CN115188187A - Roadside perception data quality monitoring system and method based on vehicle-road cooperation - Google Patents
Roadside perception data quality monitoring system and method based on vehicle-road cooperation Download PDFInfo
- Publication number
- CN115188187A CN115188187A CN202210794082.6A CN202210794082A CN115188187A CN 115188187 A CN115188187 A CN 115188187A CN 202210794082 A CN202210794082 A CN 202210794082A CN 115188187 A CN115188187 A CN 115188187A
- Authority
- CN
- China
- Prior art keywords
- vehicle
- target
- data
- roadside
- road
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 230000008447 perception Effects 0.000 title claims abstract description 84
- 238000012544 monitoring process Methods 0.000 title claims abstract description 41
- 238000000034 method Methods 0.000 title claims abstract description 37
- 238000013441 quality evaluation Methods 0.000 claims abstract description 34
- 238000011156 evaluation Methods 0.000 claims description 28
- 238000004364 calculation method Methods 0.000 claims description 10
- 238000004458 analytical method Methods 0.000 claims description 5
- 238000013135 deep learning Methods 0.000 claims description 5
- 230000002159 abnormal effect Effects 0.000 claims description 4
- 238000001914 filtration Methods 0.000 claims description 3
- 230000006870 function Effects 0.000 claims description 3
- 238000004891 communication Methods 0.000 claims description 2
- 238000001514 detection method Methods 0.000 claims description 2
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims 1
- 230000005856 abnormality Effects 0.000 abstract description 4
- 238000010276 construction Methods 0.000 abstract description 4
- 238000011161 development Methods 0.000 abstract description 3
- 238000012423 maintenance Methods 0.000 abstract description 2
- 230000003993 interaction Effects 0.000 description 6
- 230000006855 networking Effects 0.000 description 6
- 238000012937 correction Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 239000000284 extract Substances 0.000 description 3
- 230000004927 fusion Effects 0.000 description 3
- 230000002452 interceptive effect Effects 0.000 description 3
- 238000013528 artificial neural network Methods 0.000 description 2
- 230000002093 peripheral effect Effects 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 241000497429 Obus Species 0.000 description 1
- 238000007792 addition Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000013210 evaluation model Methods 0.000 description 1
- 238000009440 infrastructure construction Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 239000013589 supplement Substances 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/01—Detecting movement of traffic to be counted or controlled
-
- G—PHYSICS
- G08—SIGNALLING
- G08B—SIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
- G08B21/00—Alarms responsive to a single specified undesired or abnormal condition and not otherwise provided for
- G08B21/18—Status alarms
- G08B21/24—Reminder alarms, e.g. anti-loss alarms
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/01—Detecting movement of traffic to be counted or controlled
- G08G1/0104—Measuring and analyzing of parameters relative to traffic conditions
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/01—Detecting movement of traffic to be counted or controlled
- G08G1/0104—Measuring and analyzing of parameters relative to traffic conditions
- G08G1/0108—Measuring and analyzing of parameters relative to traffic conditions based on the source of data
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/01—Detecting movement of traffic to be counted or controlled
- G08G1/0104—Measuring and analyzing of parameters relative to traffic conditions
- G08G1/0125—Traffic data processing
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Business, Economics & Management (AREA)
- Emergency Management (AREA)
- Traffic Control Systems (AREA)
Abstract
The scheme discloses a roadside sensing data quality monitoring system and method based on vehicle-road cooperation, which comprises S1, receiving roadside sensing data acquired by roadside sensing equipment; acquiring a vehicle message from an intelligent vehicle-mounted unit installed in an online vehicle; s2, extracting first track data of the target vehicle from the vehicle message; extracting second track data of the target vehicle from the roadside perception data; s3, analyzing and comparing first track data and second track data corresponding to the same target vehicle to evaluate a roadside perception data quality evaluation value and reporting the roadside perception data quality evaluation value to a cloud control platform; and S4, when the quality evaluation value of the road side sensing data exceeds a quality threshold value, sensing abnormality is given to the external alarm point. The data quality monitoring of the roadside sensing system in the construction of the intelligent network-connected road infrastructure is focused, the reliability and accuracy of the roadside sensing data are effectively guaranteed, the operation and maintenance guarantee capacity of the roadside sensing system is improved, and the development prospect and the practical value of popularization and application are achieved.
Description
Technical Field
The invention belongs to the technical field of roadside perception data quality monitoring, and particularly relates to a roadside perception data quality monitoring system and method based on vehicle-road cooperation.
