such as ID, position, size, speed, heading angle, and other information.

[0143] The video structured data and point cloud structured data can then be unified into a coordinate system and information fused. To unify roadside sensing devices such as lidar and cameras into a single coordinate system, multi-sensor calibration can be performed after data collection. The coordinates of multiple control points in the camera coordinate system and the world coordinate system representing the 3D scene structure are used to determine the absolute positional relationship between the two coordinate systems. Finally, the lidar and camera perception data can be deduplicated, and the perception data of the same matched traffic object can be fused to obtain complete traffic object information.

[0144] In this way, by fusing multiple sensor perception data in each group of roadside perception data on the edge cloud side, each group of fused roadside perception data can be obtained, which helps to comprehensively and accurately calculate the traffic object data of the target road section.

[0145] Preferably, step 103 includes:

[0146] Calculating the confidence level of the i-th group of roadside perception data based on the i-th group of roadside perception data and the M groups of BSM data;

[0147] When the confidence level of the i-th group of roadside perception data is within a preset range, it is determined that the confidence level of the i-th group of roadside perception data meets the requirement.

[0148] In one embodiment, the confidence level of each set of roadside perception data may be determined by calculating the confidence level of each set of roadside perception data and then determining whether the confidence level is within a preset range. For roadside perception data whose confidence level is within the preset range, it may be determined that the confidence level meets the requirements.

[0149] Specifically, for the i-th set of roadside perception data, the i-th set of roadside perception data can be compared with the M sets of BSM data, and the confidence level of the i-th set of roadside perception data relative to each set of BSM data can be calculated based on the comparison results. If the confidence level is within a preset range, the confidence level of the i-th set of roadside perception data is determined to meet the requirements, and the i-th set of roadside perception data can be output. The preset range can be a confidence level range set based on actual confidence requirements, such as a confidence level range between 0.8 and 1.

[0150] In one approach, the i-th group of roadside perception data can be compared with the M groups of BSM data to determine the errors between the i-th group of roadside perception data and each group of BSM data, and then the confidence level of the i-th group of roadside perception data can be determined by analyzing each error. For example, by comparing the i-th group of roadside perception data with the M groups of BSM data to determine the errors between the i-th group of roadside perception data and each group of BSM data, M error terms can be obtained. The confidence level of the i-th group of roadside perception data can be determined by analyzing and determining the number or proportion of errors within the allowable range among the M error terms. If more than 90% of the errors in the M error terms are within the allowable range, the confidence level of the i-th group of roadside perception data is determined to be 0.9.

[0151] In this way, this implementation method can quickly and accurately determine whether each group of roadside perception data is credible by specifically calculating the confidence of each group of roadside perception data and comparing it with the preset range, and can improve the safety and reliability of roadside perception data by reasonably adjusting the preset range.

[0152] Preferably, the preset range is greater than or equal to the confidence threshold and less than or equal to 1;

[0153] There are multiple groups of roadside sensing devices arranged according to their positions on the target road section, each group of roadside sensing data is collected by a group of roadside sensing devices on the target road section, and the confidence threshold is calculated based on the multiple groups of roadside sensing data collected by the multiple groups of roadside sensing devices within a historical time period, and the multiple groups of BSM data on the target road section within the historical time period.

[0154] That is, in one embodiment, the preset range can be a range between the confidence threshold and 1, that is, for the i-th group of roadside perception data, only when it is calculated that the confidence of the i-th group of roadside perception data is greater than or equal to the confidence threshold and less than or equal to 1, is it considered that the confidence of the i-th group of roadside perception data meets the requirements and is credible.

[0155] Among them, the confidence threshold is calculated based on multiple groups of roadside perception data collected by multiple groups of roadside perception devices on the target road section within a historical time period, and multiple groups of BSM data on the target road section within the historical time period, that is, the confidence threshold can be pre-calculated during the debugging stage.

[0156] Specifically, during the confidence debugging stage, multiple confidence tests can be performed based on multiple groups of roadside perception data collected by multiple groups of roadside perception devices on the target road section within a certain time period, and multiple groups of BSM data of OBUs of vehicles traveling on the target road section within the same time period, to finally determine a more reasonable confidence threshold, which is the minimum threshold that must be met for the roadside perception data to be credible.

[0157] More specifically, a preset confidence calculation formula can be used, with multiple sets of roadside perception data acquired within the historical interval on the target road section as input data and multiple sets of BSM data acquired within the historical interval on the target road section as output data, substituted into the confidence calculation formula to calculate a confidence parameter in the confidence calculation formula, and the verified confidence parameter is determined as the confidence threshold. The multiple sets of roadside perception data and the multiple sets of BSM data acquired within the historical interval on the target road section can both be calibrated or debugged data with high credibility to ensure the reliability of the calculated confidence threshold.

[0158] The confidence calculation formula may be Y=aX+b, where X may represent roadside perception data, Y may represent corresponding BSM data, a and b are parameters to be determined, and a may be set as the confidence.

[0159] The following uses a specific example to illustrate how to determine the confidence threshold.

[0160] For example, in the same scenario, fused structured data can be extracted from the roadside, including the ID, location, heading angle, speed, size, and other information of traffic objects. Each output parameter can then be compared with the true value data, which can be output through a highly stable OBU or a high-precision RTK (Real-Time Kinematic) terminal. The specific method can be as follows:

[0161] Within the coverage of the roadside sensing equipment, test personnel are asked to use vehicle-mounted equipment such as OBU or RTK terminal to simulate real traffic scenarios for testing. Taking the positioning data in the roadside sensing data as an example, the position data of each traffic object perceived and identified by a single group of roadside sensing equipment in the same time period can be obtained as X = [x1, x2, ..., x n ] T , and the position data Y of each traffic object output by the vehicle-mounted equipment = [y1, y2, ..., y n ] T , input data X and Y into the confidence calculation formula Y = aX + b, and preliminarily calculate a0 and b0, and record a0 as the confidence parameter, a0 = |A0|, A0 = [a1, a2, ..., a n ], a0 can also be verified by combining multiple sets of measured data. The verified a0 satisfies |A0|≤1. After multiple verifications with measured data, it can be concluded that when the confidence level a of a set of roadside perception data satisfies a0≤a≤1, the roadside perception data is credible. Otherwise, the confidence level of the roadside perception data is too low and cannot be used temporarily.

[0162] In this way, a reasonable confidence threshold can be determined through this implementation, and the confidence range determined based on the confidence threshold can be accurately used to judge the confidence of each group of roadside perception data, thereby ensuring the safety and reliability of the roadside perception data used.

[0163] Preferably, calculating the confidence level of the i-th group of roadside perception data based on the i-th group of roadside perception data and the M groups of BSM data includes:

[0164] Calculating M confidence levels of the i-th group of roadside perception data based on the i-th group of roadside perception data and each group of BSM data in the M groups of BSM data;

[0165] When the confidence level of the i-th group of roadside perception data is within a preset range, determining that the confidence level of the i-th group of roadside perception data meets the requirement includes:

[0166] If less than L confidence levels among the M confidence levels are not within the preset range, determining that the confidence level of the i-th group of roadside sensing data meets the requirement, where L is a preset value and L is a positive integer;

[0167] After calculating the M confidence levels of the i-th group of roadside perception data, the method further includes:

[0168] If more than L of the M confidence levels are outside the preset range, it is determined that the confidence level of the i-th group of roadside perception data does not meet the requirement, and output of the i-th group of roadside perception data is stopped.

[0169] That is, in one embodiment, the confidence judgment of the i-th group of roadside perception data can be based on the i-th group of roadside perception data and each group of BSM data in the M groups of BSM data, and the confidence of the i-th group of roadside perception data relative to each group of BSM data is calculated, thereby calculating M confidences, and the confidence calculated each time can be compared with the preset range, and the number of confidences that are not within the preset range can be counted. When more than a certain number of confidences, such as L confidences are not within the preset range, it can be determined that the confidence of the i-th group of roadside perception data does not meet the requirements, and the output of the i-th group of roadside perception data can be suspended, that is, the i-th group of roadside perception data is temporarily not adopted, wherein L can be a value set according to actual needs or a more reasonable value determined through testing, such as L can take values such as 3, 5, etc. When the number of confidence levels that are not within the preset range does not exceed L, it can be determined that the confidence level of the i-th group of roadside perception data meets the requirements, that is, the i-th group of roadside perception data can be used.

