CN-122024485-A - Information acquisition method based on vehicle road cloud system

CN122024485ACN 122024485 ACN122024485 ACN 122024485ACN-122024485-A

Abstract

The invention provides an information acquisition method based on a vehicle road cloud system, which comprises the steps that a road side unit broadcasts road side perception information to a vehicle-mounted unit, the vehicle-mounted unit captures physical layer signal characteristics and CAN bus data of the road side perception information, cloud fusion data are used for positioning a vehicle, the vehicle-mounted unit acquires environment, gesture, CAN bus and communication channel data and uploads the cloud, the cloud extracts the characteristics and models the communication channel state, the cloud recognizes driver intention under event triggering to obtain driver intention data, the cloud fuses the driver intention, the communication channel state and the vehicle position, utility evaluation is carried out on the road side perception information, and information utility thermodynamic diagram is generated so as to optimize road side unit deployment or information release strategies. According to the method, quantitative evaluation of the effectiveness of the road side information is realized through multi-source data fusion and intention recognition, and data support is provided for optimal deployment and accurate service of the vehicle road cloud system.

Inventors

NIU CHAO
WU HAO
LI ZELONG
TAO YUNFEI
CHEN ZHUO
ZHENG HONG
GAO YUDE
ZHANG KENIU
Wang Yongchuang
ZHANG YUNLONG

Assignees

长春汽车检测中心有限责任公司

Dates

Publication Date: 20260512
Application Date: 20260318

Claims (9)