Background
With the development of the intelligent internet automobile technology and the promotion of the construction of national intelligent internet demonstration areas, the construction of urban intelligent internet road infrastructures is vigorous. The roadside sensing system is an important component of the construction of intelligent networking road infrastructure, the data sensing precision of the roadside sensing system is directly related to the realization of networking vehicle application scenes, and the roadside sensing system is of great importance to the guarantee of road traffic safety. At present, the roadside perception system is gradually deployed in large scale, and how to scientifically monitor and operate and maintain the system is achieved, so that the quality of roadside perception data is effectively guaranteed to become a difficult point which needs to be solved urgently.
The existing urban road intelligent equipment monitoring system mainly focuses on basic operation and maintenance information such as equipment online, equipment offline and equipment failure. However, the roadside sensing system of the intelligent internet road has a high requirement for monitoring data quality, and the existing monitoring system and method are difficult to meet.
In order to solve the quality problem of roadside sensing data, a series of researches are carried out. However, the current research only needs to evaluate the relevant quality of data, such as reasonability, volatility, interaction abnormality and the like, for example, chinese patent discloses a quality evaluation method of road traffic perception trajectory data [ application No.: CN202110789522.4], mainly comprising the following steps: acquiring road traffic perception track data, and extracting road traffic perception track data to be evaluated from the road traffic perception track data; calculating a quality evaluation index of the road traffic perception track data to be evaluated; calculating two types of evaluation indexes of track reasonability and track volatility of road traffic perception track data without an interactive relation; and calculating three evaluation indexes of track reasonability, track volatility and track interaction abnormity aiming at road traffic perception track data with an interaction relation. Aiming at a road traffic perception track data set Trnone without an interactive relation, a self-adaptive fusion quality evaluation model MA is established, wherein the input of the model MA is a track rationality index and a track volatility index, and a quality evaluation score can be output after nonlinear fitting is carried out through a neural network; aiming at a road traffic perception track data set Tract with an interactive relation, a self-adaptive fusion quality evaluation model MB is established, wherein the quality evaluation model MB consists of a model MA and a correction network: inputting the track rationality index and the track volatility index into the model MA, outputting the track rationality index and the track volatility index into a quality evaluation score, inputting the quality evaluation score and the track interaction abnormity index into a correction network together, performing fitting correction through a neural network, and outputting a final quality evaluation score.
After the sensing data are obtained, the quality of the sensing data is evaluated in real time by using the corresponding evaluation model, and the method is mainly used for evaluating the rationality, volatility, interaction abnormality and the like of the data. However, sometimes, when the device moves slightly due to severe weather such as wind and rain, the device calibration may be caused to be problematic, which may affect the sensing accuracy of the data to a certain extent, and may not cause problems such as data reasonability, volatility, and interaction abnormality, that is, quality problems in a general sense may not occur.
Disclosure of Invention
The invention aims to provide a roadside sensing data quality monitoring system and method based on vehicle-road cooperation, which can ensure the quality of monitoring data from another angle and supplement and improve the data quality in the prior art so as to ensure the functional safety of the roadside sensing system in scale application.
In order to achieve the purpose, the invention adopts the following technical scheme:
a roadside perception data quality monitoring method based on vehicle-road cooperation comprises the following steps:
s1, receiving roadside sensing data acquired by roadside sensing equipment;
acquiring a vehicle message from an intelligent vehicle-mounted unit installed in an internet vehicle;
s2, extracting first track data of the target vehicle from the vehicle message;
extracting second track data of the target vehicle from the roadside perception data;
s3, analyzing and comparing first track data and second track data corresponding to the same target vehicle to evaluate a roadside perception data quality evaluation value and reporting the roadside perception data quality evaluation value to a cloud control platform;
and S4, when the quality evaluation value of the road side perception data exceeds a quality threshold value, the external alarm point is abnormal in perception.
In the roadside sensing data quality monitoring method based on vehicle-road cooperation, the roadside sensing equipment in the step S1 comprises a roadside radar and a video camera, wherein the roadside sensing data comprises real-time point cloud data acquired by the roadside radar and video image data acquired by the video camera;
and acquiring the vehicle message from an intelligent vehicle-mounted unit installed in the networked vehicle through the intelligent road side unit.
In the roadside sensing data quality monitoring method based on vehicle-road cooperation, in step S2, a target vehicle is determined by the following method:
extracting a set central point, extracting a set radius, and taking a region formed by taking the central point as an original point and the set radius as the radius as a sensing range;
taking the networked vehicle which is positioned in the sensing range and provided with the intelligent vehicle-mounted unit as a target vehicle; and judging whether the corresponding vehicle is positioned in the sensing range or not according to the vehicle message, and when the corresponding vehicle enters the sensing range and when the corresponding vehicle leaves the sensing range. The networked vehicle capable of receiving the vehicle message is determined whether the networked vehicle is located in the sensing range, and the networked vehicle passing through the corresponding road section is determined to be the target vehicle in the time period from the time of entering the sensing range to the time of leaving the sensing range according to the vehicle message, and the first track data set of the networked vehicle is the set of track data in all vehicle messages received by the edge calculation unit in the time period.