[0170] In this way, through this implementation, the confidence level of each set of roadside perception data can be judged more accurately, thereby ensuring the safety and reliability of the roadside perception data.

[0171] Preferably, after calculating the M confidence levels of the i-th group of roadside perception data, the method further includes:

[0172] When it is calculated based on the i-th group of roadside perception data and the j-th group of BSM data that the j-th confidence level of the i-th group of roadside perception data is not within the preset range, and no more than L confidence levels among the M confidence levels are not within the preset range, it is determined that the confidence level of the j-th group of BSM data does not meet the requirements, and a warning message is sent to the target vehicle-mounted terminal corresponding to the j-th group of BSM data, where j is an integer between 1 and M.

[0173] That is, in one embodiment, when the confidence level of the i-th group of roadside perception data meets the requirements, but the confidence level between a certain group of BSM data and the i-th group of roadside perception data is not within the preset range, it can be determined that the confidence level of the BSM data does not meet the requirements, that is, the BSM data is unreliable, and there may be a problem with the OBU corresponding to the BSM data. Therefore, an early warning message can be sent to the OBU corresponding to the BSM data to remind the vehicle user corresponding to the OBU that there may be an abnormality or fault in the OBU and it needs to be checked.

[0174] In this way, through this implementation, it is possible to determine whether the i-th group of roadside perception data or the j-th group of BSM data is unreliable based on the M confidence levels of the i-th group of roadside perception data relative to the M groups of BSM data, which helps to promptly discover problematic equipment and solve abnormal equipment problems in a timely manner.

[0175] Preferably, after determining that the confidence level of the i-th group of roadside perception data does not meet the requirement and stopping outputting the i-th group of roadside perception data, the method further includes:

[0176] The i-th group of roadside perception data is input into a pre-established confidence correction model to obtain the i-th group of roadside perception data after correction output by the confidence correction model.

[0177] That is, in one embodiment, a confidence correction model can be pre-established to correct roadside perception data whose confidence does not meet the requirements, wherein the confidence correction model can be pre-modeled based on multiple sets of credible roadside perception data and multiple sets of BSM data obtained within the same historical time period of the target road section. Specifically, the multiple sets of credible roadside perception data can be used as input data of the confidence correction model, and the multiple sets of BSM data can be used as the true value output of the confidence correction model. The parameters of the confidence correction model are calculated through data fitting, and then the confidence correction model with determined parameters is obtained.

[0178] In this way, when it is determined that the confidence of the i-th group of roadside perception data does not meet the requirements, the i-th group of roadside perception data can be corrected based on the confidence correction model. Specifically, the i-th group of roadside perception data can be input into the pre-established confidence correction model for correction calculation to obtain the corrected i-th group of roadside perception data output by the confidence correction model. The corrected i-th group of roadside perception data can continue to be output.

[0179] Through this implementation, the confidence correction model can be used to correct roadside perception data whose confidence does not meet the requirements, thereby improving the availability and safety reliability of the roadside perception data.

[0180] Preferably, the confidence correction model is Y=c0X 5 +d0X 4 +e0X 3 +f0X 2 +g0X+h0, where X is the roadside perception data before correction, Y is the roadside perception data after correction, and c0, d0, e0, f0, g0, and h0 are model parameters.

[0181] That is, in a specific implementation, a binary quintic equation can be used as a confidence correction model. When preparing to establish the confidence correction model, Y=cX 5 +dX 4 +eX 3 +fX 2 +gX+h is the confidence correction model with the initial parameters to be determined. Then, multiple sets of reliable roadside perception data are used as the model input data X, and the corresponding multiple sets of BSM data are used as the true value output Y of the model. The accurate relationship model between Y and X is calculated, and the model parameters [c0, d0, e0, f0, g0, h0] are determined, thus obtaining the parameter-determined confidence correction model Y=c0X 5 +d0X 4 +e0X 3 +f0X 2 +g0X+h0. The confidence correction model is an order determined through experiments that can ensure the accuracy of the correction while avoiding the calculation of an overly complex correction model, that is, the confidence correction model can take into account both latency and performance.

[0182] When the i-th group of roadside perception data needs to be corrected, the i-th group of roadside perception data can be used as input data X, and after correction calculation by the confidence correction model, the output data Y of the confidence correction model is obtained as the corrected i-th group of roadside perception data.

[0183] In this way, by selecting the confidence correction model of this implementation to correct the roadside perception data whose confidence does not meet the requirements, it is possible to ensure the accuracy of the correction results and a faster correction speed, taking into account both processing delay and model performance.

[0184] Preferably, after inputting the i-th group of roadside perception data into a pre-established confidence correction model to obtain the corrected i-th group of roadside perception data output by the confidence correction model, the method further includes:

[0185] Obtaining M′ groups of BSM data for the target road section, where M′ is a positive integer;

[0186] Calculating M′ confidence levels of the i-th group of roadside perception data based on the corrected i-th group of roadside perception data and each group of BSM data in the M′ groups of BSM data;

[0187] When more than L confidence levels among the M′ confidence levels are within the preset range, it is determined that the confidence level of the corrected i-th group of roadside perception data meets the requirement, and the corrected i-th group of roadside perception data is output.

[0188] In one embodiment, the i-th group of roadside perception data corrected by the confidence correction model may be further tested to determine whether the confidence of the corrected i-th group of roadside perception data meets the requirements.

[0189] Specifically, the M′ groups of BSM data of the target road section can be obtained in real time, that is, the latest M′ groups of BSM data of the target road section can be obtained, and the confidence calculation can be continued with the corrected i-th group of roadside perception data to verify whether the confidence of the corrected i-th group of roadside perception data meets the requirements. If it is determined that its confidence meets the requirements, the corrected i-th group of roadside perception data can be adopted. If it is determined that its confidence still does not meet the requirements, the output of the corrected i-th group of roadside perception data can be suspended, and the roadside perception equipment corresponding to the i-th group of roadside perception data can be checked for abnormalities. Among them, the specific confidence calculation method and confidence judgment method can be similar to the aforementioned implementation method. For details, please refer to the aforementioned related introduction and will not be repeated here.

[0190] In this way, by correcting and verifying the roadside perception data whose confidence does not meet the requirements, the reliability of the adopted roadside perception data can be further ensured.

[0191] For example, Figure 2 As shown, in one embodiment, the confidence calculation of the i-th group of roadside perception data and the 1st to M-th groups of BSM data in the M groups of BSM data can be performed in sequence to obtain the confidence a i,1 、a i,2 、a i,3 ,…,a i,M , and each time a confidence level is calculated, it is determined whether the confidence level is within the preset range. When the confidence level a between the jth group of BSM data and the ith group of roadside sensing data is greater than i,j When it is not within the preset range [a0,1], the output of the i-th group of roadside perception data can be suspended, and the confidence calculation of other groups of BSM data with adjacent time, such as the j+1-th group of BSM data, the j+2-th group of BSM data, etc., can be performed with the i-th group of roadside perception data to obtain the confidence a i,j+1 、ai,j+2 etc., if a i,j+1 and a i,j+2 If the confidence levels calculated for more than three groups of stable and reliable BSM data and the i-th group of roadside perception data are not within the preset range, then it can be determined that the i-th group of roadside perception data is unreliable, that is, its confidence level does not meet the requirements.

[0192] The confidence calculation is performed on the j+1th group of BSM data, the j+2th group of BSM data, etc., and the ith group of roadside perception data, and the confidence a is obtained. i,j+1 、a i,j+2 etc. are all within the preset range, that is, the confidences calculated from the other two groups of BSM data and the i-th group of roadside perception data all meet the requirements, and only the confidence relative to the j-th group of BSM data does not meet the requirements, then it can be determined that the j-th group of BSM data is unreliable, and the situation is fed back to the vehicle that outputs the j-th group of BSM data, and the j-th group of BSM data can be continuously compared and judged with the perception data of other groups of roadside perception devices, and fed back to this vehicle.

[0193] If it does not belong to the above two cases, for example, the confidence level a of the neighbor i,j+1 、a i,j+2 If some of the data are within the preset range and some are not within the preset range, it is necessary to continue to suspend the output of the i-th group of roadside sensing data, and call for manual review in time to troubleshoot the problem.