1. The utility model provides an information acquisition method based on car way cloud system, its characterized in that, car way cloud system includes vehicle, road side unit and high in the clouds, be provided with on-vehicle unit on the vehicle, on-vehicle unit carries out wireless data interaction with road side unit, high in the clouds respectively, the method includes: S101, broadcasting road side perception information to a vehicle-mounted unit by a road side unit in the running process of the vehicle, capturing physical layer signal characteristics of the road side perception information and CAN bus data of the vehicle by the vehicle-mounted unit, and positioning the position of the vehicle by a cloud based on the road side perception information, the physical layer signal characteristics and the CAN bus data; S102, acquiring vehicle surrounding environment data, vehicle posture data, CAN bus data and communication channel data between the vehicle-mounted unit and a road side unit by a vehicle-mounted unit, summarizing the data into a multisource data packet uploading cloud, extracting characteristics of the multisource data packet by the cloud, and modeling a communication channel state; S103, the cloud identifies the intention of the driver under the triggering of the event to obtain intention data of the driver; s104, the cloud end fuses the driver intention data, the communication channel state and the vehicle position, carries out utility evaluation on the road side perception information, generates an information utility thermodynamic diagram according to the utility evaluation result, and optimizes road side unit deployment or information release strategies.
2. The vehicle road cloud system-based information acquisition method according to claim 1, wherein the S101 specifically includes the following operations: s201, periodically broadcasting road side perception information to the outside by a road side unit, wherein the road side perception information comprises signal lamp information and road section traffic information; S202, when a vehicle enters a broadcasting range of a road side unit, a vehicle-mounted unit continuously monitors the broadcasting of the road side unit, extracts physical layer signal characteristics for signals successfully received by each frame, wherein the physical layer signal characteristics comprise arrival time, arrival angle, channel impulse response, road side unit identity and position information; S203, the vehicle-mounted unit acquires CAN bus data of the vehicle, acquires vehicle acceleration information through the inertia measurement unit, and uploads physical layer signal characteristics, CAN bus data and vehicle acceleration information to the cloud; s204, the cloud end estimates the position of the vehicle relative to the road side unit in real time based on physical layer signal characteristics, CAN bus data and vehicle acceleration information.
3. The information acquisition method based on the vehicle-road cloud system according to claim 2, wherein when the vehicle-mounted units on a plurality of vehicles simultaneously receive the broadcast of the same road side unit, the cloud terminal gathers the physical layer signal characteristics extracted by each vehicle, and calibrates the position of the road side unit through the cooperative optimization of the vehicles, and specifically comprises the following operations: s301, according to the physical layer signal characteristics uploaded by each vehicle, vehicles which receive the broadcast of the same road side unit in the same time period are screened out, and all the screened vehicles are recorded as collaborative optimization vehicles; s302, constructing a nonlinear least square problem with a road side unit position as a solution target based on the arrival time, the arrival angle and the self position uploaded by the collaborative optimization vehicle; S303, solving a nonlinear least square problem to obtain the accurate position of the road side unit; S304, calculating whether the deviation between the accurate position of the road side unit and the currently recorded position information of the road side unit is larger than a preset threshold value, if so, updating the position information of the road side unit, and synchronizing the updated position information of the road side unit to the corresponding road side unit.
4. The vehicle-road cloud system-based information acquisition method according to claim 1, wherein the vehicle-mounted unit acquires vehicle surrounding environment data, vehicle posture data and communication channel data between the vehicle-mounted unit and the road side unit and gathers the data into a multi-source data packet uploading cloud, and specifically comprises the following operations: s401, monitoring communication channel data between a communication module and a road side unit in real time, and outputting a communication channel parameter data stream with a time stamp; S402, monitoring whether parameters in a communication channel parameter data stream meet an abnormal event triggering condition, if so, determining a time window corresponding to an abnormal event; s403, collecting communication channel parameters, vehicle positions, vehicle posture data and CAN bus data in a time window, aligning the communication channel parameters, the vehicle positions, the vehicle posture data and the CAN bus data according to uniform time stamps, and organizing the collected data into a multi-source data sequence arranged in time; s404, acquiring an image frame in a time window through a vehicle-mounted camera, and processing the image frame through an image recognition model to obtain an environment semantic tag set; s405, integrating the multi-element data sequence, the environment semantic tag set and the road side unit identity information into a data tuple, packaging the data tuple into a multi-source data packet, and uploading the multi-source data packet to the cloud.
5. The information acquisition method based on the vehicle road cloud system as set forth in claim 4, wherein the cloud end performs feature extraction on the multi-source data packet, models the state of the communication channel, and specifically includes the following operations: S501, grouping and cleaning data tuples in a multi-source data packet according to the identity information of the road side unit and the time stamp; S502, performing feature engineering on the cleaned data tuple, and outputting a feature set containing features and labels; S503, training a preset algorithm model through a feature set to obtain a channel state analysis model.
6. The vehicle road cloud system-based information acquisition method according to claim 5, wherein the feature engineering is performed on the cleaned data tuples, and a feature set including features and labels is output, and specifically comprises: s601, extracting statistical features from communication channel parameters of a data tuple; s602, performing single-heat coding on the environment semantic tag set; S603, recognizing a vehicle motion state according to the CAN bus data, and aligning the vehicle motion state, the vehicle posture data and the statistical characteristics; s604, constructing a supervised learning sample composed of features and labels, wherein the features comprise a coded environment semantic label set, vehicle attitude data, a vehicle motion state and a vehicle position, the labels are communication channel quality obtained based on statistical feature analysis, and each supervised learning sample corresponds to a time window or a one-time triggering event; s605, outputting a feature set consisting of a plurality of supervised learning samples.
7. The vehicle road cloud system-based information acquisition method according to claim 1, wherein the step S103 specifically includes the following operations: S701, when a vehicle-mounted unit receives road side perception information and accords with an event triggering condition, acquiring communication channel parameters, CAN bus data and vehicle position information in a road side perception information receiving time window, and integrating the communication channel parameters, the CAN bus data and the vehicle position information with the road side perception information to form an event data packet to be uploaded to a cloud; S702, the cloud inputs communication channel parameters into a channel state analysis model to obtain a channel quality label; S703, the cloud analyzes the road side perception information and determines traffic scene information corresponding to the road side perception information; s704, calculating an intention response index of a driver according to traffic scene information and CAN bus data; s705, outputting the intention response index as the driver intention data.
8. The vehicle road cloud system-based information acquisition method of claim 7, wherein the cloud end merges driver intention data, a communication channel state and a vehicle position, performs utility evaluation on road side perception information, and specifically comprises the following operations: S801, judging whether the channel quality label meets the minimum requirement of the channel quality, if so, executing the next step; s802, acquiring historical CAN bus data of the same position, the same period and when no road side perception information is received, and determining a historical intent response index baseline according to the historical CAN bus data; S803, calculating utility scores of the road side perception information according to the historical intent response index base line and the intent response index.
9. The vehicle road cloud system-based information acquisition method of claim 1, wherein the cloud end monitors the health degree of the road side unit automatically through data uploaded by a plurality of vehicle-mounted units, and specifically comprises the following operations: S901, the vehicle-mounted unit records interaction data of the vehicle-mounted unit and the road side unit and uploads the interaction data to the cloud; S902, summarizing the interaction data according to a time window by the cloud to form a time sequence data set of each road side unit, and simultaneously, carrying out space alignment on tracks of different vehicles in the same period to form a vehicle track density map in a broadcasting range of the road side unit; s903, calculating a multi-dimensional health index of the road side unit based on the time sequence data set; s904, detecting whether the road side units are abnormal or not based on the historical health degree of each road side unit and the health degree distribution of other road side units; S905, generating a patrol report and an operation and maintenance work order according to the detection result of S904.