In the roadside sensing data quality monitoring method based on vehicle-road cooperation, the set central point and the set radius are preset; and when the intersection is located at the crossroad, the central point of the intersection is set as the origin.
In the roadside perception data quality monitoring method based on vehicle-road cooperation, the first track data and the second track data respectively comprise a target ID, a target type, a target position (longitude and latitude), a target speed and a target course angle of a target vehicle;
for the first trajectory data, the vehicle message obtained from the intelligent on-board unit includes data: the method comprises the following steps that a target ID, a target type, a target position (longitude and latitude), a target speed and a target course angle of a corresponding vehicle are directly extracted from vehicle information by an edge computing unit;
for the second track data, the edge calculation unit carries out timestamp unification and sensor space coordinate calibration on all road side perception data, and then uses a corresponding deep learning algorithm to identify the vehicle and track the target to obtain a target ID, a target type, a target position (longitude and latitude), a target speed and a target course angle of the target vehicle so as to obtain the second track data of the target vehicle. Based on the roadside perception data, the vehicle is identified by using a deep learning algorithm, and the data such as the ID, the type, the position, the speed, the course angle and the like of the target vehicle are obtained by tracking the target by directly using the prior art, so that the part is not improved, and the description is omitted.
In the roadside sensing data quality monitoring method based on vehicle-road cooperation, the step S2 specifically includes:
filtering the vehicle messages from the intelligent vehicle-mounted unit to obtain vehicle messages in the perception range; it may be determined whether the vehicle is within the perception range based on the location data in the vehicle message.
Obtaining a data set { (timestamp, vehicle number, longitude, latitude, speed, course angle) } from the vehicle message; the time node when the corresponding networked vehicle enters the sensing range is t 0 Time node leaving sensing range is t k The first track data set of the corresponding networked vehicle is
The ID is the serial number of the corresponding internet vehicle,
respectively as the longitude of the corresponding time of the networked vehicle,
respectively the corresponding time latitudes of the networked vehicles,
respectively the speed of the networked vehicles at the corresponding moment,
respectively is the corresponding moment course angle of the networked vehicle;
the networked vehicles passing through the sensing range are regarded as target vehicles in a time period within the sensing range; different target vehicles have different t 0 、t k ；
For each target vehicle
At the moment of time to
Obtaining a second track data set by roadside perception data at the moment
Wherein
Respectively detecting time nodes from the side of the road side sensing system to the entrance and exit of a target vehicle i in the sensing range, and identifying the time nodes i Is the perception number of the target vehicle i,
for the road side perception system to perceive longitude of the target vehicle i at the corresponding time,
for the roadside sensing system to sense the latitude of the target vehicle i at the corresponding time,
for the roadside sensing system to sense the speed of the target vehicle i at the corresponding time,
and (4) sensing the heading angle of the target vehicle i at the corresponding moment for the roadside sensing system.
In the roadside sensing data quality monitoring method based on vehicle-road cooperation, in step S3, quality evaluation is performed on roadside sensing data at set time intervals, and each evaluation period is based on all target vehicles in the corresponding evaluation period for quality evaluation.
In the roadside perception data quality monitoring method based on vehicle-road cooperation, in step S3, a target vehicle with the minimum positioning deviation in an evaluation period is matched with an internet vehicle track, and evaluation value calculation is performed based on the matched internet vehicle track:
s31, traversing a first track data set and a second track data set obtained in an evaluation period by taking delta as a target function, and taking a vehicle ID corresponding to the minimum value as a matched internet vehicle track;
s32, calculating the average positioning error based on the track of the networked vehicles
Corresponding to the speed error of the networked vehicle with the ID of i
Course angle error
θ i ，σ i ，ω i The calculated value is an error value corresponding to the networked vehicle i,
the average value is the roadside perception data quality evaluation value.
A roadside sensing data quality monitoring system based on vehicle-road cooperation comprises roadside sensing equipment used for obtaining roadside sensing data, the roadside sensing equipment is connected to an edge computing unit, the edge computing unit is connected with an intelligent roadside unit, wireless communication with an intelligent vehicle-mounted unit installed in an internet vehicle is achieved through the intelligent roadside unit to obtain vehicle information from the intelligent vehicle-mounted unit, the edge computing unit is used for determining a target vehicle based on the vehicle information, identifying the target vehicle and extracting first track data of the target vehicle based on the roadside sensing data, extracting second track data of the target vehicle based on the vehicle information, evaluating the quality of the roadside sensing data through analysis and comparison of the first track data and the second track data corresponding to the same target vehicle, reporting an evaluation value to a cloud control platform, and when the evaluation value exceeds a quality threshold value, sending out external sensing abnormal point position alarm information by the cloud control platform.