[0194] In addition, if the confidence of the above-mentioned i-th group of roadside perception data does not meet the requirements, the i-th group of roadside perception data can be corrected using a confidence correction model, and the confidence of the corrected i-th group of roadside perception data and the stable output BSM data obtained in real time can be calculated to verify the accuracy of the corrected data. When more than three groups of confidences are within the preset range, the corrected i-th group of roadside data can be output again.

[0195] Preferably, after determining that the confidence level of the i-th group of roadside perception data does not meet the requirement and stopping outputting the i-th group of roadside perception data, the method further includes:

[0196] Performing a confidence judgment on the target roadside sensing device corresponding to the i-th group of roadside sensing data;

[0197] When the confidence level of the target roadside perception device does not meet the requirement, the roadside perception data collected by the target roadside perception device is corrected.

[0198] In another embodiment, if it is determined that the confidence level of the i-th set of roadside sensing data does not meet the requirements, it may be considered that a confidence level issue exists with the group of roadside sensing devices corresponding to the i-th set of roadside sensing data, thereby necessitating a confidence assessment for a single roadside sensing device within the group of roadside sensing devices. Alternatively, if the confidence level of the i-th set of roadside sensing data is still determined to not meet the requirements after correction using the aforementioned confidence correction formula, a confidence assessment may be performed for a single roadside sensing device within the group of roadside sensing devices corresponding to the i-th set of roadside sensing data.

[0199] If it is determined that the confidence level of a particular roadside sensing device in the group does not meet the required level, the roadside sensing data collected by that roadside sensing device may be corrected to improve the reliability of the roadside sensing data collected by the group of roadside sensing devices. The method for determining the confidence level of a single roadside sensing device is similar to the method for determining the confidence level of a single group of roadside sensing data described above and will not be further described here.

[0200] In this embodiment, a confidence correction model can be pre-established for each individual roadside sensing device in each group of roadside sensing devices. For example, in the same scenario, the perception data of each individual roadside sensing device, such as a camera or lidar, can be extracted, including information such as the ID, location, orientation angle, speed, and size of the traffic object. Each piece of perception data can be compared with the true value data through actual measurement. The true value data can be calibrated and trusted BSM data. After multiple comparisons and verifications, the precision and recall rates of the traffic object, as well as the error between each piece of perception data and the true value data, can be output. After comparing multiple groups of data, a confidence correction model can be established for each individual roadside sensing device. If it is found in an actual scenario that the confidence of a roadside sensing device does not meet the requirements, the confidence correction model of the roadside sensing device can be used to correct the roadside sensing data collected by the roadside sensing device and then output.

[0201] In this way, through this implementation, roadside perception devices with abnormal perception data can be gradually located, and the reliability of roadside perception data can be further improved by combining the confidence judgment and data correction of a single roadside perception device.

[0202] Step 104: Output K groups of roadside perception data whose confidences meet the requirements among the N groups of roadside perception data, where K is a positive integer less than or equal to N.

[0203] After making a confidence judgment on each of the N groups of roadside perception data respectively, the K groups of roadside perception data whose confidence meets the requirements can be output, and the K groups of roadside perception data can be retained for the next processing. For example, N is 5. When it is determined that the confidence of 3 groups of roadside perception data among the 5 groups of roadside perception data meets the requirements, these 3 groups of roadside perception data can be output. For the other 2 groups of roadside perception data whose confidence does not meet the requirements, the output can be suspended or discarded. For example, the acquisition of the roadside perception data of the 2 groups of roadside perception devices corresponding to the 2 groups of roadside perception data whose confidence does not meet the requirements can be stopped.

[0204] Step 105: Based on the traffic data on the target road section, provide vehicle-road collaborative information services to traffic objects on the target road section, wherein the traffic data includes the K groups of roadside perception data.

[0205] After outputting the K sets of roadside perception data that meet the required confidence levels, the edge cloud can determine the road traffic information of the target road segment based on the traffic data including the K sets of roadside perception data, and then provide vehicle-road collaborative information services to traffic objects on the target road segment based on the road traffic information. The vehicle-road collaborative information services can include dynamic traffic information broadcasting, driving risk rating, driving route recommendations, traffic flow prediction, road hazard reminders, etc. Traffic objects on the target road segment can include vehicles and pedestrian terminals traveling on the target road segment and connected to the Internet of Vehicles.

[0206] The traffic data may also include weather, traffic light data, high-precision map data, etc. in the area where the target road section is located. That is, the edge cloud can also obtain rich and multi-source traffic data from other channels, and combine it with the K groups of roadside perception data to analyze the road traffic information of the target road section, so that the provided vehicle-road collaborative information service is safer and more reliable.

[0207] Preferably, after step 104, the method further includes:

[0208] When K is greater than 1, the K groups of roadside perception data are fused to obtain global roadside fused perception data;

[0209] The traffic data includes the global roadside fusion perception data.

[0210] In one embodiment, after performing confidence judgment on the N groups of roadside perception data and determining K groups of roadside perception data whose confidence meets the requirements, the K groups of roadside perception data can be fused when K is greater than 1, that is, when there are two or more groups of roadside perception data whose confidence meets the requirements, to obtain global roadside fused perception data. Then, the edge cloud can provide traffic objects on the target road section with a wider coverage, more comprehensive and accurate vehicle-road collaborative information service based on the traffic data including the global roadside fused perception data.

[0211] Specifically, each of the K groups of roadside perception data can first be time-compensated, for example, by performing position compensation on each group of roadside perception data based on the difference between the data reception time and the processing time. Then, based on the topological relationship diagram of the groups of roadside perception devices deployed on the target road section, it can be determined which device source data needs to be fused. During the fusion process, matching can be performed based on data similarity, such as using the Hungarian algorithm to calculate the best match to obtain a matching result. The matched roadside perception data can then be weighted and fused based on the confidence judgment results described above to obtain global roadside fused perception data. The global ID of each traffic object on the target road section can also be updated, meaning that the same traffic object passing through different areas on the target road section will have a unique ID.

[0212] Through this implementation, global fusion of data of traffic objects on the target road section can be achieved, thereby improving the business coverage of the edge cloud and ensuring business reliability.

[0213] Preferably, after fusing the K groups of roadside sensing data, the method further includes:

[0214] Fusing the M groups of BSM data with the global roadside fusion perception data to obtain first fused traffic data;

[0215] The traffic data includes the first fused traffic data.

[0216] In one embodiment, after fusing the K groups of roadside perception data, the M groups of BSM data obtained can be further fused with the global roadside fusion perception data to obtain more reliable first fused traffic data, and then the edge cloud can provide more accurate vehicle-road collaborative information services to traffic objects on the target road section based on the traffic data including the first fused traffic data.

[0217] Specifically, the edge cloud can first perform time compensation on the M sets of acquired BSM data based on the difference between the data reception time and the processing time. Then, for the BSM data collected in stable mode and with higher confidence among the M sets of BSM data, the edge cloud can match the global roadside fusion perception data based on the data similarity principle, establish a correspondence between the traffic object ID stored in the OBU and the cached global traffic object ID, and then perform a weighted fusion of the BSM data and the global roadside fusion perception data based on the ID correspondence and the BSM confidence. Finally, the fusion result is output and cached to obtain road traffic data based on a global perspective.

[0218] Through this implementation, multi-source fusion of perception data from different devices on the target road section can be achieved, thereby improving the reliability of the vehicle-road collaborative information service provided by the edge cloud.

[0219] Preferably, the method further comprises:

[0220] Obtain G groups of SSM data (Sensor Sharing Message, perception sharing message) of the target road section, where G is a positive integer;

[0221] After fusing the M groups of BSM data with the global roadside fusion perception data, the method further includes:

[0222] fusing the G group of SSM data with the first fused traffic data to obtain second fused traffic data;

[0223] The traffic data includes the second fused traffic data.

[0224] In another embodiment, G group SSM data sent by vehicles traveling on the target road section can also be obtained, and after the roadside perception data and BSM data are integrated, the obtained G group SSM data are further integrated with the first integrated traffic data to obtain more complete and reliable second integrated traffic data. Then, the edge cloud can provide more accurate and reliable vehicle-road collaborative information services to traffic objects on the target road section based on the traffic data including the second integrated traffic data.

