Edge calculation method and system based on Internet of Vehicles TSP and medium

Through edge computing data layering and multimodal optimization factor data stitching algorithm, the accuracy and efficiency of multi-source data processing in the Internet of Vehicles are solved, accurate data alignment and fusion are achieved, and storage and transmission costs are reduced. Suitable for a variety of IoT applications.

CN119052292APending Publication Date: 2024-11-29GUANGZHOU HAIPERTE TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411263096.0
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-09-10
Publication Date
2024-11-29

AI Technical Summary

Technical Problem

In the Internet of Vehicles environment, existing data access methods are difficult to efficiently handle diversified data needs, especially when the data acquisition mechanism, frequency and delay are different, and the efficient processing capability of real-time data is lacking, resulting in data accuracy and Inefficient efficiency.

Method used

Using edge computing method based on Internet of Vehicles TSP, multi-source data is processed and calibrated through data hierarchical algorithms and multi-modal optimization factors to ensure that data from different sources are accurately aligned in time and process missing values , outliers and inconsistencies, combined with real-time upload and offline exchange, reduce network transmission efficiency and cost.

Benefits of technology

It realizes accurate alignment and fusion of multi-source data, improves data accuracy and visual retrieval capabilities, reduces storage and transmission costs, is suitable for large-scale data processing, and is suitable for Internet of Vehicles and other Internet of Things application scenarios.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119052292A_ABST
    Figure CN119052292A_ABST
Patent Text Reader

Abstract

The invention relates to an edge calculation method and system based on an Internet of Vehicles TSP and a medium, and the method comprises the steps: M1, collecting the CAN bus data information and clock synchronization log data information of a whole vehicle when the vehicle runs on a road, obtaining the image data information of the road in real time based on a vehicle-mounted camera, and obtaining the image data information of the vehicle; acquiring point cloud data information of a road in real time based on a vehicle-mounted laser radar; and M2, based on the CAN bus data information of the whole vehicle, the clock synchronization log data information, the image data information of the road and the point cloud data information of the road, processing multi-source data by adopting a data layering algorithm of edge calculation. The method not only solves the problem of accurate time alignment of network connection data and offline mass data from different sources under the condition of different data acquisition mechanisms, frequencies and delays, but also solves the problems of missing value processing, abnormal value processing and inconsistency, so as to ensure the accuracy of a fusion result.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of vehicle networking technology, and in particular to an edge computing method, system and medium based on vehicle networking TSP. Background Art

[0002] With the rapid development of intelligent transportation and Internet of Things technologies, massive amounts of vehicle and road condition data are generated in the Internet of Vehicles environment. How to efficiently access and process this data has become an urgent problem to be solved.

[0003] Existing data access methods mostly focus on processing a single data source, which makes it difficult to cope with diverse data needs. In addition, existing methods are insufficient in data storage and retrieval efficiency, and lack the ability to efficiently process real-time data. Summary of the invention

[0004] In view of the above problems, the present invention provides an edge computing method, system and medium based on the Internet of Vehicles TSP, which not only solves the accurate time alignment of networked data from different sources and offline massive data under different data collection mechanisms, frequencies and delays, but also handles the problems of missing values, outliers and inconsistencies to ensure the accuracy of the fusion results.

[0005] In order to achieve the above-mentioned purpose and other related purposes, the technical solution provided by the present invention is as follows:

[0006] An edge computing method based on vehicle networking TSP, the method comprising:

[0007] M1. The vehicle is driving on the road, collecting the CAN bus data information and clock synchronization log data information of the whole vehicle, and obtaining the image data information of the road in real time based on the on-board camera, and obtaining the point cloud data information of the road in real time based on the on-board laser radar;

[0008] M2. Based on the CAN bus data information of the whole vehicle, the clock synchronization log data information, the image data information of the road and the point cloud data information of the road, the multi-source data is processed using the data layering algorithm of edge computing to obtain the processed CAN bus data of the whole vehicle, the clock synchronization log data, the image data of the road and the point cloud data information of the road;

[0009] M3. Based on the processed vehicle CAN bus data, clock synchronization log data, road image and road point cloud data information, a data stitching algorithm of edge computing based on a multimodal optimization factor is used to stitch the multi-source data to obtain stitched multi-source data information;

[0010] M4. Based on the stitched multi-source data information, establish a multi-source data calibration function F, calibrate the multi-source data, and output the calibrated multi-source data information.