In the roadside sensing data quality monitoring system based on vehicle-road cooperation,
the roadside sensing equipment comprises a roadside radar and a video camera;
the first track data and the second track data respectively comprise a target ID, a target type, a target position (longitude and latitude), a target speed and a target course angle of a target vehicle;
the edge computing unit evaluates the quality of the roadside sensing data based on any roadside sensing data quality monitoring method based on vehicle-road cooperation.
The invention has the advantages that:
the not enough of trackside perception system data quality monitoring in this scheme focus intelligence networking road infrastructure construction, based on vehicle road collaborative mode, use networking vehicle positioning data as the truth, through comparing trackside perception data and networking vehicle high accuracy GNSS positioning data, can support real-time supervision trackside perception data precision, discover equipment perception unusual and in time report to the police, the reliability and the accuracy of trackside perception data have effectively been ensured, trackside perception system's fortune dimension guarantee ability has been promoted, development prospect and the realistic value that has popularization and application, can be applicable to intelligence networking demonstration district, the high-speed application scene of deploying a large amount of perception equipment that become more meticulous of wisdom, improve equipment operation supervision efficiency, reduce artifical fortune dimension cost.
Drawings
FIG. 1 is a block diagram of a system structure of a roadside sensing data quality monitoring system based on vehicle-road coordination according to the present invention;
FIG. 2 is a schematic diagram of equipment distribution of the roadside perception data quality monitoring system based on vehicle-road cooperation for an intersection scene according to the present invention;
FIG. 3 is a flow chart of a monitoring method in the roadside perception data quality monitoring system based on vehicle-road cooperation according to the present invention;
FIG. 4 is a flowchart of a method for obtaining two track sets in the roadside sensing data quality monitoring system based on vehicle-road cooperation according to the present invention;
FIG. 5 is a schematic diagram of a road side perception data quality monitoring and evaluation cycle based on vehicle-road cooperation.
Detailed Description
The present invention will be described in further detail with reference to the accompanying drawings and specific embodiments.
Taking an intersection as an example, as shown in fig. 1 and fig. 2, the roadside sensing data quality monitoring system based on vehicle-road cooperation provided in this embodiment includes roadside sensing devices (such as a video camera, a laser radar, a millimeter wave radar, and the like), an edge computing unit, an intelligent roadside unit (RSU), an intelligent on-board unit (OBU), and a cloud control platform. The method comprises the steps that various roadside sensing devices obtain intersection road environment data in real time based on respective sensor sensing characteristics, such as real-time point cloud data obtained by roadside radars, video image data obtained by video cameras, and an intelligent roadside unit (RSU) obtains high-precision positioning data of networked vehicles from peripheral intelligent vehicle-mounted units (OBUs) based on a vehicle-road cooperation mode. The method comprises the steps of carrying out space-time unification, target identification and fusion on different sensors through an edge computing unit to obtain motor vehicle real-time track data obtained based on a roadside sensing system, and finally judging roadside sensing positioning errors by comparing networked vehicle positioning data with roadside sensing data vehicle tracks so as to evaluate the quality of the roadside sensing data.
Specifically, as shown in fig. 3, the road-side sensing data quality monitoring method based on vehicle-road cooperation is as follows:
s1, the roadside sensing equipment acquires roadside sensing data and transmits the roadside sensing data to an edge computing unit. The roadside sensing equipment can be a laser radar, a video camera, a millimeter wave radar and the like, and the roadside sensing equipment is deployed inwards in a covering mode around each direction of the intersection because the intersection is taken as an example in the embodiment, and can be deployed by means of intersection electric warning rod pieces, traffic light rod pieces and the like, and a typical deployment mode is shown in fig. 2.
Meanwhile, an intelligent Road Side Unit (RSU) receives a vehicle message BSM message broadcasted by an intelligent vehicle-mounted unit (OBU) of a peripheral internet vehicle, and sends the obtained BSM message to an edge computing unit. In the embodiment, the road side sensing data and the BSM message frequency of the networked vehicle are both 10Hz.
And S2, the edge calculation unit extracts first track data of the target vehicle from the BSM message and extracts second track data of the target vehicle from the roadside sensing data.
The first track data and the second track data comprise a target ID, a target type, a target position (longitude and latitude), a target speed and a target course angle of the target vehicle;
for the first trajectory data, the vehicle message obtained from the intelligent on-board unit OBU comprises data: the method comprises the following steps that an edge computing unit directly extracts first track data from vehicle information according to a target ID, a target type, a target position (longitude and latitude), a target speed and a target course angle of a corresponding vehicle;
for the second track data, the edge calculation unit performs timestamp unification and sensor space coordinate calibration on the roadside sensing data of all the roadside sensing devices, then uses a corresponding deep learning algorithm to identify and track the vehicle, and outputs a target ID, a target type, a target position (longitude and latitude), a target speed and a target course angle of the target vehicle according to the frequency of 10Hz to obtain the second track data of the target vehicle. It should be noted that, based on the roadside sensing data, the vehicle is identified by using a deep learning algorithm, and the data such as the ID, the type, the position, the speed, the heading angle, and the like of the target vehicle are obtained by tracking the target, which is directly used in the prior art, and this part is not improved here, and therefore, the description is omitted here.