[0225] The SSM data may refer to the perception data of the road being traveled by sensors such as lidar and cameras installed on the vehicle.

[0226] Specifically, the edge cloud can first perform time compensation on the vehicle-side SSM data, performing location compensation on each set of SSM data based on the difference between the data reception time and the processing time. The G sets of SSM data can then be matched with the first fused traffic data based on data similarity, and confidence can be calculated for the matched data. If both are within the trustworthy range, the G sets of SSM data are supplemented and fused with the first fused traffic data to obtain the second fused traffic data, and the global ID of each traffic object on the target road section can be updated.

[0227] This implementation enables multi-source fusion of sensory data from various types of devices on the target road section, thereby improving the reliability of the vehicle-road collaborative information service provided by the edge cloud. Furthermore, service coverage can reach the district level, and computing processing latency can be reduced to hundreds of milliseconds.

[0228] Preferably, the step 105 includes:

[0229] determining object information and motion data of traffic objects on the target road segment based on traffic data on the target road segment;

[0230] Based on the object information and motion data of the traffic object, the motion trajectory of the traffic object is predicted, or the risk rating of the traffic object is performed.

[0231] In one embodiment, the edge cloud can determine the target data of the traffic objects on the target road section, such as traffic object information, traffic object motion data, etc., based on the collected traffic data on the target road section, specifically based on the fused traffic data, by analyzing and calculating the traffic data. The edge cloud can also predict the motion trajectory of the traffic objects based on the determined information and motion data of the traffic objects, such as predicting the driving trajectory of a vehicle in the next 5 seconds, or can perform risk rating on the traffic objects to determine whether the traffic objects have driving risks, such as analyzing whether the driver of a vehicle has dangerous driving behavior, or analyzing whether there are safety risks on the road section where pedestrians are located.

[0232] It should be noted that the edge cloud can send the predicted movement trajectory of the traffic object or the risk rating of the traffic object to the terminal of the corresponding traffic object for use by each traffic object, and can also further report the information to the central cloud, so that the central cloud with stronger processing capabilities and wider coverage can perform big data analysis on the information uploaded by the edge cloud to obtain more accurate and comprehensive dynamic traffic information on the target road section.

[0233] For example, the edge cloud can perform a 5s trajectory prediction for traffic objects on the target road section based on the fused traffic data, such as the above-mentioned global roadside fusion perception data, the first fused traffic data or the second fused traffic data. First, the number of frames of traffic data that have been acquired can be judged. For example, when the frequency of the input data and the output data are both 2 frames / second, it can be determined whether there are 10 frames, or 5 seconds of historical data. If 10 frames of data are met, the judgment process for static traffic objects can be entered. For example, if the change in the position data of a certain traffic object in the last two frames is less than or equal to the preset static threshold, the traffic object is considered to be a static target object, and its next frame trajectory can be predicted without prediction, and the ID of the static traffic object can be output. For non-static traffic objects, their relevant data can be preprocessed and their trajectories can be predicted, such as predicting their movement trajectory for the next 5 seconds based on the position, movement speed, heading angle and other data of the traffic object. Finally, based on the predicted position data of the last two frames of the traffic object, the coordinate conversion formula |x1-x2|+|y1-y2|.(x i ,y i ), i=1,2, calculate the latitude and longitude coordinates of the last two frames of the traffic object.

[0234] The edge cloud can also combine the information of each traffic object and the 5s trajectory prediction information to perform risk rating on each traffic object. Specifically, the kinetic energy field calculation can be performed by combining the type, mass, speed, acceleration and other data of the traffic object, and then the risk rating of each traffic object can be calculated. The calculation formula of the risk rating can be: Among them, G, k1 and k2 are all constants, R i is the road condition influencing factor, M i is the equivalent mass of the traffic object, r ij Represents two position coordinate points (x i ,y i ) and (x j ,y j ), the distance vector between i is the velocity direction angle, v i is the movement speed of the traffic object. In this way, based on this formula, the influence of each traffic object on its surrounding traffic objects can be calculated, and then the risk rating of each traffic object in the overall coverage area can be obtained.

[0235] Through this implementation, it is possible to predict the movement trajectory of each relevant traffic object based on the collected traffic data on the edge cloud side, or to perform risk rating on each traffic object, and then provide relevant services to each traffic object with high reliability.

[0236] Preferably, after fusing the G group SSM data with the first fused traffic data to obtain second fused traffic data, the method further includes:

[0237] Uploading the second fused traffic data to the central cloud;

[0238] Receiving traffic information of the target road section sent by the central cloud;

[0239] The traffic information is sent to traffic objects on the target road section.

[0240] The above-mentioned central cloud can cover multiple edge clouds, that is, the coverage range of the central cloud can be a collection of the coverage ranges of multiple edge clouds. The central cloud can receive multi-source fused traffic data corresponding to their respective road sections uploaded in real time by each edge cloud, and then perform big data analysis on the quasi-real-time traffic data within its coverage area to determine the dynamic traffic information of each road section within the coverage area.

[0241] In this embodiment, the edge cloud can upload multi-source fused traffic data to the central cloud in real time. Specifically, the edge cloud can obtain the dynamic traffic information of the target road section after fusion processing and risk rating calculation of multi-source data on the road side and the vehicle side, and then upload the multi-source fusion calculation data to the central cloud in real time through a dedicated network for data storage.

[0242] The central cloud can perform big data analysis on near-real-time data within its coverage area. For example, the central cloud can analyze the driving data of multiple vehicles on the road to identify and mark bad driving habits, such as violent braking, sudden acceleration, and emergency lane changes, and can incorporate these into the database for use as hazard markers and road hazard reminders. The central cloud can also calculate and generate congestion events based on the average vehicle speed within the coverage area, and analyze and predict congestion traffic big data for a certain period of time to achieve interconnection between the vast majority or even all traffic objects and the central cloud network, forming traffic big data in the central cloud. Based on this traffic big data, the central cloud can implement traffic flow analysis, communication flow prediction, driving route recommendations, traffic guidance, driving habit analysis, driving intention analysis, etc., providing road traffic objects with diversified dynamic traffic information services.

[0243] The central cloud can also send dynamic traffic information to vehicles within the coverage area. Specifically, after the central cloud calculates the quasi-real-time dynamic traffic information, it can distribute it to the corresponding edge cloud by region, and then send it to the RSU via the Uu port or PC5 port, and the RSU broadcasts it to the vehicle. That is, the edge cloud can receive the quasi-real-time dynamic traffic information corresponding to its area sent by the central cloud, and can send the received quasi-real-time dynamic traffic information to the roadside sensing devices in its coverage area, that is, the target road section. The roadside sensing devices in the target road section will then broadcast the relevant quasi-real-time dynamic traffic information to traffic objects such as vehicles and pedestrians within their respective coverage areas.

[0244] In this way, by uploading the multi-source fused traffic data on the edge cloud side to the central cloud, it is possible to achieve quasi-real-time dynamic traffic information processing at the provincial, municipal and even national levels, with a wider coverage. Moreover, due to the high reliability of the fused traffic data, the high reliability of related businesses based on the data can be guaranteed accordingly.

[0245] The embodiment of the present invention can be applied to Figure 3 The vehicle-road cooperative system structure shown in the figure can also be used to exchange information between devices and cloud services. Figure 3 Indicated by the arrow.

[0246] The vehicle-road cooperative data processing method of an embodiment of the present invention obtains N sets of roadside sensing data for a target road section, where N is a positive integer; obtains M sets of basic safety message (BSM) data for the target road section, where M is a positive integer; performs a confidence assessment on the i-th set of roadside sensing data based on the i-th set of roadside sensing data and the M sets of BSM data, where i sequentially ranges from 1 to N; outputs K sets of roadside sensing data from the N sets of roadside sensing data whose confidences meet requirements, where K is a positive integer less than or equal to N; and provides a vehicle-road cooperative information service to traffic objects on the target road section based on traffic data on the target road section, where the traffic data includes the K sets of roadside sensing data. In this way, by performing a confidence assessment on the multiple sets of roadside sensing data obtained for the target road section and providing a vehicle-road cooperative information service to traffic objects on the target road section based on the roadside sensing data whose confidences meet requirements, the security and reliability of the roadside sensing data can be improved, thereby ensuring the security and reliability of the vehicle-road cooperative information service.