[0011] Furthermore, in step M2, the data layering algorithm using edge computing is used to process multi-source data, including:

[0012] M21. Based on the CAN bus data information of the whole vehicle, the clock synchronization log data information, the image data information of the road and the point cloud data information of the road, establish an embedding function G of multi-source data,

[0013]

[0014] Among them, x1 is the CAN bus data information of the whole vehicle, x2 is the clock synchronization log data information, x3 is the image data information of the road, x4 is the point cloud data information of the road, α1, α2 and α3 are the embedding factors of the multi-source data, and the embedding matrix of the multi-source data is characterized to obtain the embedding matrix data information of the multi-source data;

[0015] M22. Based on the embedded matrix data information of the multi-source data, establish a data stratification function H of the multi-source data,

[0016]

[0017] Among them, y is the embedded matrix data information of multi-source data, β1, β2 and β3 are the stratification determining factors of multi-source data, and the multi-source data is stratified to obtain the stratified multi-source data information;

[0018] M23. Based on the layered multi-source data information, a disk array of multi-source data is constructed, and the multi-source data is processed to obtain processed vehicle CAN bus data, clock synchronization log data, road images and road point cloud data information.

[0019] Furthermore, the constraints of the multi-source data stratification determining factors β1, β2 and β3 are:

[0020]

[0021] Furthermore, the embedding factors α1, α2 and α3 of the multi-source data are respectively,

[0022]

[0023] Among them, x1 is the CAN bus data information of the whole vehicle, x2 is the clock synchronization log data information, x3 is the image data information of the road, and x4 is the point cloud data information of the road.

[0024] Furthermore, in step M3, stitching the multi-source data using the edge computing data stitching algorithm based on the multimodal optimization factor includes:

[0025] M31. Based on the processed vehicle CAN bus data, clock synchronization log data, road image and road point cloud data information, normalization processing is performed to obtain normalized vehicle CAN bus data, clock synchronization log data, road image and road point cloud data information;

[0026] M32. Based on the normalized vehicle CAN bus data, clock synchronization log data, road image and road point cloud data information, a multi-source data stitching function Q based on a multimodal optimization factor is established:

[0027]

[0028] Among them, z1 is the normalized vehicle CAN bus data information, z2 is the normalized clock synchronization log data information, z3 is the normalized road image data information, z4 is the normalized road point cloud data information, δ1, δ2 and δ3 are multimodal optimization factors;

[0029] M33. Based on the multi-source data fusion function Q based on the multi-modal optimization factor, the multi-source data are stitched to obtain stitched multi-source data information.

[0030] Furthermore, the relationship function of the multimodal optimization factors δ1, δ2 and δ3 is W,

[0031] Among them, the constraint condition of the relationship function W is (2,3).

[0032] Furthermore, the multimodal optimization factor δ1 is,

[0033]

[0034] The multimodal optimization factor δ2 is,

[0035]

[0036] The multimodal optimization factor δ3 is,

[0037]

[0038] Among them, z1 is the normalized CAN bus data information of the whole vehicle, z2 is the normalized clock synchronization log data information, z3 is the normalized road image data information, and z4 is the normalized road point cloud data information.

[0039] Furthermore, the multi-source data calibration function F is:

[0040]

[0041] Among them, r is the multi-source data information after stitching, and θ1, θ2 and θ3 are the weight coefficient factors of multi-source data.

[0042] In order to achieve the above-mentioned objectives and other related objectives, the present invention also provides an edge computing system based on the Internet of Vehicles TSP, including a computer device, which is programmed or configured to execute any step of the edge computing method based on the Internet of Vehicles TSP.

[0043] In order to achieve the above-mentioned objectives and other related objectives, the present invention also provides a computer-readable storage medium, which stores a computer program programmed or configured to execute any one of the edge computing methods based on the Internet of Vehicles TSP.

[0044] The present invention has the following positive effects:

[0045] 1. The present invention processes multi-source data by adopting the data layering algorithm of edge computing, and stitches multi-source data in combination with the data stitching algorithm of edge computing based on multimodal optimization factors, which not only solves the accurate time alignment of networked data from different sources and offline massive data under different data acquisition mechanisms, frequencies and delays, but also handles the problems of missing values, outliers and inconsistencies to ensure the accuracy of the fusion results, and the layered access module combines real-time upload with offline exchange, especially when processing large-scale data, which can greatly reduce network transmission efficiency and transmission costs through low-cost storage resources. And this layered upload strategy is very suitable for a variety of application scenarios, not only in the field of Internet of Vehicles, but also in industrial Internet of Things, smart cities, etc.

[0046] 2. The present invention calibrates multi-source data by establishing a multi-source data calibration function F, which can not only ensure that the data of different types of sensors are aligned in time, reduce errors in data fusion, and improve the accuracy of data in subsequent annotation and simulation links, but also solve the problems of data visualization retrieval and packaging (timestamp packaging method) downloading from different sources, and serve subsequent data annotation and other businesses. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 It is a schematic diagram of the method flow of the present invention;

[0048] Figure 2 A schematic diagram of a flow chart of an algorithm for data stratification of edge computing of the present invention;

[0049] Figure 3 It is a flow chart of a data stitching algorithm based on edge computing of a multimodal optimization factor of the present invention;

[0050] Figure 4 A schematic diagram of hierarchical data access according to the present invention;

[0051] Figure 5 A schematic diagram of data stitching of the present invention;

[0052] Figure 6 A data flow diagram of the data package download service of the present invention. DETAILED DESCRIPTION

[0053] The following is a description of exemplary embodiments of the present disclosure in conjunction with the accompanying drawings, including various details of the embodiments of the present disclosure to facilitate understanding, which should be considered as merely exemplary. Therefore, it should be recognized by those of ordinary skill in the art that various changes and modifications may be made to the embodiments described herein without departing from the scope and spirit of the present disclosure. Similarly, for the sake of clarity and conciseness, descriptions of well-known functions and structures are omitted in the following description.

[0054] Embodiment 1: as 1 or Figure 4 or Figure 5 or Figure 6 As shown, an edge computing method based on the Internet of Vehicles TSP, the method comprising:

[0055] M1. The vehicle is driving on the road, collecting the CAN bus data information and clock synchronization log data information of the whole vehicle, and obtaining the image data information of the road in real time based on the on-board camera, and obtaining the point cloud data information of the road in real time based on the on-board laser radar;

[0056] M2. Based on the CAN bus data information of the whole vehicle, the clock synchronization log data information, the image data information of the road and the point cloud data information of the road, the multi-source data is processed using the data layering algorithm of edge computing to obtain the processed CAN bus data of the whole vehicle, the clock synchronization log data, the image data of the road and the point cloud data information of the road;

[0057] M3. Based on the processed vehicle CAN bus data, clock synchronization log data, road image and road point cloud data information, a data stitching algorithm of edge computing based on a multimodal optimization factor is used to stitch the multi-source data to obtain stitched multi-source data information;

[0058] M4. Based on the stitched multi-source data information, establish a multi-source data calibration function F, calibrate the multi-source data, and output the calibrated multi-source data information.