S3, analyzing and comparing first track data and second track data corresponding to the same target vehicle to evaluate a roadside perception data quality evaluation value and reporting the roadside perception data quality evaluation value to a cloud control platform;
furthermore, the quality of roadside perception data is evaluated mainly by comparing the first track data and the second track data of the target vehicle, so that the first track data and the second track data of the target vehicle need to be acquired, and whether the first track data and the second track data are acquired for the non-target vehicle (the non-target vehicle comprises a vehicle which does not pass through a perception range and the vehicle which passes through the perception range is located outside the perception range) is out of the limit range of the application. The manner of determining the target vehicle here is:
and extracting a set central point, extracting a set radius, taking the central point as an original point, taking an area formed by the set radius as the radius as a sensing range, and then taking the internet vehicles which are positioned in the sensing range and are provided with the intelligent vehicle-mounted units as target vehicles.
The central point and the radius are set in advance by technicians according to conditions such as road conditions.
Specifically, whether the corresponding networked vehicle is located in the sensing range, and when the networked vehicle enters the sensing range and when the networked vehicle leaves the sensing range are judged according to the vehicle information. The networked vehicles capable of receiving the vehicle message are determined to be the target vehicles in the time period (from the time of entering the sensing range to the time of leaving the sensing range) of the corresponding road section, and the networked vehicles passing the corresponding road section are determined to be non-target vehicles at other times according to the vehicle message, and the first track data set of the networked vehicles is the set of track data in all vehicle messages received by the edge calculation unit in the time period.
When the intersection is located at the intersection, the center point of the intersection is preferably set as the origin, and the radius within 25 meters is used as the sensing range.
And S4, when the quality evaluation value of the road side sensing data exceeds the quality threshold value, sensing abnormity at the external alarm point position, and usually alarming to a service platform. The detection of low-quality sensing data and the alarm of abnormal equipment calibration are realized. Whether the roadside perception data quality evaluation value exceeds the quality threshold value or not can be judged by the cloud control platform or the edge computing unit, the quality threshold value is determined by technicians in the field according to actual conditions, for example, the quality threshold value is determined according to conditions such as vehicle speed limit and the like of a deployment scene, generally, the higher the speed limit speed is, the higher the threshold value is, and the specific situation is not limited.
Further, as shown in fig. 4, step S2 specifically includes:
filtering vehicle messages from the intelligent vehicle-mounted unit to obtain BSM messages in a perception range; it may be determined whether the vehicle is within the perception range based on the location data in the vehicle message.
Based on the foregoing, the networked vehicle passing through the sensing range is regarded as the target vehicle during the time period in the sensing range;
get the data set { (timestamp, vehicle number, longitude, latitude, speed, heading angle) } from the BSM message,
the time node when the corresponding networked vehicle enters the sensing range is t 0 Time node leaving sensing range is t k Those skilled in the art will appreciate that different target vehicles will typically have different t 0 、t k . The first track data set of the corresponding networked vehicle is
The ID is the corresponding internet vehicle number,
respectively as the longitude of the corresponding time of the networked vehicle,
respectively the corresponding time latitudes of the networked vehicles,
respectively the speed of the networked vehicle at the corresponding moment,
respectively a course angle of the networked vehicle at the corresponding moment;
obtaining t of each target vehicle 0 Time to t k Obtaining a second track data set by the roadside perception data at the moment
Wherein
Time nodes of the target vehicle i entering and leaving the perception range, namely time node t corresponding to the target vehicle i 0 ，t k 。ID i Is the perception number of the target vehicle i,
for roadside sensing systems in responseThe perceived longitude of the target vehicle i at that time,
for the roadside sensing system to sense the latitude of the target vehicle i at the corresponding time,
for the roadside sensing system to sense the speed of the target vehicle i at the corresponding time,
and (4) sensing the heading angle of the target vehicle i at the corresponding moment for the roadside sensing system.
Further, in step S3, the target vehicle with the smallest positioning deviation in the evaluation period is matched with the internet vehicle trajectory, and an evaluation value calculation is performed based on the matched internet vehicle trajectory:
s31, traversing all the first track data sets and all the second track data sets obtained in the evaluation period by taking the delta as a target function, and taking the vehicle ID corresponding to the minimum value as the matched internet vehicle track;
s32, calculating average positioning error based on the vehicle track of the internet
Corresponding to the speed error of the networked vehicle with the ID of i
Course angle error
θ i ，σ i ，ω i The calculated value is an error value corresponding to the networked vehicle i,
mean value ofAnd the road side perception data quality evaluation value is obtained.