[0247] The present invention proposes a perception data fusion solution based on confidence judgment. In the application scenario of autonomous driving vehicles, the perception computing capability can be moved from the vehicle side to the vehicle-road perception system based on the edge cloud. In addition to the requirements for network stability and speed and the accuracy of the sensing equipment, this system also needs to evaluate the credibility of the perception data to ensure the security and reliability of the perception data. Based on this solution, not only can the confidence of the roadside perception data be verified under test conditions, but the roadside perception data can also be automatically corrected in actual application scenarios, and untrusted feedback can be performed on the vehicle-side BSM data and vehicle-side SSM data to ensure the data security and stability of the entire perception link. At the same time, based on this confidence judgment, the present invention proposes a perception data fusion order of the roadside first and the vehicle side, which is more reliable in actual application environments. By promoting the application of all-view, all-weather smart road multi-fusion perception solutions, the standardized transformation of smart roads and the verification of road network carrying capacity can be achieved, thereby helping to promote the classification and standardized transformation of smart roads and fill the gaps in related fields.

[0248] Specifically, the solution in the embodiment of the present invention has the following advantages over the existing related technologies:

[0249] 1) This invention uses roadside cameras and lidar to collect video data and point cloud data from roadside sensor equipment, respectively. By perceiving, identifying, and fusing this raw data, structured data of traffic participants from the roadside perspective is obtained. The current perception and recognition algorithm supports lidar devices with 32 lines or higher, and cameras with a resolution of 1920×1080p. By fusing these two, it can identify people, motor vehicles, non-motor vehicles, and other object types, with recognition accuracy, precision, and recall rates all reaching 95% or higher.

[0250] 2) The multi-source perception fusion solution proposed in this invention is adaptable to a wide range of sensor devices. It can fuse structured data from roadside LiDAR, roadside millimeter-wave radar, roadside cameras, vehicle-mounted LiDAR, vehicle-mounted millimeter-wave radar, and vehicle BSM data to form a stable data protocol that is adaptable to a wider range of perception devices. Furthermore, it can compare perception data from devices from different manufacturers in the same environment to form a device confidence comparison model.

[0251] 3) The present invention proposes to divide the fusion steps into data fusion of a single group of roadside sensor devices, fusion of multiple groups of roadside perception data, vehicle BSM data fusion, and SSM data fusion. Considering that roadside perception data is relatively stable and controllable, and the data source is clear, confidence calculations can be performed multiple times under the test state, and confidence comparisons can be performed on the data of all connected devices, thereby outputting a more reliable roadside perception data confidence calculation model. On the other hand, more reliable perception data can be output based on the confidence correction model. In actual scenarios, SSM and BSM data come from the vehicle side, and the data source is unstable, which also means that the confidence is unstable. Therefore, it is possible to consider performing confidence calculations on SSM and BSM in actual scenarios, thereby supplementing the roadside perception data based on the confidence, thereby judging the stability of the vehicle-side data, and then fusing it with the roadside data, thereby outputting more accurate and reliable multi-source fusion data, providing overall data support for subsequent risk rating, planning decisions, and dynamic traffic information.

[0252] 4) In the present invention, traffic participants and roadside RSUs can establish highly dynamic communication links through the Uu interface and the PC5 interface. The vehicle-side data and perception device data are aggregated and uploaded through the roadside RSU on the PC5 or 5G Uu communication link of C-V2X, and the edge cloud real-time calculation and key traffic information are distributed to the vehicle side, providing overall end-to-end information services. The communication link delay can be controlled within 100ms, the roadside perception fusion delay is within 60ms, and the edge cloud side multi-source data fusion calculation delay can reach within 40ms. Therefore, combined with the communication link and calculation delay, the delay can be controlled within 200ms. By continuously optimizing link transmission and calculation delay, the overall system delay target can be controlled within 100ms, while also ensuring data stability.

[0253] 5) The edge computing in this invention is primarily based on the edge cloud, offering the advantages of strong computing power, fast computation, and wide coverage. By uploading edge cloud computing data to the central cloud, near-real-time dynamic traffic information processing can be achieved at the provincial, municipal, and even national levels, meeting the needs of large-scale perception computing and enabling application in a variety of scenarios, greatly improving scalability and compatibility.

[0254] See also Figure 4 , Figure 4 is a flow chart of the vehicle-road cooperative data processing method provided by an embodiment of the present invention, such as Figure 4 As shown, the following steps are included:

[0255] Step 401: Receive the second fused traffic data sent by the edge cloud, wherein the second fused traffic data is obtained by fusing K groups of roadside perception data that meet the confidence requirements of the target road section, M groups of BSM data, and G groups of SSM data, where K is an integer greater than 1, and M and G are both positive integers.

[0256] Step 402: Determine the traffic information of the target road section based on the second fused traffic data.

[0257] Step 403: Send the traffic information to the edge cloud.

[0258] This embodiment is as Figure 1 The implementation method of the central cloud side corresponding to the embodiment shown in the figure can be found in the Figure 1 To avoid repetition, the relevant introduction in the illustrated embodiments will not be repeated here.

[0259] The vehicle-road collaborative data processing method of an embodiment of the present invention receives second fused traffic data sent by the edge cloud, wherein the second fused traffic data is obtained by fusing K groups of roadside perception data that meet the confidence requirements, M groups of BSM data, and G groups of SSM data of the target road section, where K is an integer greater than 1, and M and G are both positive integers; based on the second fused traffic data, the traffic information of the target road section is determined; and the traffic information is sent to the edge cloud. In this way, by determining the traffic information of the target road section based on the multi-source fused traffic data sent by the edge cloud and subjected to confidence judgment, the security and reliability of the vehicle-road collaborative information service can be guaranteed.

[0260] The embodiment of the present invention also provides an edge cloud. Figure 5 , Figure 5 This is a structural diagram of the edge cloud provided by an embodiment of the present invention. Since the principle of solving the problem by the edge cloud is similar to the vehicle-road cooperative data processing method in the embodiment of the present invention, the implementation of the edge cloud can refer to the implementation of the method, and the repeated parts will not be repeated.

[0261] like Figure 5 As shown, the edge cloud 500 includes:

[0262] The first acquisition module 501 is configured to acquire N sets of roadside sensing data of a target road section, where N is a positive integer;

[0263] The second acquisition module 502 is configured to acquire M groups of basic safety message (BSM) data of the target road section, where M is a positive integer;

[0264] A first processing module 503 is configured to perform a confidence judgment on the i-th group of roadside sensing data based on the i-th group of roadside sensing data and the M groups of BSM data, where i is 1 to N in sequence;

[0265] An output module 504 is configured to output K groups of roadside sensing data whose confidences meet the requirements among the N groups of roadside sensing data, where K is a positive integer less than or equal to N;

[0266] The execution module 505 is used to provide vehicle-road cooperative information services to traffic objects on the target road section based on the traffic data on the target road section, wherein the traffic data includes the K groups of roadside perception data.

[0267] Preferably, the first processing module 503 includes:

[0268] a calculation submodule, configured to calculate the confidence level of the i-th group of roadside perception data based on the i-th group of roadside perception data and the M groups of BSM data;

[0269] The first determination submodule is configured to determine whether the confidence level of the i-th group of roadside perception data meets the requirement when the confidence level of the i-th group of roadside perception data is within a preset range.

[0270] Preferably, the preset range is greater than or equal to the confidence threshold and less than or equal to 1;

[0271] There are multiple groups of roadside sensing devices arranged according to their positions on the target road section, each group of roadside sensing data is collected by a group of roadside sensing devices on the target road section, and the confidence threshold is calculated based on the multiple groups of roadside sensing data collected by the multiple groups of roadside sensing devices within a historical time period, and the multiple groups of BSM data on the target road section within the historical time period.