Description

Information acquisition method based on vehicle road cloud system Technical Field The invention relates to the technical field of road information management systems, in particular to an information acquisition method based on a vehicle road cloud system. Background The vehicle road cloud system is an information physical system for fully integrating related information of vehicles, roads and cloud computing platforms, and the vehicles can not only sense the surrounding environment but also acquire remote information such as actual traffic conditions, route planning, driving environments and the like through network connection energization. In addition, the energy consumption can be reduced, and intelligent route planning and driving advice are beneficial to reducing unnecessary acceleration and braking, so that the fuel consumption and emission are reduced. However, the existing vehicle road cloud system has the defects that firstly, vehicle ends and road end data are synchronized in dependence on GPS time and are easy to be misaligned under the shielding of tunnels and overhead, so that time-space dislocation occurs in fusion data, secondly, the vehicle ends in the conventional implementation scheme only collect content provided by the road ends and do not pay attention to the state of a communication link, thirdly, the fault of road side equipment is required to be inspected manually or by a special vehicle at fixed points, the cost is high and the period is long, thirdly, in the conventional implementation scheme, only what information is sent by the road side and what target is 'seen' by the vehicle ends are paid attention to, and the real response of drivers to the road side information is not paid attention to, so that the utility value of the road side information cannot be evaluated. Disclosure of Invention The invention aims to provide an information acquisition method based on a vehicle road cloud system, which aims to solve or at least partially solve the defects of the existing vehicle road cloud system. In order to achieve the above purpose, the technical scheme provided by the invention is as follows: an information acquisition method based on a vehicle road cloud system, wherein the vehicle road cloud system comprises a vehicle, a road side unit and a cloud end, a vehicle-mounted unit is arranged on the vehicle, and the vehicle-mounted unit performs wireless data interaction with the road side unit and the cloud end respectively, and the method comprises the following steps: S101, broadcasting road side perception information to a vehicle-mounted unit by a road side unit in the running process of the vehicle, capturing physical layer signal characteristics of the road side perception information and CAN bus data of the vehicle by the vehicle-mounted unit, and positioning the position of the vehicle by a cloud based on the road side perception information, the physical layer signal characteristics and the CAN bus data; S102, acquiring vehicle surrounding environment data, vehicle posture data, CAN bus data and communication channel data between the vehicle-mounted unit and a road side unit by a vehicle-mounted unit, summarizing the data into a multisource data packet uploading cloud, extracting characteristics of the multisource data packet by the cloud, and modeling a communication channel state; S103, the cloud identifies the intention of the driver under the triggering of the event to obtain intention data of the driver; s104, the cloud end fuses the driver intention data, the communication channel state and the vehicle position, carries out utility evaluation on the road side perception information, generates an information utility thermodynamic diagram according to the utility evaluation result, and optimizes road side unit deployment or information release strategies. Further, the step S101 specifically includes the following operations: s201, periodically broadcasting road side perception information to the outside by a road side unit, wherein the road side perception information comprises signal lamp information and road section traffic information; S202, when a vehicle enters a broadcasting range of a road side unit, a vehicle-mounted unit continuously monitors the broadcasting of the road side unit, extracts physical layer signal characteristics for signals successfully received by each frame, wherein the physical layer signal characteristics comprise arrival time, arrival angle, channel impulse response, road side unit identity and position information; S203, the vehicle-mounted unit acquires CAN bus data of the vehicle, acquires vehicle acceleration information through the inertia measurement unit, and uploads physical layer signal characteristics, CAN bus data and vehicle acceleration information to the cloud; s204, the cloud end estimates the position of the vehicle relative to the road side unit in real time based on physical layer signal characteristics, CAN bus data and vehic