Preferably, an evaluation period is set for the system, that is, quality evaluation is performed on the roadside sensing data at set time intervals, which may be once a day, once every 2 hours, once every 12 hours, and so on. And each evaluation period is used for carrying out quality evaluation on all target vehicles in the corresponding evaluation period, specifically, a target vehicle with the minimum positioning error is selected from all target vehicles in the evaluation period, and then the quality evaluation is carried out on the basis of the target vehicle. As shown in fig. 5, four evaluation periods T0-T1, T1-T2, T2-T3, and T3-T4 are divided, and taking the evaluation period T0-T1 as an example, three vehicles a, B, and C pass through the sensing range in the current evaluation period.
The time for the vehicle A to enter and leave is ta0 and tak respectively;
the time when the vehicle B enters and leaves is tb0 and tbk respectively;
the time for the vehicle C to enter and leave is tc0 and tck respectively;
the first track data set of the A vehicle in a time period of ta0-tak is A [ ], and the second track data set is A' [ ];
the first track data set of the vehicle B is B [ ] and the second track data set is B' [ ] in the time periods tb 0-tbk;
the first track data set of the vehicle C in the time period tc0-tck is C [ ], and the second track data set is C' [ ];
then the evaluation mode is as follows:
2A]And A' 2]By passing
The analysis and comparison are carried out, and the obtained result is theta a ,σ a ,ω a ；
2 [ B ]],B'[]By passing
The analysis and comparison are carried out, and the obtained result is theta b ,σ b ,ω b ；
2C],C'[]By passing
The analysis and comparison are carried out, and the obtained result is theta c ,σ c ,ω c ；
Is calculated to obtain
The roadside sensing data quality evaluation value of the quality evaluation is averaged to obtain the difference theta, sigma and omega, and when any one of theta, sigma and omega is higher than the quality threshold value, an alarm is given.
According to the system and the method, the motor vehicle target track data in the sensing range is obtained through the edge computing unit, and is compared and analyzed with the BSM (vehicle-to-vehicle) information of the networked vehicle obtained based on the intelligent Road Side Unit (RSU), and the road side sensing data quality is evaluated.
The specific embodiments described herein are merely illustrative of the spirit of the invention. Various modifications or additions may be made to the described embodiments, or alternatives may be employed, by those skilled in the art, without departing from the spirit or ambit of the invention as defined in the appended claims.
Although terms such as intelligent on-board unit, intelligent road side unit, road side sensing device, edge computing unit, cloud control platform, service platform, etc. are used more herein, the possibility of using other terms is not excluded. These terms are used merely to more conveniently describe and explain the nature of the present invention; they are to be construed as being without limitation to the spirit of the present invention.
Claims (10)
1. A roadside perception data quality monitoring method based on vehicle-road cooperation is characterized by comprising the following steps:
s1, receiving roadside sensing data acquired by roadside sensing equipment;
acquiring a vehicle message from an intelligent vehicle-mounted unit installed in an internet vehicle;
s2, extracting first track data of the target vehicle from the vehicle message;
extracting second track data of the target vehicle from the roadside sensing data;
s3, analyzing and comparing first track data and second track data corresponding to the same target vehicle to evaluate a road side perception data quality evaluation value and reporting the road side perception data quality evaluation value to a cloud control platform;
and S4, when the quality evaluation value of the road side perception data exceeds a quality threshold value, the external alarm point is abnormal in perception.
2. The road-side perception data quality monitoring method based on vehicle-road cooperation according to claim 1, wherein in step S1, the road-side perception device includes a road-side radar and a video camera, and the road-side perception data includes real-time point cloud data acquired by the road-side radar and video image data acquired by the video camera;
and acquiring the vehicle message from an intelligent vehicle-mounted unit installed in the networked vehicle through the intelligent road side unit.
3. The road-side perception data quality monitoring method based on vehicle-road cooperation according to claim 1 or 2, wherein in step S2, the target vehicle is determined by:
extracting a set central point, extracting a set radius, and taking an area formed by taking the set central point as an original point and the set radius as the radius as a sensing range;
taking the networked vehicle which is positioned in the sensing range and provided with the intelligent vehicle-mounted unit as a target vehicle;
and judging whether the corresponding vehicle is positioned in the perception range or not according to the vehicle message, and when the corresponding vehicle enters the perception range and when the corresponding vehicle leaves the perception range.