[0272] Preferably, the calculation submodule includes:

[0273] a calculation unit, configured to calculate M confidence levels of the i-th group of roadside perception data based on the i-th group of roadside perception data and each group of BSM data in the M groups of BSM data;

[0274] The first determination submodule is configured to determine that the confidence level of the i-th group of roadside sensing data meets the requirement when less than L confidence levels among the M confidence levels are not within the preset range, where L is a preset value and is a positive integer;

[0275] Edge Cloud 500 also includes:

[0276] The second processing module is used to determine that the confidence level of the i-th group of roadside perception data does not meet the requirements and stop outputting the i-th group of roadside perception data when more than L confidence levels among the M confidence levels are not within the preset range.

[0277] Preferably, the edge cloud 500 further includes:

[0278] The third processing module is used to determine that the confidence level of the j-th group of BSM data does not meet the requirements when calculating, based on the i-th group of roadside perception data and the j-th group of BSM data, that the j-th confidence level of the i-th group of roadside perception data is not within the preset range, and no more than L confidence levels among the M confidence levels are not within the preset range, and send a warning message to the target vehicle-mounted terminal corresponding to the j-th group of BSM data, where j is an integer between 1 and M.

[0279] Preferably, the edge cloud 500 further includes:

[0280] The first correction module is used to input the i-th group of roadside perception data into a pre-established confidence correction model to obtain the corrected i-th group of roadside perception data output by the confidence correction model.

[0281] Preferably, the edge cloud 500 further includes:

[0282] A third acquisition module is configured to acquire M′ groups of BSM data of the target road section, where M′ is a positive integer;

[0283] a calculation module, configured to calculate M′ confidence levels of the i-th group of roadside perception data based on the corrected i-th group of roadside perception data and each group of BSM data in the M′ groups of BSM data;

[0284] The fourth processing module is used to determine that the confidence level of the corrected i-th group of roadside perception data meets the requirements when more than L confidence levels among the M′ confidence levels are within the preset range, and output the corrected i-th group of roadside perception data.

[0285] Preferably, the confidence correction model is Y=c0X 5 +d0X 4 +e0X 3 +f0X 2+g0X+h0, where X is the roadside perception data before correction, Y is the roadside perception data after correction, and c0, d0, e0, f0, g0, and h0 are model parameters.

[0286] Preferably, the edge cloud 500 further includes:

[0287] a fifth processing module, configured to perform a confidence judgment on the target roadside sensing device corresponding to the i-th group of roadside sensing data;

[0288] The second correction module is used to correct the roadside perception data collected by the target roadside perception device when the confidence level of the target roadside perception device does not meet the requirements.

[0289] Preferably, a plurality of groups of roadside sensing devices are arranged according to positions on the target road section, each group of roadside sensing data is collected by a group of roadside sensing devices on the target road section, a group of roadside sensing devices includes a plurality of roadside sensors, and each group of roadside sensing data includes a plurality of sensor sensing data;

[0290] Edge Cloud 500 also includes:

[0291] The first fusion module is used to fuse multiple sensor perception data in the i-th group of roadside perception data among the N groups of roadside perception data to obtain the fused i-th group of roadside perception data.

[0292] Preferably, the edge cloud 500 further includes:

[0293] A second fusion module is configured to fuse the K groups of roadside perception data to obtain global roadside fusion perception data when K is greater than 1;

[0294] The traffic data includes the global roadside fusion perception data.

[0295] Preferably, the edge cloud 500 further includes:

[0296] a third fusion module, configured to fuse the M groups of BSM data with the global roadside fusion perception data to obtain first fused traffic data;

[0297] The traffic data includes the first fused traffic data.

[0298] Preferably, the edge cloud 500 further includes:

[0299] A fourth acquisition module is used to obtain G groups of SSM data of the target road section, where G is a positive integer;

[0300] a fourth fusion module, configured to fuse the G group SSM data with the first fused traffic data to obtain second fused traffic data;

[0301] The traffic data includes the second fused traffic data.

[0302] Preferably, the execution module 505 includes:

[0303] a second determining submodule, configured to determine object information and motion data of traffic objects on the target road section based on the traffic data on the target road section;

[0304] The processing submodule is used to predict the motion trajectory of the traffic object based on the object information and motion data of the traffic object, or to perform risk rating on the traffic object.

[0305] Preferably, the edge cloud 500 further includes:

[0306] A second sending module, configured to upload the second fused traffic data to a central cloud;

[0307] A second receiving module is used to receive the traffic information of the target road section sent by the central cloud;

[0308] The third sending module is used to send the traffic information to the traffic objects on the target road section.

[0309] The edge cloud provided by the embodiment of the present invention can execute Figure 1 The implementation principle and technical effects of the method embodiment shown are similar, and will not be described in detail here.

[0310] The edge cloud 500 of an embodiment of the present invention obtains N sets of roadside sensing data for a target road section, where N is a positive integer; obtains M sets of basic safety message (BSM) data for the target road section, where M is a positive integer; performs a confidence assessment on the i-th set of roadside sensing data based on the i-th set of roadside sensing data and the M sets of BSM data, where i sequentially ranges from 1 to N; outputs K sets of roadside sensing data from the N sets of roadside sensing data whose confidences meet requirements, where K is a positive integer less than or equal to N; and provides a vehicle-road cooperative information service to traffic objects on the target road section based on traffic data on the target road section, where the traffic data includes the K sets of roadside sensing data. In this way, by performing confidence assessments on the multiple sets of roadside sensing data obtained for the target road section and providing a vehicle-road cooperative information service to traffic objects on the target road section based on the roadside sensing data whose confidences meet requirements, the security and reliability of the roadside sensing data can be improved, thereby ensuring the security and reliability of the vehicle-road cooperative information service.

[0311] The embodiment of the present invention also provides a central cloud. Figure 6 , Figure 6This is a structural diagram of the central cloud provided by an embodiment of the present invention. Since the principle of solving the problem by the central cloud is similar to the vehicle-road cooperative data processing method in the embodiment of the present invention, the implementation of the central cloud can refer to the implementation of the method, and the repeated parts will not be repeated.

[0312] like Figure 6 As shown, the central cloud 600 includes:

[0313] The first receiving module 601 is configured to receive second fused traffic data sent by the edge cloud, wherein the second fused traffic data is obtained by fusing K groups of roadside perception data that meet confidence requirements, M groups of BSM data, and G groups of SSM data of the target road section, where K is an integer greater than 1, and M and G are both positive integers;

[0314] A determination module 602 is configured to determine traffic information of the target road segment based on the second fused traffic data;

[0315] The first sending module 603 is configured to send the traffic information to the edge cloud.

[0316] The central cloud provided by the embodiment of the present invention can execute Figure 4 The implementation principle and technical effects of the method embodiment shown are similar, and will not be described in detail here.

[0317] The central cloud 600 of an embodiment of the present invention receives second fused traffic data sent by the edge cloud, wherein the second fused traffic data is obtained by fusing K groups of roadside perception data that meet confidence requirements, M groups of BSM data, and G groups of SSM data for a target road section, where K is an integer greater than 1, and M and G are both positive integers; based on the second fused traffic data, determines traffic information for the target road section; and transmits the traffic information to the edge cloud. In this way, by determining the traffic information for the target road section based on the confidence-assessed multi-source fused traffic data sent by the edge cloud, the security and reliability of the vehicle-road collaborative information service can be ensured.

[0318] The embodiment of the present invention also provides a network side device. Since the principle of the network side device to solve the problem is similar to the vehicle-road cooperative data processing method in the embodiment of the present invention, the implementation of the network side device can refer to the implementation of the method, and the repeated parts will not be repeated. Figure 7 As shown, the network side device of the embodiment of the present invention includes:

[0319] Processor 700, when the network-side device is an edge cloud-side device, is configured to read the program in memory 720 and execute the following process:

[0320] Obtain N sets of roadside sensing data for the target road section, where N is a positive integer;

[0321] Obtain M groups of BSM data for the target road section, where M is a positive integer;

[0322] Performing a confidence judgment on the i-th group of roadside sensing data based on the i-th group of roadside sensing data and the M groups of BSM data, where i takes values from 1 to N in sequence;

[0323] Outputting K groups of roadside perception data whose confidence meets the requirement among the N groups of roadside perception data, where K is a positive integer less than or equal to N;

[0324] Based on the traffic data on the target road section, a vehicle-road collaborative information service is provided to traffic objects on the target road section, wherein the traffic data includes the K groups of roadside perception data.

[0325] The transceiver 710 is configured to receive and send data under the control of the processor 700 .

[0326] Among them, Figure 7 In the embodiment, the bus architecture may include any number of interconnected buses and bridges, specifically linking together various circuits of one or more processors represented by processor 700 and memory represented by memory 720. The bus architecture may also link together various other circuits such as peripheral devices, voltage regulators, and power management circuits, which are all well known in the art and therefore will not be described further herein. The bus interface provides an interface. The transceiver 710 may be a plurality of components, i.e., a transmitter and a transceiver, providing a unit for communicating with various other devices on a transmission medium. The processor 700 is responsible for managing the bus architecture and general processing, and the memory 720 may store data used by the processor 700 when performing operations.

[0327] Preferably, the processor 700 is further configured to read a program in the memory 720 and execute the following steps:

[0328] Calculating the confidence level of the i-th group of roadside perception data based on the i-th group of roadside perception data and the M groups of BSM data;

[0329] When the confidence level of the i-th group of roadside perception data is within a preset range, it is determined that the confidence level of the i-th group of roadside perception data meets the requirement.

[0330] Preferably, the preset range is greater than or equal to the confidence threshold and less than or equal to 1;

[0331] There are multiple groups of roadside sensing devices arranged according to their positions on the target road section, each group of roadside sensing data is collected by a group of roadside sensing devices on the target road section, and the confidence threshold is calculated based on the multiple groups of roadside sensing data collected by the multiple groups of roadside sensing devices within a historical time period, and the multiple groups of BSM data on the target road section within the historical time period.

[0332] Preferably, the processor 700 is further configured to read a program in the memory 720 and execute the following steps:

[0333] Calculating M confidence levels of the i-th group of roadside perception data based on the i-th group of roadside perception data and each group of BSM data in the M groups of BSM data;

[0334] If less than L confidence levels among the M confidence levels are not within the preset range, determining that the confidence level of the i-th group of roadside sensing data meets the requirement, where L is a preset value and L is a positive integer;

[0335] If more than L of the M confidence levels are outside the preset range, it is determined that the confidence level of the i-th group of roadside perception data does not meet the requirement, and output of the i-th group of roadside perception data is stopped.

[0336] Preferably, the processor 700 is further configured to read a program in the memory 720 and execute the following steps:

[0337] When it is calculated based on the i-th group of roadside perception data and the j-th group of BSM data that the j-th confidence level of the i-th group of roadside perception data is not within the preset range, and no more than L confidence levels among the M confidence levels are not within the preset range, it is determined that the confidence level of the j-th group of BSM data does not meet the requirements, and a warning message is sent to the target vehicle-mounted terminal corresponding to the j-th group of BSM data, where j is an integer between 1 and M.

[0338] Preferably, the processor 700 is further configured to read a program in the memory 720 and execute the following steps:

[0339] The i-th group of roadside perception data is input into a pre-established confidence correction model to obtain the i-th group of roadside perception data after correction output by the confidence correction model.

[0340] Preferably, the processor 700 is further configured to read a program in the memory 720 and execute the following steps:

[0341] Obtaining M′ groups of BSM data for the target road section, where M′ is a positive integer;

[0342] Calculating M′ confidence levels of the i-th group of roadside perception data based on the corrected i-th group of roadside perception data and each group of BSM data in the M′ groups of BSM data;

[0343] When more than L confidence levels among the M′ confidence levels are within the preset range, it is determined that the confidence level of the corrected i-th group of roadside perception data meets the requirement, and the corrected i-th group of roadside perception data is output.

[0344] Preferably, the confidence correction model is Y=c0X 5 +d0X 4 +e0X 3 +f0X 2 +g0X+h0, where X is the roadside perception data before correction, Y is the roadside perception data after correction, and c0, d0, e0, f0, g0, and h0 are model parameters.

[0345] Preferably, the processor 700 is further configured to read a program in the memory 720 and execute the following steps:

[0346] Performing a confidence judgment on the target roadside sensing device corresponding to the i-th group of roadside sensing data;

[0347] When the confidence level of the target roadside perception device does not meet the requirement, the roadside perception data collected by the target roadside perception device is corrected.

[0348] Preferably, a plurality of groups of roadside sensing devices are arranged according to positions on the target road section, each group of roadside sensing data is collected by a group of roadside sensing devices on the target road section, a group of roadside sensing devices includes a plurality of roadside sensors, and each group of roadside sensing data includes a plurality of sensor sensing data;

[0349] The processor 700 is further configured to read the program in the memory 720 and execute the following steps:

[0350] The plurality of sensor perception data in the i-th group of roadside perception data among the N groups of roadside perception data are fused to obtain the i-th group of fused roadside perception data.

[0351] Preferably, the processor 700 is further configured to read a program in the memory 720 and execute the following steps:

[0352] When K is greater than 1, the K groups of roadside perception data are fused to obtain global roadside fused perception data;

[0353] The traffic data includes the global roadside fusion perception data.

[0354] Preferably, the processor 700 is further configured to read a program in the memory 720 and execute the following steps:

[0355] Fusing the M groups of BSM data with the global roadside fusion perception data to obtain first fused traffic data;

[0356] The traffic data includes the first fused traffic data.

[0357] Preferably, the processor 700 is further configured to read a program in the memory 720 and execute the following steps:

[0358] Obtain G groups of SSM data for the target road section, where G is a positive integer;

[0359] fusing the G group of SSM data with the first fused traffic data to obtain second fused traffic data;

[0360] The traffic data includes the second fused traffic data.

[0361] Preferably, the processor 700 is further configured to read a program in the memory 720 and execute the following steps:

[0362] determining object information and motion data of traffic objects on the target road segment based on traffic data on the target road segment;

[0363] Based on the object information and motion data of the traffic object, the motion trajectory of the traffic object is predicted, or the risk rating of the traffic object is performed.

[0364] Preferably, the processor 700 is further configured to read a program in the memory 720 and execute the following steps:

[0365] Uploading the second fused traffic data to the central cloud via the transceiver 710;

[0366] Receive the traffic information of the target road section sent by the central cloud through the transceiver 710;

[0367] The traffic information is sent to traffic objects on the target road section via the transceiver 710 .

[0368] When the network-side device is a central cloud-side device, the processor 700 is configured to read the program in the memory 720 and execute the following process:

[0369] Receiving, through the transceiver 710, second fused traffic data sent by the edge cloud, wherein the second fused traffic data is obtained by fusing K groups of roadside perception data that meet confidence requirements, M groups of BSM data, and G groups of SSM data of the target road section, where K is an integer greater than 1, and M and G are both positive integers;

[0370] determining traffic information of the target road section based on the second fused traffic data;

[0371] The traffic information is sent to the edge cloud via the transceiver 710 .

[0372] The network side device provided in the embodiment of the present invention can execute the above method embodiment, and its implementation principle and technical effects are similar, which will not be repeated in this embodiment.

[0373] In addition, the computer-readable storage medium of the embodiment of the present invention is used to store a computer program, and the computer program can be executed by a processor to implement Figure 1 Each step in the method embodiment shown, or implementation as follows Figure 4 The various steps in the method embodiment are shown.

[0374] In the several embodiments provided in this application, it should be understood that the disclosed methods and devices can be implemented in other ways. For example, the device embodiments described above are merely schematic. For example, the division of the units is merely a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection of some interfaces, devices or units, which can be electrical, mechanical or other forms.

[0375] In addition, the functional units in the various embodiments of the present invention may be integrated into a single processing unit, each unit may be physically included separately, or two or more units may be integrated into a single unit. The aforementioned integrated units may be implemented in the form of hardware or hardware plus software functional units.

[0376] The above-mentioned integrated unit implemented in the form of a software functional unit can be stored in a computer-readable storage medium. The above-mentioned software functional unit is stored in a storage medium and includes a number of instructions for causing a computer device (which can be a personal computer, server, or network device, etc.) to perform some steps of the sending and receiving methods described in various embodiments of the present invention. The aforementioned storage medium includes: a USB flash drive, a mobile hard drive, ROM (Read-Only Memory), RAM (Random Access Memory), a magnetic disk, or an optical disk, etc., various media that can store program code.

[0377] The above is a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications should also be regarded as within the scope of protection of the present invention.

Claims

1. A vehicle-road cooperative data processing method, characterized in that: include: Obtain N sets of roadside sensing data for the target road section, where N is a positive integer; Obtaining M groups of basic safety message (BSM) data for the target road section, where M is a positive integer; Performing a confidence judgment on the i-th group of roadside sensing data based on the i-th group of roadside sensing data and the M groups of BSM data, where i takes values from 1 to N in sequence; Outputting K groups of roadside perception data whose confidence meets the requirement among the N groups of roadside perception data, where K is a positive integer less than or equal to N; Based on the traffic data on the target road section, providing a vehicle-road cooperative information service to traffic objects on the target road section, wherein the traffic data includes the K groups of roadside perception data; The step of performing confidence judgment on the i-th group of roadside perception data based on the i-th group of roadside perception data and the M groups of BSM data includes: Calculating the confidence level of the i-th group of roadside perception data based on the i-th group of roadside perception data and the M groups of BSM data; When the confidence level of the i-th group of roadside sensing data is within a preset range, determining that the confidence level of the i-th group of roadside sensing data meets the requirement; The calculating, based on the i-th group of roadside perception data and the M groups of BSM data, the confidence level of the i-th group of roadside perception data includes: Calculating M confidence levels of the i-th group of roadside perception data based on the i-th group of roadside perception data and each group of BSM data in the M groups of BSM data; When the confidence level of the i-th group of roadside perception data is within a preset range, determining that the confidence level of the i-th group of roadside perception data meets the requirement includes: If less than L confidence levels among the M confidence levels are not within the preset range, determining that the confidence level of the i-th group of roadside sensing data meets the requirement, where L is a preset value and L is a positive integer; After calculating the M confidence levels of the i-th group of roadside perception data, the method further includes: If more than L of the M confidence levels are outside the preset range, it is determined that the confidence level of the i-th group of roadside perception data does not meet the requirement, and output of the i-th group of roadside perception data is stopped.

2. The method according to claim 1, characterized in that The preset range is greater than or equal to the confidence threshold and less than or equal to 1; There are multiple groups of roadside sensing devices arranged according to their positions on the target road section, each group of roadside sensing data is collected by a group of roadside sensing devices on the target road section, and the confidence threshold is calculated based on the multiple groups of roadside sensing data collected by the multiple groups of roadside sensing devices within a historical time period, and the multiple groups of BSM data on the target road section within the historical time period.

3. The method according to claim 1, characterized in that After calculating the M confidence levels of the i-th group of roadside perception data, the method further includes: When it is calculated based on the i-th group of roadside perception data and the j-th group of BSM data that the j-th confidence level of the i-th group of roadside perception data is not within the preset range, and no more than L confidence levels among the M confidence levels are not within the preset range, it is determined that the confidence level of the j-th group of BSM data does not meet the requirements, and a warning message is sent to the target vehicle-mounted terminal corresponding to the j-th group of BSM data, where j is an integer between 1 and M.

4. The method according to claim 1, wherein After determining that the confidence level of the i-th group of roadside perception data does not meet the requirement and stopping outputting the i-th group of roadside perception data, the method further includes: The i-th group of roadside perception data is input into a pre-established confidence correction model to obtain the i-th group of roadside perception data after correction output by the confidence correction model.

5. The method according to claim 4, characterized in that The confidence correction model is Y=c0X 5 +d0X 4 +e0X 3 +f0X 2 +g0X+h0, where X is the roadside perception data before correction, Y is the roadside perception data after correction, and c0, d0, e0, f0, g0, and h0 are model parameters.

6. The method according to claim 1, characterized in that After determining that the confidence level of the i-th group of roadside perception data does not meet the requirement and stopping outputting the i-th group of roadside perception data, the method further includes: Performing a confidence judgment on the target roadside sensing device corresponding to the i-th group of roadside sensing data; When the confidence level of the target roadside perception device does not meet the requirement, the roadside perception data collected by the target roadside perception device is corrected.

7. The method according to claim 1, characterized in that A plurality of groups of roadside sensing devices are arranged on the target road section according to their positions, each group of roadside sensing data is collected by a group of roadside sensing devices on the target road section, a group of roadside sensing devices includes a plurality of roadside sensors, and each group of roadside sensing data includes a plurality of sensor sensing data; After acquiring N groups of roadside perception data of the target road section, and before performing confidence judgment on the i-th group of roadside perception data based on the i-th group of roadside perception data and the M groups of BSM data, the method further includes: The plurality of sensor perception data in the i-th group of roadside perception data among the N groups of roadside perception data are fused to obtain the i-th group of fused roadside perception data.

8. The method according to claim 1, characterized in that After outputting K groups of roadside perception data whose confidences meet the requirements among the N groups of roadside perception data, the method further includes: When K is greater than 1, the K groups of roadside perception data are fused to obtain global roadside fused perception data; The traffic data includes the global roadside fusion perception data.

9. The method according to claim 8, characterized in that After fusing the K groups of roadside sensing data, the method further includes: Fusing the M groups of BSM data with the global roadside fusion perception data to obtain first fused traffic data; The traffic data includes the first fused traffic data.

10. The method according to claim 9, characterized in that The method further comprises: Obtain G group of perception sharing message SSM data of the target road section, where G is a positive integer; After fusing the M groups of BSM data with the global roadside fusion perception data, the method further includes: fusing the G group of SSM data with the first fused traffic data to obtain second fused traffic data; The traffic data includes the second fused traffic data.

11. The method according to any one of claims 1 to 10, characterized in that The providing of a vehicle-road cooperative information service to traffic objects on the target road section based on the traffic data on the target road section includes: determining object information and motion data of traffic objects on the target road segment based on traffic data on the target road segment; Based on the object information and motion data of the traffic object, the motion trajectory of the traffic object is predicted, or the risk rating of the traffic object is performed.

12. The method according to claim 10, characterized in that After fusing the G group SSM data with the first fused traffic data to obtain second fused traffic data, the method further includes: Uploading the second fused traffic data to the central cloud; Receiving traffic information of the target road section sent by the central cloud; The traffic information is sent to traffic objects on the target road section.

13. An edge cloud, characterized in that: include: A first acquisition module is used to acquire N groups of roadside sensing data of a target road section, where N is a positive integer; A second acquisition module is configured to acquire M groups of basic safety message (BSM) data of the target road section, where M is a positive integer; A first processing module is configured to perform a confidence judgment on the i-th group of roadside sensing data based on the i-th group of roadside sensing data and the M groups of BSM data, where i is 1 to N in sequence; an output module, configured to output K groups of roadside perception data whose confidences meet the requirements among the N groups of roadside perception data, where K is a positive integer less than or equal to N; an execution module, configured to provide a vehicle-road cooperative information service to traffic objects on the target road section based on traffic data on the target road section, wherein the traffic data includes the K groups of roadside perception data; The first processing module includes: a calculation submodule, configured to calculate the confidence level of the i-th group of roadside perception data based on the i-th group of roadside perception data and the M groups of BSM data; A first determining submodule, configured to determine whether the confidence level of the i-th group of roadside sensing data meets a requirement when the confidence level of the i-th group of roadside sensing data is within a preset range; The calculation submodule includes: a calculation unit, configured to calculate M confidence levels of the i-th group of roadside perception data based on the i-th group of roadside perception data and each group of BSM data in the M groups of BSM data; The first determination submodule is configured to determine that the confidence level of the i-th group of roadside sensing data meets the requirement when less than L confidence levels among the M confidence levels are not within the preset range, where L is a preset value and is a positive integer; The edge cloud further includes: The second processing module is used to determine that the confidence level of the i-th group of roadside perception data does not meet the requirements and stop outputting the i-th group of roadside perception data when more than L confidence levels among the M confidence levels are not within the preset range.

14. A network-side device, comprising: A transceiver, a memory, a processor, and a computer program stored in the memory and executable on the processor; wherein the processor is configured to read the program in the memory to implement the steps of the vehicle-road cooperative data processing method as described in any one of claims 1 to 12.

15. A computer-readable storage medium for storing a computer program, characterized in that: When the computer program is executed by a processor, the steps in the vehicle-road cooperative data processing method as described in any one of claims 1 to 12 are implemented.

Images