[0059] In this embodiment, if Figure 2 As shown, in step M2, the data layering algorithm using edge computing is used to process multi-source data, including:

[0060] M21. Based on the CAN bus data information of the whole vehicle, the clock synchronization log data information, the image data information of the road and the point cloud data information of the road, establish an embedding function G of multi-source data,

[0061]

[0062] Among them, x1 is the CAN bus data information of the whole vehicle, x2 is the clock synchronization log data information, x3 is the image data information of the road, x4 is the point cloud data information of the road, α1, α2 and α3 are the embedding factors of the multi-source data, and the embedding matrix of the multi-source data is characterized to obtain the embedding matrix data information of the multi-source data;

[0063] M22. Based on the embedded matrix data information of the multi-source data, establish a data stratification function H of the multi-source data,

[0064]

[0065] Among them, y is the embedded matrix data information of multi-source data, β1, β2 and β3 are the stratification determining factors of multi-source data, and the multi-source data is stratified to obtain the stratified multi-source data information;

[0066] M23. Based on the layered multi-source data information, a disk array of multi-source data is constructed, and the multi-source data is processed to obtain processed vehicle CAN bus data, clock synchronization log data, road images and road point cloud data information.

[0067] In this embodiment, the constraints of the multi-source data stratification determining factors β1, β2 and β3 are:

[0068]

[0069] In this embodiment, the embedding factors α1, α2 and α3 of the multi-source data are respectively:

[0070] Among them, x1 is the CAN bus data information of the whole vehicle, x2 is the clock synchronization log data information, x3 is the image data information of the road, and x4 is the point cloud data information of the road.

[0071] In this embodiment, a visualization tool (such as Grafana, Tableau, or a custom web application) is used to display the fused data. The visualization content can be divided into different panels to display video streams, radar data, and message information. Users can use the quick query interface based on the time range to obtain all data packets returned by the system within the time period.

[0072] Example 2: Based on the edge computing method based on the Internet of Vehicles TSP in Example 1, the present invention is further illustrated and described below.

[0073] Such as 1 or Figure 4 or Figure 5 or Figure 6 As shown, an edge computing method based on the Internet of Vehicles TSP, the method comprising:

[0074] M1. The vehicle is driving on the road, collecting the CAN bus data information and clock synchronization log data information of the whole vehicle, and obtaining the image data information of the road in real time based on the on-board camera, and obtaining the point cloud data information of the road in real time based on the on-board laser radar;

[0075] M2. Based on the CAN bus data information of the whole vehicle, the clock synchronization log data information, the image data information of the road and the point cloud data information of the road, the multi-source data is processed using the data layering algorithm of edge computing to obtain the processed CAN bus data of the whole vehicle, the clock synchronization log data, the image data of the road and the point cloud data information of the road;

[0076] M3. Based on the processed vehicle CAN bus data, clock synchronization log data, road image and road point cloud data information, a data stitching algorithm of edge computing based on a multimodal optimization factor is used to stitch the multi-source data to obtain stitched multi-source data information;

[0077] M4. Based on the stitched multi-source data information, establish a multi-source data calibration function F, calibrate the multi-source data, and output the calibrated multi-source data information.

[0078] In this embodiment, if Figure 3 As shown, in step M3, stitching multi-source data using a data stitching algorithm based on edge computing of a multimodal optimization factor includes:

[0079] M31. Based on the processed vehicle CAN bus data, clock synchronization log data, road image and road point cloud data information, normalization processing is performed to obtain normalized vehicle CAN bus data, clock synchronization log data, road image and road point cloud data information;

[0080] M32. Based on the normalized vehicle CAN bus data, clock synchronization log data, road image and road point cloud data information, a multi-source data stitching function Q based on a multimodal optimization factor is established:

[0081]

[0082] Among them, z1 is the normalized vehicle CAN bus data information, z2 is the normalized clock synchronization log data information, z3 is the normalized road image data information, z4 is the normalized road point cloud data information, δ1, δ2 and δ3 are multimodal optimization factors;

[0083] M33. Based on the multi-source data fusion function Q based on the multi-modal optimization factor, the multi-source data are stitched to obtain stitched multi-source data information.

[0084] In this embodiment, the relationship function of the multimodal optimization factors δ1, δ2 and δ3 is W,

[0085]

[0086] Among them, the constraint condition of the relationship function W is (2,3).

[0087] In this embodiment, the multimodal optimization factor δ1 is,

[0088]

[0089] The multimodal optimization factor δ2 is,

[0090]

[0091] The multimodal optimization factor δ3 is,

[0092]

[0093] Among them, z1 is the normalized CAN bus data information of the whole vehicle, z2 is the normalized clock synchronization log data information, z3 is the normalized road image data information, and z4 is the normalized road point cloud data information.

[0094] In this embodiment, the data stitching service is responsible for analyzing and calibrating the video, radar and other data collected by the entire vehicle based on the network application message timeline and combined with the clock synchronization log to ensure that the application message at a certain moment can match the video and radar original data at the corresponding moment.

[0095] The detailed steps are as follows:

[0096] 1) Parse the PTP / NTP clock synchronization log, obtain the key-value relationship between the application message timestamp and the video and radar data timestamp, and send the application message and clock synchronization data to different topics of Kafka.

[0097] 2) Use Flink to consume key-value relationships from Kafka and obtain the original file index objects of video and radar data by timestamp.

[0098] 3) Obtain the additional timestamp in the original data packet through frame sampling, compare the difference between the data packet timestamp and the key-value, if (application message timestamp - X milliseconds) ≤ data packet timestamp ≤ application message timestamp, mark it as a normal data packet, otherwise, mark it as an abnormal data packet.

[0099] X milliseconds can be customized based on data scenarios. The common scenario reference range is as follows:

[0100] a. Fast scenarios (such as autonomous driving): the recommended time interval is between 10-100 milliseconds;

[0101] b. General scenarios (such as intelligent transportation and monitoring): the recommended time interval is between 100-500 milliseconds;

[0102] c. Complex scenarios (more factors need to be considered): The recommended time range is between 500 milliseconds and 1 second.

[0103] 4) The calibrated data packets and timestamp key-values ​​are marked and stored in a database (such as a time series database, NoSQL database, etc.) according to the following data structure to facilitate subsequent rapid retrieval.

[0104] Select an appropriate time window (such as the 10-1000 milliseconds mentioned above) according to the characteristics of different data sources. For each data source, keep the most recent frames of data for subsequent fusion work.

[0105] In this embodiment, the multi-source data calibration function F is:

[0106]

[0107] Among them, r is the multi-source data information after stitching, and θ1, θ2 and θ3 are the weight coefficient factors of multi-source data.

[0108] In this embodiment, the present invention provides an edge computing system based on the Internet of Vehicles TSP, including a computer device, which is programmed or configured to execute any one of the steps of the edge computing method based on the Internet of Vehicles TSP.

[0109] In this embodiment, the present invention provides a computer-readable storage medium, on which is stored a computer program programmed or configured to execute any one of the edge computing methods based on the Internet of Vehicles TSP.

[0110] Any reference to memory, storage, database or other media used in the embodiments provided in the present application may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM) and memory bus dynamic RAM (RDRAM).

[0111] In summary, the present invention not only solves the problem of accurate temporal alignment of networked data from different sources and offline massive data under different data collection mechanisms, frequencies and delays, but also handles the problems of missing values, outliers and inconsistencies to ensure the accuracy of the fusion results.

[0112] The above specific implementations do not constitute a limitation on the protection scope of the present disclosure. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present disclosure shall be included in the protection scope of the present disclosure.

Claims

1. An edge computing method based on TSP of Internet of Vehicles, characterized in that: The method comprises: M1. The vehicle is driving on the road, collecting the CAN bus data information and clock synchronization log data information of the whole vehicle, and obtaining the image data information of the road in real time based on the on-board camera, and obtaining the point cloud data information of the road in real time based on the on-board laser radar; M2. Based on the CAN bus data information of the whole vehicle, the clock synchronization log data information, the image data information of the road and the point cloud data information of the road, the multi-source data is processed using the data layering algorithm of edge computing to obtain the processed CAN bus data of the whole vehicle, the clock synchronization log data, the image data of the road and the point cloud data information of the road; M3. Based on the processed vehicle CAN bus data, clock synchronization log data, road image and road point cloud data information, a data stitching algorithm of edge computing based on a multimodal optimization factor is used to stitch the multi-source data to obtain stitched multi-source data information; M4. Based on the stitched multi-source data information, establish a multi-source data calibration function F, calibrate the multi-source data, and output the calibrated multi-source data information.

2. The edge computing method based on the Internet of Vehicles TSP according to claim 1 is characterized in that: In step M2, the data layering algorithm using edge computing is used to process multi-source data, including: M21. Based on the CAN bus data information of the whole vehicle, the clock synchronization log data information, the image data information of the road and the point cloud data information of the road, establish an embedding function G of multi-source data, Among them, x1 is the CAN bus data information of the whole vehicle, x2 is the clock synchronization log data information, x3 is the image data information of the road, x4 is the point cloud data information of the road, α1, α2 and α3 are the embedding factors of the multi-source data, and the embedding matrix of the multi-source data is characterized to obtain the embedding matrix data information of the multi-source data; M22. Based on the embedded matrix data information of the multi-source data, establish a data stratification function H of the multi-source data, Among them, y is the embedded matrix data information of multi-source data, β1, β2 and β3 are the stratification determining factors of multi-source data, and the multi-source data is stratified to obtain the stratified multi-source data information; M23. Based on the layered multi-source data information, a disk array of multi-source data is constructed, and the multi-source data is processed to obtain processed vehicle CAN bus data, clock synchronization log data, road images and road point cloud data information.

3. The edge computing method based on the Internet of Vehicles TSP according to claim 2 is characterized in that: The constraints of the multi-source data stratification determining factors β1, β2 and β3 are:

4. The edge computing method based on the Internet of Vehicles TSP according to claim 2 is characterized in that: The embedding factors α1, α2 and α3 of the multi-source data are respectively: Among them, x1 is the CAN bus data information of the whole vehicle, x2 is the clock synchronization log data information, x3 is the image data information of the road, and x4 is the point cloud data information of the road.

5. The edge computing method based on the Internet of Vehicles TSP according to claim 1 is characterized in that: In step M3, stitching multi-source data using a data stitching algorithm based on edge computing of a multimodal optimization factor includes: M31. Based on the processed vehicle CAN bus data, clock synchronization log data, road image and road point cloud data information, normalization processing is performed to obtain normalized vehicle CAN bus data, clock synchronization log data, road image and road point cloud data information; M32. Based on the normalized vehicle CAN bus data, clock synchronization log data, road image and road point cloud data information, a multi-source data stitching function Q based on a multi-modal optimization factor is established. Among them, z1 is the normalized vehicle CAN bus data information, z2 is the normalized clock synchronization log data information, z3 is the normalized road image data information, z4 is the normalized road point cloud data information, δ1, δ2 and δ3 are multimodal optimization factors; M33. Based on the multi-source data fusion function Q based on the multi-modal optimization factor, the multi-source data are stitched to obtain stitched multi-source data information.

6. The edge computing method based on the Internet of Vehicles TSP according to claim 5 is characterized in that: The relationship function of the multimodal optimization factors δ1, δ2 and δ3 is W, Among them, the constraint condition of the relationship function W is (2,3).

7. The edge computing method based on the Internet of Vehicles TSP according to claim 5 is characterized in that: The multimodal optimization factor δ1 is, The multimodal optimization factor δ2 is, The multimodal optimization factor δ3 is, Among them, z1 is the normalized CAN bus data information of the whole vehicle, z2 is the normalized clock synchronization log data information, z3 is the normalized road image data information, and z4 is the normalized road point cloud data information.

8. The edge computing method based on the Internet of Vehicles TSP according to claim 1 is characterized in that: The multi-source data calibration function F is: Among them, r is the multi-source data information after stitching, and θ1, θ2 and θ3 are the weight coefficient factors of the multi-source data.

9. An edge computing system based on the Internet of Vehicles TSP, comprising a computer device, characterized in that: The computer device is programmed or configured to execute the steps of the edge computing method based on the Internet of Vehicles TSP as described in any one of claims 1 to 8.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program that is programmed or configured to execute the edge computing method based on the Internet of Vehicles TSP as described in any one of claims 1 to 8.