4. The roadside perception data quality monitoring method based on vehicle-road cooperation according to claim 3, wherein the set central point and the set radius are preset; and when the intersection is located at the intersection, the center point of the intersection is set as the origin.
5. The roadside perception data quality monitoring method based on vehicle-road coordination according to claim 3, characterized in that,
the first track data and the second track data comprise a target ID, a target type, a target position (longitude and latitude), a target speed and a target course angle of a target vehicle;
for the first trajectory data, the vehicle message obtained from the intelligent on-board unit includes data: the method comprises the following steps that a target ID, a target type, a target position (longitude and latitude), a target speed and a target course angle of a corresponding vehicle are directly extracted from vehicle information by an edge computing unit;
for the second track data, the edge calculation unit carries out timestamp unification and sensor space coordinate calibration on all road side perception data, and then uses a corresponding deep learning algorithm to identify the vehicle and track the target to obtain a target ID, a target type, a target position (longitude and latitude), a target speed and a target course angle of the target vehicle so as to obtain the second track data of the target vehicle.
6. The roadside perception data quality monitoring method based on vehicle-road coordination according to claim 5, characterized in that,
the step S2 specifically includes:
filtering the vehicle message from the intelligent vehicle-mounted unit to obtain the vehicle message in the perception range;
obtaining a data set { (timestamp, vehicle number, longitude, latitude, speed, heading angle) } from the vehicle message,
the time node when the corresponding networked vehicle enters the sensing range is t 0 Time node of departure sensing range is t k The first track data set of the corresponding networked vehicle is
The ID is the corresponding internet vehicle number,
respectively as the longitude of the corresponding time of the networked vehicle,
respectively the corresponding time latitudes of the networked vehicles,
respectively the speed of the networked vehicle at the corresponding moment,
respectively is the corresponding moment course angle of the networked vehicle;
obtaining t of each target vehicle 0 Time to t k Obtaining a second track data set by roadside perception data at the moment
Wherein
Respectively time nodes from the detection side of the road side sensing system to the entrance and exit of the target vehicle i in the sensing range, ID i Is the perception number of the target vehicle i,
the perception warp of the roadside perception system to the target vehicle i at the corresponding momentThe degree of the water is measured by the following method,
for the roadside sensing system to sense the latitude of the target vehicle i at the corresponding time,
for the roadside sensing system to sense the speed of the target vehicle i at the corresponding time,
and (4) sensing the heading angle of the target vehicle i at the corresponding moment for the roadside sensing system.
7. The road-side perception data quality monitoring method based on vehicle-road cooperation according to claim 6, wherein in step S3, quality evaluation is performed on the road-side perception data at set time intervals, and each evaluation period is based on quality evaluation of all target vehicles in the corresponding evaluation period.
8. The roadside perception data quality monitoring method based on vehicle-road cooperation according to claim 7, wherein in step S3, a target vehicle with minimum positioning deviation in an evaluation period is matched with a vehicle track of the internet, and evaluation value calculation is performed based on the matched vehicle track of the internet:
s31, traversing all the first track data sets and all the second track data sets obtained in the evaluation period by taking the delta as a target function, and taking the vehicle ID corresponding to the minimum value as the matched internet vehicle track;
s32: calculating average positioning error based on internet vehicle track
Corresponding to a network with ID iSpeed error of connected vehicle
Course angle error
θ i ，σ i ，ω i The calculated value is an error value corresponding to the internet vehicle i,
the average value is the roadside perception data quality evaluation value.
9. The system is characterized by comprising roadside sensing equipment, wherein the roadside sensing equipment is used for acquiring roadside sensing data, the roadside sensing equipment is connected with an edge computing unit, the edge computing unit is connected with an intelligent roadside unit, wireless communication with an intelligent vehicle-mounted unit installed in an internet vehicle is achieved through the intelligent roadside unit to acquire vehicle information from the intelligent vehicle-mounted unit, the edge computing unit is used for determining a target vehicle based on the vehicle information, identifying the target vehicle and extracting first track data of the target vehicle based on the roadside sensing data, extracting second track data of the target vehicle based on the vehicle information, evaluating the quality of the roadside sensing data through an analysis ratio of the first track data and the second track data corresponding to the same target vehicle, reporting an evaluation value to a cloud control platform, and when the evaluation value exceeds a quality threshold value, sending sensing point location alarm information to the outside by the cloud control platform.
10. The road side perceptual data quality monitoring system of the vehicle-road synergy based on claim 9,
the roadside sensing equipment comprises a roadside radar and a video camera;
the first track data and the second track data comprise a target ID, a target type, a target position (longitude and latitude), a target speed and a target course angle of a target vehicle;
the edge calculation unit evaluates the quality of roadside perception data based on the roadside perception data quality monitoring method based on vehicle-road cooperation of any one of claims 1 to 8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210794082.6A CN115188187A (en) | 2022-07-05 | 2022-07-05 | Roadside perception data quality monitoring system and method based on vehicle-road cooperation |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210794082.6A CN115188187A (en) | 2022-07-05 | 2022-07-05 | Roadside perception data quality monitoring system and method based on vehicle-road cooperation |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115188187A true CN115188187A (en) | 2022-10-14 |
Family
ID=83517711
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210794082.6A Pending CN115188187A (en) | 2022-07-05 | 2022-07-05 | Roadside perception data quality monitoring system and method based on vehicle-road cooperation |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115188187A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116614841A (en) * | 2023-07-17 | 2023-08-18 | 中汽智联技术有限公司 | Road side data quality assessment method and electronic equipment |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113779174A (en) * | 2021-11-05 | 2021-12-10 | 华砺智行(武汉)科技有限公司 | Method, system, equipment and medium for improving perception precision of roadside sensor |
| CN114173307A (en) * | 2021-12-17 | 2022-03-11 | 浙江海康智联科技有限公司 | Roadside perception fusion system based on vehicle-road cooperation and optimization method |
| CN114357019A (en) * | 2021-12-03 | 2022-04-15 | 同济大学 | Method for monitoring data quality of road side sensing unit in intelligent networking environment |
| CN114648231A (en) * | 2022-03-29 | 2022-06-21 | 北京清研宏达信息科技有限公司 | Quality evaluation method for road-side data of vehicle-road cooperation |
-
2022
- 2022-07-05 CN CN202210794082.6A patent/CN115188187A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113779174A (en) * | 2021-11-05 | 2021-12-10 | 华砺智行(武汉)科技有限公司 | Method, system, equipment and medium for improving perception precision of roadside sensor |
| CN114357019A (en) * | 2021-12-03 | 2022-04-15 | 同济大学 | Method for monitoring data quality of road side sensing unit in intelligent networking environment |
| CN114173307A (en) * | 2021-12-17 | 2022-03-11 | 浙江海康智联科技有限公司 | Roadside perception fusion system based on vehicle-road cooperation and optimization method |
| CN114648231A (en) * | 2022-03-29 | 2022-06-21 | 北京清研宏达信息科技有限公司 | Quality evaluation method for road-side data of vehicle-road cooperation |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116614841A (en) * | 2023-07-17 | 2023-08-18 | 中汽智联技术有限公司 | Road side data quality assessment method and electronic equipment |
| CN116614841B (en) * | 2023-07-17 | 2023-10-27 | 中汽智联技术有限公司 | Road side data quality assessment method and electronic equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11836985B2 (en) | Identifying suspicious entities using autonomous vehicles | |
| US11380105B2 (en) | Identification and classification of traffic conflicts | |
| CN112700470B (en) | Target detection and track extraction method based on traffic video stream | |
| CN110164130B (en) | Traffic incident detection method, device, equipment and storage medium | |
| CN113706737B (en) | Road surface inspection system and method based on automatic driving vehicle | |
| CN114333330B (en) | Intersection event detection system based on road side edge holographic sensing | |
| CN112433519A (en) | Unmanned driving detection system and vehicle driving detection method | |
| US20220254249A1 (en) | Traffic event and road condition identification and classification | |
| CN115618932A (en) | Traffic incident prediction method and device based on internet automatic driving and electronic equipment | |
| US20220237919A1 (en) | Method, Apparatus, and Computing Device for Lane Recognition | |
| Wang et al. | Realtime wide-area vehicle trajectory tracking using millimeter-wave radar sensors and the open TJRD TS dataset | |
| CN111477011A (en) | Detection device and detection method for road intersection early warning | |
| CN114771548A (en) | Data logging for advanced driver assistance system testing and verification | |
| CN115188187A (en) | Roadside perception data quality monitoring system and method based on vehicle-road cooperation | |
| CN114973659A (en) | Method, device and system for detecting indirect event of expressway | |
| Sabour et al. | Deepflow: Abnormal traffic flow detection using siamese networks | |
| Zheng | Developing a traffic safety diagnostics system for unmanned aerial vehicles usingdeep learning algorithms | |
| CN115527364B (en) | Traffic accident tracing method and system based on radar data fusion | |
| KR20230166855A (en) | Antenna monitoring and selection | |
| CN104408942A (en) | Intelligent vehicle speed measuring device and method | |
| CN117178309A (en) | Method for creating a map with collision probability | |
| CN109685010B (en) | Highway slope protection network vulnerability position positioning method and system | |
| CN114968189A (en) | Platform for perception system development of an autopilot system | |
| US20220165151A1 (en) | Traffic jam information providing device, traffic jam information processing method, and recording medium | |
| CN116129631A (en) | Vehicle-road cooperative data processing method and related equipment |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |