WO2022001618A1 - 一种车辆的车道保持控制方法、装置及系统 - Google Patents

一种车辆的车道保持控制方法、装置及系统 Download PDF

Info

Publication number
WO2022001618A1
WO2022001618A1 PCT/CN2021/099499 CN2021099499W WO2022001618A1 WO 2022001618 A1 WO2022001618 A1 WO 2022001618A1 CN 2021099499 W CN2021099499 W CN 2021099499W WO 2022001618 A1 WO2022001618 A1 WO 2022001618A1
Authority
WO
WIPO (PCT)
Prior art keywords
vehicle
geographic information
virtual geographic
road
information
Prior art date
Application number
PCT/CN2021/099499
Other languages
English (en)
French (fr)
Inventor
徐哲
Original Assignee
华为技术有限公司
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by 华为技术有限公司 filed Critical 华为技术有限公司
Publication of WO2022001618A1 publication Critical patent/WO2022001618A1/zh

Links

Images

Classifications

    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B60—VEHICLES IN GENERAL
    • B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
    • B60W30/00—Purposes of road vehicle drive control systems not related to the control of a particular sub-unit, e.g. of systems using conjoint control of vehicle sub-units
    • B60W30/10—Path keeping
    • B60W30/12—Lane keeping
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B60—VEHICLES IN GENERAL
    • B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
    • B60W40/00—Estimation or calculation of non-directly measurable driving parameters for road vehicle drive control systems not related to the control of a particular sub unit, e.g. by using mathematical models
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B60—VEHICLES IN GENERAL
    • B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
    • B60W40/00—Estimation or calculation of non-directly measurable driving parameters for road vehicle drive control systems not related to the control of a particular sub unit, e.g. by using mathematical models
    • B60W40/02—Estimation or calculation of non-directly measurable driving parameters for road vehicle drive control systems not related to the control of a particular sub unit, e.g. by using mathematical models related to ambient conditions
    • B60W40/06—Road 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
    • G—PHYSICS
    • G08—SIGNALLING
    • G08G—TRAFFIC CONTROL SYSTEMS
    • G08G1/00—Traffic control systems for road vehicles
    • G08G1/09—Arrangements for giving variable traffic instructions
    • G08G1/0962—Arrangements for giving variable traffic instructions having an indicator mounted inside the vehicle, e.g. giving voice messages
    • G08G1/0967—Systems involving transmission of highway information, e.g. weather, speed limits
    • G—PHYSICS
    • G08—SIGNALLING
    • G08G—TRAFFIC CONTROL SYSTEMS
    • G08G1/00—Traffic control systems for road vehicles
    • G08G1/16—Anti-collision systems

Definitions

  • the present application relates to the field of automatic driving, and in particular, to a vehicle lane keeping control method, device and system.
  • Lane Keep Assist is one of the important functions to realize autonomous driving.
  • the lane line of the current driving lane is identified, and the lane line information is transmitted to the vehicle computing system to determine whether the vehicle is running within the lane range in real time, thereby ensuring driving safety.
  • LKA is a key technology to realize ADAS (Advanced Driving Assistance System) and Level 2-Level 5 autonomous driving, and it is also one of the important technologies to ensure the driving safety of autonomous vehicles.
  • the data input of this technology mainly relies on on-board sensors to detect the lane lines on both sides of the road, and the on-board sensors are easily affected by weather and have a very limited sensing range, so they cannot effectively obtain lane information in the far area ahead.
  • some areas of urban roads and highways have lane lines worn or even no lane lines, which leads to the fact that the lane range cannot be accurately perceived by the on-board equipment, which ultimately limits the use range of the lane keeping function. Effect.
  • the present application provides a vehicle lane keeping control method, device and system, which reduces the influence of the weather environment on the lane keeping function and effectively expands the use range of the lane keeping function.
  • the present application provides a vehicle lane keeping control method.
  • the method includes: acquiring location information of real road elements in a vehicle driving environment that affect the vehicle's driving; and generating virtual geographic information according to the location information, where the virtual geographic information includes First virtual geographic information corresponding to real road elements and second virtual first information corresponding to missing road elements; sending virtual geographic information to the vehicle to instruct the vehicle to keep lanes according to the virtual geographic information and data collected by the vehicle perception system.
  • the method is mainly applied to various computing-capable devices such as cloud servers or roadside computing units. The above method reduces the influence of the weather environment on the lane keeping function by sending virtual geographic information to the vehicle to assist the vehicle to keep the lane.
  • the virtual geographic information also includes the second virtual geographic information corresponding to the missing road elements, so that the vehicle can drive normally even in the scenario where the original lane keeping function fails, which expands the use range of the lane keeping function and improves the driving performance of the vehicle. security.
  • the virtual geographic information includes at least one of a virtual lane line, a virtual driving guide line, and a drivable area of the vehicle.
  • the virtual driving guide line is the middle line of the virtual lane lines on both sides.
  • the vehicle can realize the function of lane keeping according to various types of virtual geographic information.
  • the missing road element when the missing road element is located in the lane area, the missing road element is a worn lane line; when the missing road element is located in the intersection area, the missing road element is the missing intersection lane line.
  • the lane line In the real driving environment of the vehicle, the lane line may be worn due to long-term use and lack of maintenance, or there is no lane line in some intersection areas, which often lead to the failure of the vehicle's lane keeping function.
  • the method provided by the embodiments of the present application can generate virtual geographic information corresponding to these missing road elements, improve the accuracy of the lane keeping function, and reduce human intervention.
  • acquiring the location information of the real road element includes: accepting perception data sent by a roadside perception system, where the perception system includes at least one of a camera, a millimeter-wave radar, and a lidar; Perceptual data generates location information for real road elements.
  • the roadside perception system is a camera
  • the image coordinates of the real road elements can be obtained through the image captured by the camera, and then the position information of the real road elements can be obtained by converting the image coordinates of the real road elements into global coordinates according to the calibration relationship.
  • the roadside perception system is radar (or lidar)
  • the global position information of real road elements is obtained by scanning the radar.
  • acquiring the location information of the real road element includes: generating the location information of the real road element according to map data, where the map data includes road information of the driving area of the vehicle.
  • map data of the service area is stored in the cloud or roadside system, and the map data includes road information (such as road edge position, road gradient, etc.). Therefore, the location information of real road elements can be extracted directly from the map data. This implementation improves computational efficiency.
  • the method further includes: receiving road traffic information, where the road traffic information is used to indicate whether a road is available, and correcting the virtual geographic information according to the road traffic information.
  • the road traffic information may be sent by the traffic management platform, or reported by the construction unit or the traffic police. According to the road traffic information, it can be determined that some lanes are abnormally occupied, and then the virtual geographic information is corrected.
  • the generated virtual geographic information is corrected according to the road traffic information, the accuracy of the virtual geographic information is improved, and the driving safety of the vehicle is ensured.
  • the road traffic information includes at least one of road construction information, traffic accident information, road damage information and traffic flow information.
  • generating the virtual geographic information according to the position information of the real road element includes: generating first virtual geographic information according to the position information of the real road element, and then generating second virtual geographic information according to the first virtual geographic information .
  • generating the second virtual geographic information according to the first virtual geographic information includes: generating the endpoints and intersections of each lane line in the first virtual geographic information, and when the missing road element is located in the lane area, using A straight line directly connects the opposite end points; when the missing road element is located in the intersection area, a curve is used to connect the opposite end points, and the curvature of the curve can be determined according to the driving habits of the driver.
  • generating virtual geographic information according to the location information of real road elements includes: inputting the location information of real road elements into a pre-trained machine learning model to generate virtual geographic information.
  • the pre-trained machine learning model can be a support vector machine model or a neural network model. Generate virtual geographic information using pre-trained machine learning models, increasing computational efficiency.
  • the real road elements include one or more of real lane lines, road central dividing lines, road central green belts, real road edge lines, and real stop lines.
  • the vehicle implements the lane keeping function, it will use the on-board sensors to detect the location information of the above-mentioned real road elements to achieve lane keeping.
  • the present application provides a lane keeping control method for a vehicle, the method comprising: the vehicle receives virtual geographic information, where the virtual geographic information includes first virtual geographic information corresponding to real road elements and first virtual geographic information corresponding to missing road elements Second virtual geographic information; the vehicle performs lane keeping according to the virtual geographic information and data collected by the vehicle-mounted perception system.
  • this method expands the perception range of the vehicle and improves the stability and reliability of the lane keeping function.
  • the vehicle determines that it deviates from the lane according to data collected by the vehicle-mounted perception system or virtual geographic information, and pushes a warning prompt to the user or the vehicle management platform to control the vehicle to drive back into the lane.
  • the vehicle determines that it deviates from the expected lane range, it directly intervenes in the driving of the vehicle or reminds the driver to intervene to reduce the occurrence of traffic safety accidents.
  • the vehicle displays the received virtual geographic information to the user.
  • the user can see the operation of the vehicle in the virtual geographic information through the in-vehicle user interface. Improved user experience.
  • the virtual geographic information includes at least one of a virtual lane line, a virtual driving guide line, and a drivable area of the vehicle.
  • performing lane keeping according to the virtual geographic information and data collected by the vehicle-mounted perception system includes: the vehicle collects the coordinates of real road elements in a relative coordinate system through the vehicle-mounted perception system, and the relative coordinate system The origin is the vehicle; the vehicle converts the coordinates of the virtual geographic information into coordinates in the above-mentioned relative coordinate system; when the confidence level of the position data perceived by the on-board sensor is lower than the preset threshold, the received virtual geographic information is used for lane keeping .
  • the virtual geographic information includes lane lines, driving guidance lines, or a global coordinate system (eg, a northeast sky or a northeast ground coordinate system) of a drivable area of the vehicle.
  • the coordinates of the virtual geographic information need to be converted to realize lane keeping.
  • the vehicle's lane keeping strategy can be based on the position coordinates detected by its own sensors.
  • the confidence of the coordinates is lower than a preset threshold (for example, when the lane line is worn or the lane line is blocked by the vehicle ahead), the virtual geographic information for lane keeping.
  • the present application provides a lane keeping control method for a vehicle.
  • virtual geographic information is calculated by the vehicle.
  • the vehicle generates virtual geographic information based on the location information of real road elements, and performs lane keeping according to the virtual geographic information and data collected by on-board sensors. The application range and accuracy of the lane keeping function have been improved.
  • the vehicle determines that it deviates from the lane according to data collected by the vehicle-mounted perception system or virtual geographic information, and pushes a warning prompt to the user or the vehicle management platform to control the vehicle to drive back into the lane.
  • the vehicle displays the received virtual geographic information to the user.
  • the user can see the operation of the vehicle in the virtual geographic information through the in-vehicle user interface.
  • the virtual geographic information includes at least one of a virtual lane line, a virtual driving guide line, and a drivable area of the vehicle.
  • performing lane keeping according to the virtual geographic information and data collected by the vehicle-mounted perception system includes: the vehicle collects the coordinates of real road elements in a relative coordinate system through the vehicle-mounted perception system, and the relative coordinate system The origin is the vehicle; the vehicle converts the coordinates of the virtual geographic information into coordinates in the above-mentioned relative coordinate system; when the confidence level of the position data perceived by the on-board sensor is lower than the preset threshold, the received virtual geographic information is used for lane keeping .
  • the present application provides a device for controlling a vehicle to keep a lane
  • the device includes a communication module and a processing module, and the processing module is used to obtain position information of real road elements, wherein the real road elements are affected by the driving environment of the vehicle.
  • the actual existing road elements in which the vehicle is traveling; the processing module is further configured to generate virtual geographic information according to the location information, and the virtual geographic information includes the first virtual geographic information corresponding to the real road elements and the first virtual geographic information corresponding to the missing road elements.
  • Virtual geographic information the communication module is configured to send the virtual geographic information to the vehicle to instruct the vehicle to keep a lane according to the virtual geographic information.
  • the virtual geographic information includes at least one of a virtual lane line, a virtual driving guide line, and a drivable area of the vehicle.
  • the missing road element when the missing road element is located in a lane area, the missing road element is a worn lane line; when the missing road element is located in an intersection area, the missing road element is The road element is the missing lane line at the intersection.
  • the communication module is further configured to receive sensing data sent by the roadside perception system, where the roadside perception system includes at least one of a camera, a millimeter-wave radar, and a lidar; the processing module is further configured to The perception data generates location information for the real road elements.
  • the processing module is further configured to generate position information of the real road element according to map data, where the map data includes road information of the driving area of the vehicle.
  • the communication module is further configured to receive road traffic information, where the road traffic information is used to indicate whether a road is available; the processing module is further configured to correct the virtual geographic information according to the road traffic information.
  • the road traffic information includes at least one of road construction information, traffic accident information, road damage information and traffic flow information.
  • the processing module is further configured to: generate first virtual geographic information according to the position information of the real road element; and generate second virtual geographic information according to the first virtual geographic information.
  • the processing module is further configured to: input the location information into a pre-trained machine learning model to generate the virtual geographic information.
  • the real road elements include one or more of real lane lines, road central dividing lines, road central green belts, real road edge lines, and real stop lines.
  • the present application provides a vehicle with a lane keeping control function, the vehicle including a communication module, a processing module and an on-board sensor.
  • the communication module is configured to receive virtual geographic information, where the virtual geographic information includes first virtual geographic information corresponding to real road elements and second virtual geographic information corresponding to missing road elements;
  • the processing module is configured to receive virtual geographic information according to the virtual geographic information and
  • the data collected by the on-board sensors is used for lane keeping.
  • the processing module is further configured to determine the lane departure according to the data collected by the vehicle-mounted perception system or the virtual geographic information; the communication module is further configured to push an alarm prompt to the user or the vehicle management platform to control all drive the vehicle back into the specified lane.
  • the vehicle further includes a display module: the display module is configured to display the virtual geographic information to the user.
  • the virtual geographic information includes at least one of a virtual lane line, a virtual driving guide line, and a drivable area of the vehicle.
  • the processing module is further configured to: determine the coordinates of road elements within a preset range in front of the vehicle in a relative coordinate system according to data collected by the vehicle-mounted perception system, where the origin of the relative coordinate system is the vehicle; convert the coordinates of the virtual geographic information into the coordinates under the above-mentioned relative coordinate system; when the confidence of the coordinates of the road elements within the preset range in front of the vehicle is lower than the preset threshold, carry out according to the relative coordinates of the virtual geographic information. Lane keeping.
  • the present application provides another vehicle with a lane keeping control function.
  • the vehicle includes: a processing module and an on-board sensor.
  • the processing module is used to obtain position information of real road elements, wherein the real road elements are real road elements that affect the driving of the vehicle in the driving environment of the vehicle; it is also used to generate virtual geographic information according to the position information, and the virtual geographic information It includes the first virtual geographic information corresponding to the real road element and the second virtual geographic information corresponding to the missing road element; and is also used for lane keeping according to the virtual geographic information and the data collected by the vehicle-mounted sensor.
  • the processing module is further configured to determine the lane departure according to the data collected by the vehicle-mounted perception system or the virtual geographic information.
  • the vehicle further includes a communication module, and the communication module is configured to push a warning prompt to a user or a vehicle management platform to control the vehicle to drive back within a specified lane range.
  • the vehicle further includes a display module: the display module is configured to display the virtual geographic information to the user.
  • the virtual geographic information includes at least one of a virtual lane line, a virtual driving guide line, and a drivable area of the vehicle.
  • the processing module is further configured to: determine the coordinates of road elements within a preset range in front of the vehicle in a relative coordinate system according to data collected by the vehicle-mounted perception system, where the origin of the relative coordinate system is the vehicle; convert the coordinates of the virtual geographic information into the coordinates under the above-mentioned relative coordinate system; when the confidence of the coordinates of the road elements within the preset range in front of the vehicle is lower than the preset threshold, carry out according to the relative coordinates of the virtual geographic information. Lane keeping.
  • the present application provides a device for controlling a vehicle to implement lane keeping, the device comprising a processor and a memory, wherein: the memory stores computer instructions; the processor executes the computer instructions, so as to realize the above-mentioned first aspect and possibility Implement the method described in any one of the modes.
  • the present application provides a vehicle, which includes a processor and a memory, wherein: computer instructions are stored in the memory; the processor executes the computer instructions to implement various possibilities in the second and third aspects above any of the methods in the implementation.
  • the present application provides a computer-readable storage medium, including computer-executable instructions, which, when the computer-executable instructions are run on a computer, cause the computer to execute any one of the first aspect, the second aspect, or the third aspect. method.
  • the present application provides a computer program product comprising computer-executable instructions, which, when the computer-executable instructions are run on a computer, cause the computer to perform any one of the methods provided in the first aspect, the second aspect or the third aspect.
  • the present application provides a system for controlling a vehicle to achieve lane keeping, comprising: a device for controlling a vehicle to achieve lane keeping and a vehicle; wherein the device for controlling a vehicle to achieve lane keeping is configured to perform any one of the above-mentioned first aspect.
  • the vehicle is configured to execute any one of the methods provided in the second aspect above.
  • FIG. 1 is a schematic diagram of implementing a lane keeping function using an on-board sensor provided by an embodiment of the present application
  • FIG. 2 is a schematic diagram of a system architecture provided by an embodiment of the present application.
  • FIG. 2A is a schematic diagram of an application in a highway scenario provided by an embodiment of the present application.
  • 2B is a schematic diagram of an application in an urban road scene provided by an embodiment of the present application.
  • FIG. 3 is a schematic overall flowchart of a vehicle lane keeping control method provided by an embodiment of the present application.
  • FIG. 4 is a schematic flowchart of another vehicle lane keeping control method according to an embodiment of the present application.
  • 5A is a schematic diagram of first virtual geographic information provided by an embodiment of the present application.
  • 5B is a schematic diagram of an endpoint or an intersection point of the first virtual geographic information provided by an embodiment of the present application.
  • 5C is a schematic diagram of virtual geographic information provided by an embodiment of the present application.
  • 5D is a schematic diagram of another virtual geographic information provided by an embodiment of the present application.
  • 5E is a schematic diagram of another virtual geographic information provided by an embodiment of the present application.
  • FIG. 6 is a schematic diagram of corrected virtual geographic information provided by an embodiment of the present application.
  • FIG. 7 is a schematic flowchart of another vehicle lane keeping control method according to an embodiment of the present application.
  • FIG. 8 is a schematic diagram of a lane keeping control device according to an embodiment of the present application.
  • FIG. 9 is a schematic diagram of a vehicle provided by an embodiment of the present application.
  • FIG. 10 is a schematic diagram of a lane keeping control device for a vehicle according to an embodiment of the present application.
  • the present application provides a vehicle lane keeping control method, device and system, which pushes virtual geographic information to the vehicle through a roadside system or a cloud service platform, so as to solve the problem of lane keeping when the vehicle is driving in an area without lane lines or in an area with worn lane lines.
  • the problem of functional failure ensures driving safety.
  • Lane Keep Assist It is a function provided by the intelligent driving system, hereinafter referred to as "lane keep”. This function is mainly to perceive the lane lines through the on-board sensors and keep the vehicle driving within the specified lane range. When the vehicle approaches the identified marking line and may leave the driving lane, the driver can be alerted by vibration of the steering wheel or sound, or the vehicle controller can directly intervene in the vehicle's deviation behavior by controlling the brake coordination device to help the vehicle keep driving within the lane.
  • Real road elements real road elements that affect the driving of vehicles in the driving environment. For example, lane lines on roads, road medians/median strips, road edge lines, intersection stop lines, etc. By obtaining the location information of these road elements, the vehicle can be assisted to drive within the specified lane range.
  • Virtual Geographic Information Based on real road elements or pre-stored maps, virtual data containing real or non-real geographic information is constructed.
  • the virtual geographic information mainly includes a virtual lane line, a vehicle driving guide line, a vehicle travelable area, and the like. This information is mainly used to guide the vehicle to achieve the lane keeping function.
  • Roadside Sensors System Deployed on both sides of the road, including one or more sensing devices.
  • the perception device is used to perceive real road information (positions of real road elements) and road traffic conditions.
  • the sensing device may be a video detector (eg, a camera), or a radar detector (eg, millimeter-wave radar, lidar, microwave radar, ultrasonic radar, etc.).
  • Road Side Unit A communication device deployed on the road side of the road, which can send and receive messages with vehicles and cloud service platforms.
  • Edge computing (Multi-access Edge Computing, MEC) unit a computing service node deployed on the roadside or the edge of the core network, with the ability to process images, video data, and analyze and calculate virtual geographic information.
  • MEC Multi-access Edge Computing
  • Cloud Service Platform Deployed on one or more remote servers, providing powerful computing capabilities, it can communicate with vehicles and send and receive messages.
  • LKA Combined with ACC (Adaptive Cruise Control, active cruise control), LKA is the key technology to realize ADAS (Advanced Driving Assistance System, advanced assisted driving) and L2-L5 autonomous driving, and it is also an important guarantee for the safe driving of autonomous vehicles.
  • the vehicle realizes lane keeping through its own sensors and control system, including the following specific steps: (1) the on-board perception device obtains road images; (2) image processing, extracting and analyzing lane line information; (3) according to The calculated lane line information determines the lane keeping control strategy of the vehicle.
  • the above lane keeping control methods mainly rely on on-board equipment for real-time perception, the perception range is small, and it is easily affected by light and weather.
  • in-vehicle sensing devices When encountering low-visibility scenes such as glare and fog, in-vehicle sensing devices often fail to identify valid lane line information.
  • real road conditions often affect the implementation of the lane keeping function. For example, some lane markings may be worn out due to age, or at some traffic intersections there are no lane markings, or the lane markings are obscured by other vehicles. In the above scenario, the vehicle-mounted sensing device cannot perceive the real lane line information, which leads to the failure of the lane keeping function and requires manual intervention.
  • the embodiment of the present application provides a lane keeping control method for a vehicle, which can help the vehicle to use the lane keeping function normally even when the lane line is abnormal (for example, the lane line is missing, worn, blocked, etc.), thereby improving the driving performance. Safety.
  • the present application provides a system architecture including a cloud service platform 21 , a vehicle 22 , and a roadside system 23 .
  • the cloud service platform 21 is deployed on a cloud server, can provide various computing services, and can communicate with the vehicle 22 and the roadside system 23 .
  • the cloud service platform 21 is mainly used to calculate and generate virtual geographic information and send it to the vehicle 22 , or forward the virtual geographic information generated by the calculation to the vehicle 22 through the roadside system 23 .
  • the vehicle 22 includes a communication module 221 , a positioning module 222 and a processing module 223 .
  • the positioning module 222 is used to obtain the position information of the vehicle 22 itself.
  • the communication module 221 is used for communicating with the cloud service platform 21 or the roadside system 23, reporting the position information of the vehicle, and receiving virtual geographic information.
  • the processing module 223 is mainly used for processing and calculating according to the received virtual geographic information, so as to control the vehicle to realize the lane keeping function. It should be noted that the virtual geographic information may be sent to the vehicle after being calculated by the cloud service platform or the roadside system, or may be generated by the vehicle itself.
  • the roadside system 23 includes a roadside perception system 231 , a roadside unit RSU232 and an edge computing node MEC233 .
  • Roadside sensing system 231 includes one or more sensing devices, such as cameras or radars.
  • the data (video image data or radar data) collected by the roadside perception system 231 is used to realize two functions, one is used to detect the position information of real road elements, and the other is used to calculate the road Travel information.
  • the roadside unit RSU232 is a communication module that uses vehicle-road communication (including but not limited to LVE-V, 4G/5G, DSRC) technology to communicate with the on-board unit (OBU, On Board Unit), using wired network or mobile cellular network, etc.
  • vehicle-road communication including but not limited to LVE-V, 4G/5G, DSRC
  • the wireless network communicates with the cloud server.
  • the edge computing node MEC233 provides computing capabilities for computing road traffic information based on the data collected by the perception system.
  • the edge computing node MEC is further configured to calculate and generate virtual geographic information according to the location information of real road elements.
  • FIG. 2A and FIG. 2B show two specific application scenarios provided by the embodiments of the present application.
  • FIG. 2A shows a schematic diagram of an application scenario of an embodiment of the present application in a highway scenario.
  • the system mainly includes vehicle A, vehicle B, cloud service platform 21 and roadside system 23 .
  • the lane line in front of vehicle A is worn.
  • the lane keeping function may fail and cause a series of safety problems.
  • the method provided by this embodiment of the present application can generate virtual geographic information ( 242 ) according to the position information of real road elements (eg, lane lines in 241 ), and send the virtual geographic information to vehicle A, so that the lane keeping function of the vehicle can be Normal use.
  • there is a high-speed intersection area in front of vehicle B and there is no lane line itself.
  • the method provided by the present application can generate virtual geographic information 252 in advance according to the position information of lane lines or expressway edge lines in 251, and push it to vehicle B, so that the vehicle can control itself to run within a reasonable lane range. It should be noted that the vehicles in 241, 242, 251, and 252 are only used to identify the location where the vehicle is traveling, and do not belong to virtual geographic information.
  • FIG. 2B shows a schematic diagram of a system in an urban road scene according to an embodiment of the present application. Similar to FIG. 2A , when the lane lines are worn or the lane lines do not exist at the urban intersection itself, the cloud service platform 21 or the computing unit 233 in the roadside system and other devices that can provide computing power can determine the location of the real road elements.
  • the information generates virtual geographic information 261 , 271 .
  • the generated virtual geographic information fills in the missing lane line information, thereby ensuring that the vehicle can control itself to drive within a reasonable lane range during the whole process of driving, reducing manual intervention and greatly improving safety.
  • the overall process of the vehicle lane keeping control method provided by the embodiment of the present application is described with reference to FIG. 3 .
  • the method can be applied to various devices that can provide computing capabilities, such as a cloud service platform or a computing unit 233 in a roadside system. As shown in the figure, the method includes steps S32-S36:
  • the real road elements are the road elements that affect the driving of the vehicle in the driving environment.
  • the real road elements include real road elements on the road, such as lane lines, road medians/median dividers, road edge lines, stop lines, and the like.
  • Manner 1 Obtain location information of real road elements through the roadside perception system 231 .
  • the image captured by the roadside camera is input into a pre-trained neural network (such as VGG-16, Faster-RCNN, etc.) to extract the coordinates of each line in the image, and then the above-mentioned real roads are classified according to the calibration relationship of the camera.
  • the image coordinates of the elements are converted into world coordinates, and finally the global position information of the real road elements can be obtained.
  • Method 2 The location information of real road elements is directly extracted from the map (which can be a high-precision map).
  • the pre-stored high-precision map can be directly called, and the position information of the lane line can be obtained according to the map data.
  • the local storage device can be searched for whether map data exists, and if it exists, it is called directly, and if it does not exist, the management platform is requested to obtain map data.
  • the virtual geographic information includes first virtual geographic information corresponding to real road elements and second virtual geographic information corresponding to missing road elements.
  • the global position information of the real road element can be obtained, and then the missing part of the road element can be determined (for example, a worn lane line or a lane line that does not exist in the intersection area itself).
  • Complete virtual geographic information can then be generated using basic mathematical methods or methods such as machine learning.
  • S36 Send the virtual geographic information to the vehicle, so that the vehicle performs lane keeping according to the virtual geographic information.
  • the vehicle drives into a preset area and reports its location, and then sends virtual geographic information based on the vehicle's location.
  • virtual geographic information within the expected driving range can be sent to the vehicle.
  • virtual geographic information within the size of a semi-circle with a radius of 500 meters in front of the vehicle is sent to the sender. If the virtual geographic information is calculated by the roadside system, the roadside system can directly push the virtual geographic information to vehicles entering the service range.
  • the cloud service platform can directly push the virtual geographic information to the vehicle through the mobile cellular data network or wireless network such as WI-FI, or it can first send the virtual geographic information to the road in different areas according to the location of each roadside system.
  • the side system pushes the virtual geographic information of the corresponding area, which is then forwarded to the vehicle by the roadside system.
  • this application does not specifically limit the device for calculating virtual geographic information and the device for sending virtual geographic information, depending on the actual situation.
  • step S36 before pushing the virtual geographic information to the vehicle (step S36), it further includes step S35:
  • the virtual geographic information generated in step S33 may be corrected according to the road traffic information.
  • the road traffic information is used to indicate whether the lane can be used normally, and exemplarily, includes traffic flow information, road construction information, road damage information, traffic accident information, and the like. Exemplarily, there are several ways to obtain road traffic information:
  • Method 1 Obtain through the roadside perception system. Surveillance cameras collect video data on the road, and can analyze the traffic flow data or traffic accident data on the road through deep learning or trajectory fitting.
  • Method 2 The traffic management platform pushes traffic accident information or traffic flow information.
  • Method 3 The construction unit reports the construction time and area.
  • the device that corrects the virtual geographic information and the device that calculates and generates the virtual geographic information are not necessarily the same device.
  • the cloud service platform may calculate and generate virtual geographic information, and then the cloud service platform pushes the information to the roadside system, and the roadside system corrects the virtual geographic information according to the real-time road traffic information and sends it to the vehicle.
  • the virtual geographic information can also be generated by the cloud service platform, and then corrected according to the real-time road traffic information sent by the roadside perception system, and then sent to the vehicle.
  • Steps S32-S36 generate virtual geographic information and send it to the vehicle, and the vehicle implements lane keeping in combination with the data collected by the vehicle-mounted perception system and the received virtual geographic information.
  • the vehicle can adopt a variety of strategies to achieve lane keeping.
  • the data collected by the in-vehicle perception system may be the main data, and when the confidence level of the collected data is lower than the preset threshold, the received virtual geographic information is used for lane keeping.
  • the lane departure warning system will issue a warning to remind the driver to perform manual intervention; or, according to the user's original settings, Directly control the vehicle back to the lane line.
  • the lane keeping system when the lane keeping system detects that the vehicle is in an abnormal driving condition and deviates significantly from the driving lane, it can also report the abnormal deviation to the vehicle management platform, and the vehicle management platform can adjust the driving conditions of the vehicle and the emergency degree of the abnormal event.
  • the operating vehicles shall take early warning measures or takeover measures to avoid serious traffic accidents caused by abnormal driving, such as driving into the opposite lane or falling into the lake.
  • the virtual geographic information generated in this embodiment of the present application includes virtual geographic information corresponding to real lane lines and virtual geographic information corresponding to missing lane lines. Pushed to the vehicle, the vehicle can still achieve the function of lane keeping normally when the lane line is abnormal (lane line wear, no lane line at intersection, lane line is blocked); at the same time, the vehicle calculates the cloud system or roadside system
  • the virtual geographic information generated by the device is combined with the data collected by the vehicle perception system, which effectively avoids the impact of extreme weather on the vehicle perception system, and expands the applicable area and applicable scenarios of the lane keeping system.
  • the embodiment of the present application further corrects the generated virtual geographic information according to the road traffic information, which greatly improves the accuracy of the virtual geographic information and reduces the impact of traffic accidents, road construction, etc. on the vehicle lane keeping function.
  • the computing unit 233 in the roadside system generates virtual geographic information to assist the vehicle in lane keeping
  • the virtual geographic information is calculated and generated by a computing unit (for example, MEC) in the roadside system and pushed to the vehicle, which specifically includes the following steps:
  • S41 The vehicle enters the service area of the roadside system, establishes a communication connection with the roadside equipment unit RSU and authenticates, reports the vehicle location information and requests service.
  • the vehicle may actively request the roadside system to push the virtual geographic information of the corresponding location, or the roadside system may actively push the virtual geographic information of the corresponding location to the vehicle according to the vehicle location after the roadside system detects the vehicle.
  • S42 The computing unit in the roadside system acquires the position information of the real road elements in the driving area of the vehicle.
  • real road elements include real lane lines, road medians/median strips, road edge lines, stop lines, and so on.
  • the computing unit in the roadside system can obtain the location information corresponding to the real road elements in the following two ways:
  • the local storage unit of the roadside system stores map data corresponding to the driving area of the vehicle.
  • the map can be a high-precision map or an ordinary electronic map.
  • the computing unit in the roadside system can directly extract the first virtual geographic information from the locally stored map data (as shown in Fig. 5A).
  • Method 2 Generate location information corresponding to real road elements in the area through image data uploaded by the roadside perception system.
  • the roadside perception system can be a camera, and the captured image data can be input into the pre-trained neural network to extract the feature information of the real road elements, and the image coordinates of the real road elements can be converted into world coordinates to generate The first virtual geographic information shown in FIG. 5A.
  • S43 Generate virtual geographic information according to the position information of the real road element.
  • FIG. 5A is the first virtual geographic information generated according to the position information of real road elements.
  • the wear of the lane lines in the lane area and the lack of lane lines in the intersection area can easily lead to the failure of the lane keeping function of the vehicle. Therefore, the embodiments of the present application generate virtual geographic information corresponding to real road elements and missing road elements, so as to ensure the normal use of the lane keeping function.
  • Virtual geographic information can be generated using basic mathematical methods or deep learning methods. Exemplarily, taking a basic mathematical method as an example, generating virtual geographic information according to the location information of real road elements includes the following steps:
  • each lane line and the stop line is generated according to the first virtual geographic information.
  • the missing part of the road element is in the lane area, it is directly connected by a straight line; if the missing part of the road element is in the intersection area, it is connected by a curve, and the curvature of the curve can be determined according to the driving habits of the driver.
  • the virtual geographic information includes not only the lane lines, but also the driving guidance lines (as shown in 5D) and the drivable area of the vehicle (as shown in 5E).
  • the location information of real road elements can also be directly input into the pre-trained machine learning model to generate virtual geographic information as shown in Figures 5C, 5D, and 5E.
  • the machine learning model can be a deep learning neural network such as ResNet and GoogleNet, or a machine learning classification model such as a support vector machine.
  • the road traffic data may be captured road image data or radar detection data.
  • the camera continuously collects road traffic flow, and generates a traffic flow trajectory by fitting. When it is found that there is no traffic flow trajectory in the lane with traffic flow, it means that the lane is abnormal, and the information is sent to the roadside computing unit.
  • S45 The roadside computing unit generates corrected virtual geographic information.
  • the roadside computing unit corrects the virtual geographic information generated in step S53 according to the road traffic information uploaded by the roadside perception system. As shown in Figure 6, when car E passes through the intersection, there are originally 1, 2, and 3 lanes to choose from, that is, three sets of virtual lane lines can be generated originally, but since lanes 2 and 3 have been occupied for various reasons, the road The side computing unit finally retains only the virtual lane line corresponding to lane 1.
  • S47 The vehicle realizes lane keeping in combination with the data collected by the vehicle-mounted perception system and the received virtual geographic information.
  • Virtual geographic information is global data, and lane lines are also coordinates in the global coordinate system. Therefore, when a vehicle receives virtual geographic information, it first needs to convert the virtual geographic information into relative coordinates centered on the vehicle, so as to achieve lane keeping.
  • the virtual geographic information is displayed on the user interface, so that the driver or passenger can pay attention to the driving state of the vehicle in real time and improve the user experience.
  • the present application does not specifically limit the lane keeping strategy for the vehicle to combine the vehicle-mounted perception data with the virtual geographic information data, and it depends on the situation.
  • the vehicle system may preset a threshold, and when the confidence level of the vehicle-mounted perception system in detecting the surrounding lane line information is lower than this threshold, the lane line data provided by the virtual geographic information is used, and when the confidence level is higher than the threshold, it is used. Data detected by in-vehicle perception systems. When the confidence level of the virtual geographic information is high, only the virtual geographic information may be used in the whole driving process.
  • step S41 the vehicle first reports the location request service, and then the roadside system generates virtual geographic information within a preset size range including the location according to the location of the vehicle.
  • the time for the vehicle to report the location is not limited to before the virtual geographic information is generated, and the vehicle can also report the location information after the roadside system has generated all the virtual geographic information within its service area. Request service.
  • the embodiments of the present application do not specifically limit the time when the vehicle reports the location to request the service.
  • the correction of the virtual geographic information may be performed continuously, and may be corrected according to the latest road condition data when the vehicle requests service.
  • the embodiments of the present application use the technical method of generating virtual geographic information such as virtual lane lines and guide lines and pushing them to the vehicle, to assist vehicle perception to obtain road lane information more comprehensively, and to ensure lanes even in areas without lane lines/blocked lanes Maintaining the normal use of the function reduces the influence of the weather environment and the limitations of the vehicle perception system on the lane keeping function, greatly expands the application scenarios, and improves driving safety.
  • the roadside perception system is set up high on the road, and the viewing angle is farther and better, which effectively improves the perception distance and accuracy of the vehicle.
  • the traditional lane keeping function does not correct the lane information perceived by the vehicle or obtained from the map.
  • the embodiment of the present application corrects the virtual geographic information according to the real-time road conditions to help the vehicle reduce traffic accidents, road construction and other events to the vehicle's lane keeping system. functional impact.
  • the cloud service platform generates virtual geographic information to assist vehicles in lane keeping
  • the virtual geographic information is calculated and generated by the cloud service platform and pushed to the vehicle, which specifically includes the following steps:
  • S71 The vehicle is connected to the cloud service platform, and after authentication, the location information is reported, and related services are requested.
  • the cloud service platform matches the location information of the vehicle with the map pre-stored in the background to determine the driving area of the vehicle.
  • the size of the area may be the size of a preset range.
  • the driving area of the vehicle may be a semicircle with the vehicle as the center and a radius of 500 meters in front of the vehicle.
  • the cloud service platform obtains the location information of the real road elements in the area from the map.
  • the cloud service platform will store global map data, which can be ordinary electronic maps or high-precision maps.
  • the cloud service platform can directly extract the first virtual geographic information corresponding to the real road elements in the driving area of the vehicle from the map.
  • the virtual geographic information extracted from the map still has missing parts (such as intersections), so the virtual geographic information needs to be supplemented.
  • it can be supplemented by the most basic geometric method, and the curvature of the intersection curve is determined according to the driving habits of the driver.
  • S74 The roadside perception system reports road traffic information to the cloud service platform.
  • Roadside perception systems can identify abnormal areas on the road based on real-time detection data.
  • the public platform of the public security can also report accident information or road construction conditions directly reported by road construction units.
  • the embodiments of the present application do not specifically limit the source of the real-time road condition information.
  • S75 The cloud service platform corrects the virtual geographic information according to the road traffic information.
  • S76 The cloud service platform pushes virtual geographic information to the vehicle.
  • the cloud service platform can push directly to the vehicle or forward it via the roadside system.
  • S77 The vehicle realizes lane keeping in combination with the data collected by the vehicle-mounted perception system and the received virtual geographic information.
  • the virtual geographic information is composed of coordinate points in the global coordinate system. Before the vehicle uses the virtual geographic information, it is necessary to first convert the coordinates of the virtual geographic information in the global coordinate system into relative coordinates centered on the vehicle, which is convenient for use.
  • This embodiment of the present application does not specifically limit the vehicle's lane keeping strategy based on on-board perception and virtual geographic information.
  • the vehicle system pre-stores a threshold, and the on-board perception system generates a confidence level when detecting the lane line. When the confidence level is lower than the above When the threshold is exceeded, virtual geographic information is used for lane keeping, and when the confidence level is higher than the above threshold, the vehicle perception data is used for lane keeping.
  • the cloud system determines the area according to the position reported by the vehicle, and then generates the virtual geographic information of the area.
  • the cloud system can also calculate the global virtual geographic information in advance and store it in the storage unit.
  • the vehicle reports the location to request service
  • the road traffic data of the corresponding area is obtained to the pre-stored virtual geographic information of the area.
  • the geographic information is corrected and then pushed to the vehicle.
  • virtual geographic information such as virtual lane lines and guide lines is generated through a cloud service platform and pushed to the vehicle to help the vehicle obtain lane information more comprehensively. It can also ensure that the lane keeping function is available in scenarios such as occlusion, etc.), which expands the application scenarios; compared with only using vehicle perception for lane keeping, the lane keeping strategy of vehicle perception combined with virtual geographic information provided by the embodiment of the present application greatly improves the vehicle's performance.
  • the perception distance improves driving safety; the generated virtual geographic information is corrected through road condition data, reducing the impact of traffic accidents, road construction and other events on the lane keeping function.
  • the vehicle can also be used as a computing device to generate virtual geographic information, and implement the lane keeping function according to the virtual geographic information.
  • the method includes: the vehicle is based on data or maps collected by the roadside perception system. The data generates position information of real road elements, and then generates virtual geographic information according to the position information, wherein the virtual geographic information includes the first virtual geographic information corresponding to the real road elements and the second virtual geographic information corresponding to the missing road elements; The data collected by the in-vehicle perception system and virtual geographic information realize lane keeping.
  • the vehicle sends the calculated virtual geographic information to the roadside system or the cloud service system. When other vehicles enter the area, the roadside system or the cloud service system first performs the virtual geographic information on the roadside system or the cloud service system. Correction, and then send the corrected virtual geographic information to other vehicles.
  • Communication module 801 and processing module 802 may be included:
  • the processing module 802 is configured to obtain location information of real road elements, where the real road elements are real road elements that affect the driving of the vehicle in the vehicle driving environment, and is also configured to generate virtual geographic information according to the location information, the The virtual geographic information includes first virtual geographic information corresponding to the real road element and second virtual geographic information corresponding to the missing road element;
  • the communication module 801 is configured to send the virtual geographic information to the vehicle to instruct the vehicle to keep a lane according to the virtual geographic information.
  • the communication module 801 is further configured to receive sensing data sent by a roadside sensing system, where the roadside sensing system includes at least one of a camera, a millimeter-wave radar, and a lidar; the processing module 802 also uses for generating the position information of the real road element according to the perception data.
  • the processing module 802 is further configured to generate position information of the real road element according to map data, where the map data includes road information of the driving area of the vehicle.
  • the communication module 801 is further configured to receive road traffic information, where the road traffic information is used to indicate whether a road is available; the processing module 802 is further configured to correct the virtual geographic information according to the road traffic information.
  • the processing module 802 is further configured to: generate the first virtual geographic information according to the location information, and generate the second virtual geographic information according to the first virtual geographic information.
  • the processing module 802 is further configured to: input the location information into a pre-trained machine learning model to generate the virtual geographic information.
  • a vehicle with a lane keeping control function includes: a communication module 901 , a processing module 902 and an on-board sensor 903 .
  • the communication module 901 is configured to receive virtual geographic information, where the virtual geographic information includes first virtual geographic information corresponding to real road elements and second virtual geographic information corresponding to missing road elements;
  • the processing module 902 is configured to The virtual geographic information and the data collected by the vehicle-mounted sensor 903 are used for lane keeping.
  • the processing module 902 is further configured to determine lane departure according to the data collected by the vehicle-mounted perception system or the virtual geographic information; the communication module 901 is further configured to push an alarm prompt to the user or the vehicle management platform to The vehicle is controlled to return within the lane range.
  • the vehicle further includes a display module 904 for displaying the virtual geographic information to the user.
  • the processing module 902 is further configured to determine, according to the data collected by the vehicle-mounted perception system, the coordinates of road elements within a preset range in front of the vehicle in a relative coordinate system, where the origin of the relative coordinate system is the vehicle. ; Convert the coordinates of the virtual geographic information into coordinates under the relative coordinate system, and the origin of the relative coordinate system is the vehicle; when the confidence level of the coordinates of the road elements within the preset range in front of the vehicle is low At a preset threshold, lane keeping is performed according to the relative coordinates of the virtual geographic information.
  • the device 1000 includes a processor 1001 , a storage medium 1002 , a communication interface 1003 and a memory unit 1004 .
  • the processor 1001 , the storage medium 1002 , the communication interface 1003 , and the memory unit 1004 communicate through the bus 1005 , and can also communicate through other means such as wireless transmission.
  • the memory 1002 is used for storing instructions, and the processor 1001 is used for executing the instructions stored in the memory 1002 .
  • the memory 1002 stores program codes, and the processor 1001 can call the program codes stored in the memory 1002 to perform the following operations:
  • the virtual geographic information is sent to the vehicle, so that the vehicle performs lane keeping according to the data collected by the vehicle-mounted perception system and the virtual geographic information.
  • the processor 1001 may be a CPU, and the processor 1001 may also be other general-purpose processors, digital signal processors (DSPs), application specific integrated circuits (ASICs), field programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc.
  • DSPs digital signal processors
  • ASICs application specific integrated circuits
  • FPGA field programmable gate arrays
  • a general purpose processor may be a microprocessor or any conventional processor or the like.
  • the memory 1002 may include read only memory and random access memory, and provides instructions and data to the processor 1001 .
  • Memory 1002 may also include non-volatile random access memory.
  • the memory 1002 may also store information of device type.
  • the memory 1002 may be volatile memory or non-volatile memory, or may include both volatile and non-volatile memory.
  • the non-volatile memory may be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (erasable PROM, EPROM), electrically programmable Erase programmable read-only memory (electrically EPROM, EEPROM) or flash memory.
  • Volatile memory may be random access memory (RAM), which acts as an external cache.
  • RAM static random access memory
  • DRAM dynamic random access memory
  • SDRAM synchronous dynamic random access memory
  • Double data rate synchronous dynamic random access memory double data date SDRAM, DDR SDRAM
  • enhanced synchronous dynamic random access memory enhanced SDRAM, ESDRAM
  • synchronous link dynamic random access memory direct rambus RAM, DR RAM
  • bus 1005 may also include a power bus, a control bus, a status signal bus, and the like.
  • bus 105 the various buses are labeled as bus 105 in the figure.
  • the device 1000 according to the embodiment of the present application may correspond to the management apparatus 800 in the embodiment of the present application, and may correspond to the execution of the roadside MEC in FIG. 4 or the cloud service platform in FIG. 7 according to the embodiment of the present application , and are not repeated here for brevity.
  • the present application also provides a vehicle with a lane keeping control function.
  • the structure of the vehicle includes a processor, a storage medium, a communication interface and a memory unit.
  • the processor, the storage medium, the communication interface, and the memory unit communicate through a bus, and can also communicate through other means such as wireless transmission.
  • the memory is used to store instructions, and the processor is used to execute the instructions stored by the memory.
  • the memory stores program codes, and the processor 1001 can call the program codes stored in the memory 1002 to execute the operation steps performed by the vehicle in the method shown in FIG. 4 or FIG. 7 , which are not repeated here for brevity.
  • the present application also provides a lane keeping control system.
  • the system includes a roadside system as shown in Figure 4 and a vehicle.
  • the system includes a cloud service platform and a vehicle as shown in FIG. 7 .
  • Each device in the system executes the method shown in each figure, which is not repeated here for brevity.
  • the above embodiments may be implemented in whole or in part by software, hardware, firmware or any other combination.
  • the above-described embodiments may be implemented in whole or in part in the form of a computer program product.
  • the computer program product includes one or more computer instructions. When the computer program instructions are loaded or executed on a computer, all or part of the processes or functions described in the embodiments of the present application are generated.
  • the computer may be a general purpose computer, special purpose computer, computer network, or other programmable device.
  • the computer instructions may be stored in or transmitted from one computer readable storage medium to another computer readable storage medium, for example, the computer instructions may be downloaded from a website site, computer, server or data center Transmission to another website site, computer, server, or data center is by wire (eg, coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (eg, infrared, wireless, microwave, etc.).
  • the computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server, a data center, or the like that contains one or more sets of available media.
  • the usable media may be magnetic media (eg, floppy disks, hard disks, magnetic tapes), optical media (eg, DVDs), or semiconductor media.
  • the semiconductor medium may be a solid state drive (SSD).

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Automation & Control Theory (AREA)
  • Transportation (AREA)
  • Mechanical Engineering (AREA)
  • Mathematical Physics (AREA)
  • General Physics & Mathematics (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Atmospheric Sciences (AREA)
  • Traffic Control Systems (AREA)

Abstract

一种车辆（22）的车道保持控制方法、装置及系统，该方法包括：根据真实道路元素的位置信息生成虚拟地理信息，以使得车辆（22）根据该虚拟地理信息和车载感知系统采集的数据进行车道保持。其中，虚拟地理信息包括真实道路元素对应的第一虚拟地理信息以及缺失的道路元素对应的第二虚拟地理信息。

Description

一种车辆的车道保持控制方法、装置及系统 技术领域
本申请涉及自动驾驶领域，尤其涉及一种车辆的车道保持控制方法、装置及系统。
背景技术
车道保持辅助(Lane Keep Assistant，简称LKA)是实现自动驾驶的重要功能之一。车辆在行驶过程中，通过车载摄像头或者激光雷达，识别当前行驶车道的车道线，并将车道线信息传输到车辆计算系统中，实时判断车辆是否运行在车道范围内，进而确保行车安全。
LKA是实现ADAS(Advanced Driving Assistance System，高级辅助驾驶)和Level2-Level5自动驾驶的关键技术，也是确保自动驾驶车辆行车安全的重要技术之一。该技术的数据输入主要依赖于车载传感器对道路两旁的车道线进行检测，而车载传感器易受天气影响且感知范围十分有限，无法有效地获取前方较远区域的车道信息。此外，在车辆实际的行驶过程中，城市道路和高速公路的部分区域存在车道线磨损甚至无车道线的情况，进而导致车道范围无法被车载设备准确感知，最终限制了车道保持功能的使用范围和使用效果。
发明内容
本申请提供了一种车辆的车道保持控制方法、装置及系统，减少了天气环境对车道保持功能的影响，有效地扩大了车道保持功能的使用范围。
第一方面，本申请提供一种车辆的车道保持控制方法，该方法包括：获取车辆行车环境中影响车辆行驶的真实道路元素的位置信息；根据该位置信息生成虚拟地理信息，该虚拟地理信息包括真实道路元素对应的第一虚拟地理信息以及缺失的道路元素对应的第二虚拟第一信息；向车辆发送虚拟地理信息以指示车辆根据该虚拟地理信息和车载感知系统采集的数据进行车道保持。该方法主要应用于云端服务器或者路侧计算单元等各种具备计算能力的设备。上述方法通过向车辆发送虚拟地理信息以辅助车辆进行车道保持，减少了天气环境等对车道保持功能的影响。此外，虚拟地理信息中还包含了缺失的道路元素对应的第二虚拟地理信息，使得车辆在原本车道保持功能失效的场景下也能正常行驶，扩大了车道保持功能的使用范围，提高了车辆行驶的安全性。
在一种可能的实现方式中，虚拟地理信息包括虚拟车道线、虚拟行车引导线、车辆可行驶区域中的至少一种。其中，虚拟行车引导线是两侧虚拟车道线的中间线。车辆可以根据多种类型的虚拟地理信息实现车道保持的功能。
在另一种可能的实现方式中，当缺失的道路元素位于车道区域时，该缺失的道路元素为被磨损的车道线；当缺失的道路元素位于路口区域时，该缺失的道路元素为路口缺失的车道线。在车辆真实的行车环境中，车道线可能因为长期使用缺乏维护而被磨损，或者在一些路口区域本身就不存在车道线，这些情况往往会导致车辆的车道保持功能失效。而本 申请实施例提供的方法可以生成这些缺失道路元素对应的虚拟地理信息，提高车道保持功能的准确性，减少人为干预。
在另一种可能的实现方式中，获取真实道路元素的位置信息包括：接受路侧感知系统发送的感知数据，其中，感知系统包括摄像机、毫米波雷达、激光雷达中的至少一种；根据该感知数据生成真实道路元素的位置信息。当路侧感知系统为摄像机时，通过该摄像机拍摄的图像即可得到真实道路元素的图像坐标，再根据标定关系将真实道路元素的图像坐标转换成全局坐标即可得到真实道路元素的位置信息。当路侧感知系统为雷达(或者激光雷达)时，通过该雷达扫描得到真实道路元素的全局位置信息。
在另一种可能的实现方式中，获取真实道路元素的位置信息包括：根据地图数据生成所述真实道路元素的位置信息，所述地图数据包括所述车辆的行驶区域的道路信息。通常，云端或者路侧系统中会存储有服务区域的地图数据，该地图数据包括了道路的信息(比如道路边缘位置、道路坡度等等)。因此，直接从地图数据中就可以提取真实道路元素的位置信息。该实现方式提高了计算效率。
在另一种可能的实现方式中，所述方法还包括：接收道路交通信息，该道路交通信息用于指示道路是否可用，根据所述道路交通信息校正所述虚拟地理信息。该道路交通信息可以是交通管理平台发送的，也可以是施工单位或者交通警察上报的。根据道路交通信息可以判断出部分车道被异常占用，进而对虚拟地理信息进行校正。在本实现方式中，根据道路交通信息对生成的虚拟地理信息进行了校正，提高了虚拟地理信息的准确度，确保了车辆的行车安全。
在另一种可能的实现方式中，道路交通信息包括：道路施工信息、交通事故信息、道路损坏信息和车流信息中的至少一种。
在另一种可能的实现方式中，根据真实道路元素的位置信息生成虚拟地理信息包括：根据真实道路元素的位置信息生成第一虚拟地理信息，然后根据第一虚拟地理信息生成第二虚拟地理信息。
在另一种可能的实现方式中，根据第一虚拟地理信息生成第二虚拟地理信息包括：生成第一虚拟地理信息中各车道线的端点以及交点，当缺失的道路元素位于车道区域时，使用直线直接连接相对的端点；当缺失的道路元素位于路口区域时，使用曲线连接相对的端点，曲线的曲率可以根据驾驶员的行车习惯所确定。
在另一种可能的实现方式中，根据真实道路元素的位置信息生成虚拟地理信息包括：将真实道路元素的位置信息输入至预训练的机器学习模型中生成虚拟地理信息。预训练的机器学习模型可以是支持向量机模型或者神经网络模型等。使用预训练的机器学习模型生成虚拟地理信息，提高了计算效率。
在另一种可能的实现方式中，真实道路元素包括真实的车道线、道路中央分隔线、道路中央绿化带、真实的道路边缘线、真实的停车线中的一种或多种。车辆在实现车道保持功能的时候，会使用车载传感器探测上述真实道路元素的位置信息以实现车道保持。
第二方面，本申请提供一种车辆的车道保持控制方法，所述方法包括：车辆接收虚拟地理信息，该虚拟地理信息包括真实的道路元素对应的第一虚拟地理信息以及缺失的道路元素对应的第二虚拟地理信息；该车辆根据所述虚拟地理信息以及车载感知系统采集的数 据进行车道保持。比起单纯地仅仅通过车载传感器采集的数据进行车道保持，该方法扩大了车辆的感知范围，提高了车道保持功能的稳定性以及可靠性。
在一种实现方式中，车辆根据车载感知系统采集的数据或者虚拟地理信息确定自身偏离车道，向用户或者车辆管理平台推送告警提示以控制所述车辆驶回所述车道范围内。当车辆判断自身偏离预计的车道范围时，直接干预车辆的行驶或者提醒驾驶员介入以减少交通安全事故的发生。
在另一种可能实现方式中，车辆向用户显示接收到的虚拟地理信息。用户可以通过车载用户界面看到车辆在虚拟地理信息中的运行情况。提高了用户体验。
在另一种可能的实现方式中，虚拟地理信息包括虚拟车道线、虚拟行车引导线、车辆可行驶区域中的至少一种。
在另一种可能的实现方式中，根据所述虚拟地理信息以及车载感知系统采集的数据进行车道保持包括：车辆通过车载感知系统采集真实道路元素在相对坐标系下的坐标，该相对坐标系的原点为该车辆；车辆将虚拟地理信息的坐标转换成上述相对坐标系下的坐标；当通过车载传感器感知的位置数据的置信度低于预设阈值时，采用接收到的虚拟地理信息进行车道保持。虚拟地理信息包括的是车道线、行车引导线或者车辆可行驶区域的全局坐标系(例如东北天或者北东地坐标系)。因此，当车辆接收到虚拟地理信息时，需要将虚拟地理信息的坐标进行转换，以实现车道保持。车辆的车道保持策略可以是以自身传感器探测到的位置坐标为主，当该坐标的置信度低于预设阈值时(例如车道线磨损、车道线被前方车辆遮挡的情况下)，采用虚拟地理信息进行车道保持。
第三方面，本申请提供一种车辆的车道保持控制方法，在该方法中，虚拟地理信息由车辆计算，所述方法包括:车辆获取真实道路元素的位置信息，其中，真实道路元素为车辆行车环境中影响所述车辆行驶的真实存在的道路元素；根据该位置信息生成虚拟地理信息，所述虚拟地理信息包括真实的道路元素对应的第一虚拟地理信息以及缺失的道路元素对应的第二虚拟地理信息；根据虚拟地理信息以及车载感知系统采集的数据进行车道保持。在该实施例中，车辆基于真实道路元素的位置信息生成了虚拟地理信息，并且根据该虚拟地理信息和车载传感器采集的数据进行车道保持。提高了车道保持功能的应用范围以及准确度。
在一种可能的实现方式中，车辆根据车载感知系统采集的数据或者虚拟地理信息确定自身偏离车道，向用户或者车辆管理平台推送告警提示以控制所述车辆驶回所述车道范围内。
在另一种可能的实现方式中，车辆向用户显示接收到的虚拟地理信息。用户可以通过车载用户界面看到车辆在虚拟地理信息中的运行情况。
在另一种可能的实现方式中，虚拟地理信息包括虚拟车道线、虚拟行车引导线、车辆可行驶区域中的至少一种。
在另一种可能的实现方式中，根据所述虚拟地理信息以及车载感知系统采集的数据进行车道保持包括：车辆通过车载感知系统采集真实道路元素在相对坐标系下的坐标，该相对坐标系的原点为该车辆；车辆将虚拟地理信息的坐标转换成上述相对坐标系下的坐标；当通过车载传感器感知的位置数据的置信度低于预设阈值时，采用接收到的虚拟地理信息 进行车道保持。
第四方面，本申请提供一种控制车辆进行车道保持的装置，该装置包括通信模块和处理模块，处理模块用于获取真实道路元素的位置信息，其中，真实道路元素为车辆行车环境中影响所述车辆行驶的真实存在的道路元素；处理模块还用于根据所述位置信息生成虚拟地理信息，虚拟地理信息包括所述真实的道路元素对应的第一虚拟地理信息以及缺失的道路元素对应的第二虚拟地理信息；通信模块用于向车辆发送该虚拟地理信息以指示车辆根据所述虚拟地理信息进行车道保持。
在另一种可能的实现方式中，虚拟地理信息包括虚拟车道线、虚拟行车引导线、车辆可行驶区域中的至少一种。
在另一种可能的实现方式中，当所述缺失的道路元素位于车道区域时，所述缺失的道路元素为被磨损的车道线；当所述缺失的道路元素位于路口区域时，所述缺失的道路元素为所述路口缺失的车道线。
在另一种可能的实现方式中，通信模块还用于接收路侧感知系统发送的感知数据，路侧感知系统包括摄像机、毫米波雷达、激光雷达中的至少一种；处理模块还用于根据所述感知数据生成所述真实道路元素的位置信息。
在另一种可能的实现方式中，处理模块还用于根据地图数据生成所述真实道路元素的位置信息，所述地图数据包括所述车辆的行驶区域的道路信息。
在另一种可能的实现方式中，通信模块还用于接收道路交通信息，所述道路交通信息用于指示道路是否可用；处理模块还用于根据所述道路交通信息校正所述虚拟地理信息。
在另一种可能的实现方式中，道路交通信息包括：道路施工信息、交通事故信息、道路损坏信息和车流信息中的至少一种。
在另一种可能的实现方式中，处理模块还用于：根据真实道路元素的位置信息生成第一虚拟地理信息；根据第一虚拟地理信息生成第二虚拟地理信息。
在另一种可能的实现方式中，处理模块还用于：将所述位置信息输入至预训练的机器学习模型生成所述虚拟地理信息。
在另一种可能的实现方式中，真实道路元素包括真实的车道线、道路中央分隔线、道路中央绿化带、真实的道路边缘线、真实的停车线中的一种或多种。
第五方面，本申请提供一种具有车道保持控制功能的车辆，该车辆包括通信模块、处理模块以及车载传感器。通信模块用于，接收虚拟地理信息，该虚拟地理信息包括真实的道路元素对应的第一虚拟地理信息以及缺失的道路元素对应的第二虚拟地理信息；处理模块用于根据所述虚拟地理信息以及所述车载传感器采集的数据进行车道保持。
在另一种可能的实现方式中，处理模块还用于，根据车载感知系统采集的数据或者所述虚拟地理信息确定偏离车道；通信模块还用于向用户或者车辆管理平台推送告警提示以控制所述车辆驶回规定的车道范围内。
在另一种可能的实现方式中，该车辆还包括显示模块：显示模块用于向用户显示所述虚拟地理信息。
在另一种可能的实现方式中，虚拟地理信息包括虚拟车道线、虚拟行车引导线、车辆可行驶区域中的至少一种。
在另一种可能的实现方式中，处理模块还用于：根据车载感知系统采集的数据确定车辆前方预设范围内的道路元素在相对坐标系下的坐标，该相对坐标系的原点为所述车辆；将虚拟地理信息的坐标转换成上述相对坐标系下的坐标；当该车辆前方预设范围内的道路元素的坐标的置信度低于预设阈值时，根据该虚拟地理信息的相对坐标进行车道保持。
第六方面，本申请提供另一种具有车道保持控制功能的车辆。该车辆包括：处理模块以及车载传感器。处理模块用于获取真实道路元素的位置信息，其中，真实道路元素为车辆行车环境中影响所述车辆行驶的真实存在的道路元素；还用于根据所述位置信息生成虚拟地理信息，虚拟地理信息包括所述真实的道路元素对应的第一虚拟地理信息以及缺失的道路元素对应的第二虚拟地理信息；还用于根据该虚拟地理信息和车载传感器采集的数据进行车道保持。
在另一种可能的实现方式中，处理模块还用于，根据车载感知系统采集的数据或者所述虚拟地理信息确定偏离车道。该车辆还包括通信模块，通信模块用于向用户或者车辆管理平台推送告警提示以控制所述车辆驶回规定的车道范围内。
在另一种可能的实现方式中，该车辆还包括显示模块：显示模块用于向用户显示所述虚拟地理信息。
在另一种可能的实现方式中，虚拟地理信息包括虚拟车道线、虚拟行车引导线、车辆可行驶区域中的至少一种。
在另一种可能的实现方式中，处理模块还用于：根据车载感知系统采集的数据确定车辆前方预设范围内的道路元素在相对坐标系下的坐标，该相对坐标系的原点为所述车辆；将虚拟地理信息的坐标转换成上述相对坐标系下的坐标；当该车辆前方预设范围内的道路元素的坐标的置信度低于预设阈值时，根据该虚拟地理信息的相对坐标进行车道保持。
第七方面，本申请提供一种控制车辆实现车道保持的设备，该设备包括处理器和存储器，其中：存储器中存储有计算机指令；处理器执行所述计算机指令，以实现上述第一方面及可能实现方式中任一所述的方法。
第八方面，本申请提供一种车辆，该车辆包括处理器和存储器，其中：存储器中存储有计算机指令；处理器执行所述计算机指令，以实现上述第二方面、第三方面中各种可能实现方式中的任一所述方法。
第九方面，本申请提供一种计算机可读存储介质，包括计算机执行指令，当该计算机执行指令在计算机上运行时，使得计算机执行第一方面、第二方面或第三方面提供的任意一种方法。
第十方面，本申请提供一种计算机程序产品，包含计算机执行指令，当该计算机执行指令在计算机上运行时，使得计算机执行第一方面、第二方面或第三方面提供的任意一种方法。
第十一方面，本申请提供一种控制车辆实现车道保持的系统，包括：控制车辆实现车道保持的装置和车辆；其中，控制车辆实现车道保持的装置用于执行上述第一方面提供的任意一种方法，车辆用于执行上述第二方面提供的任意一种方法。
第三方面至第十一方面中的任一种实现方式所带来的技术效果可参见第一方面或第二方面中对应实现方式所带来的技术效果，此处不再赘述。
需要说明的是，在方案不矛盾的前提下，上述各个方面中的方案均可以结合。
附图说明
图1为本申请实施例提供的利用车载传感器实现车道保持功能的示意图；
图2为本申请实施例提供的系统架构示意图；
图2A为本申请实施例提供的在高速公路场景下的应用示意图；
图2B为本申请实施例提供的在城市道路场景下的应用示意图；
图3为本申请实施例提供的车辆的车道保持控制方法的整体流程示意图；
图4为本申请实施例提供的另一种车辆的车道保持控制方法的流程示意图；
图5A为本申请实施例提供的第一虚拟地理信息示意图；
图5B为本申请实施例提供的第一虚拟地理信息的端点或交点示意图；
图5C为本申请实施例提供的虚拟地理信息示意图；
图5D为本申请实施例提供的另一种虚拟地理信息示意图；
图5E为本申请实施例提供的另一种虚拟地理信息示意图；
图6为本申请实施例提供的校正的虚拟地理信息示意图；
图7为本申请实施例提供的另一种车辆的车道保持控制方法的流程示意图；
图8为本申请实施例提供的一种车道保持控制装置示意图；
图9为本申请实施例提供的一种车辆示意图；
图10为本申请实施例提供的车辆的车道保持控制设备示意图。
具体实施方式
本申请提供一种车辆的车道保持控制方法、装置及系统，通过路侧系统或者云服务平台向车辆推送虚拟地理信息，解决了车辆行驶在无车道线区域或车道线磨损区域时车辆的车道保持功能失效的问题，确保了行车安全。
在介绍本申请提供的实施例之前，首先对本申请实施例中出现的一些术语进行简单介绍：
车道保持辅助(Lane Keep Assistant)：是智能驾驶系统提供的一种功能，以下简称“车道保持”。该功能主要是通过车载传感器对车道线进行感知，保持车辆行驶在规定的车道范围内。当车辆靠近识别到的标记线并可能脱离行驶车道，可通过方向盘的振动，或者声音来提醒驾驶员注意，或车载控制器直接通过控制刹车协调装置对车辆的偏离行为进行干预，帮助车辆保持行驶在车道范围内。
真实道路元素：行车环境中影响车辆行驶的真实存在的道路元素。例如道路上的车道线、道路中央分隔线/中央分隔带、道路边缘线、路口停车线等等。通过获取这些道路元素的位置信息，可以辅助车辆行驶在规定的车道范围内。
虚拟地理信息(Virtual Geographic Information)：基于真实的道路元素或预先存储的地图，构建的包含真实存在或非真实存在的地理信息的虚拟数据。虚拟地理信息在本申请的实施例中主要包括虚拟车道线、车辆行驶引导线、车辆可行驶区域等。这些信息主要用于引导车辆实现车道保持功能。
路侧感知系统(Roadside Sensors System)：部署在道路两旁，包括一种或多种感知设备。其中，感知设备用于感知真实的道路信息(真实道路元素的位置)以及道路交通状况。示例性的，感知设备可以是视频检测器(例如摄像机)，也可以是雷达检测器(例如毫米波雷达、激光雷达、微波雷达、超声波雷达等)。
路侧单元(Road Side Unit，RSU)：部署在道路路侧的通信设备，可与车辆和云服务平台进行消息的发送和接收。
边缘计算(Multi-access Edge Computing，MEC)单元:部署在路侧或核心网边缘的计算服务节点，具备处理图像、视频数据以及分析计算虚拟地理信息的能力。
云服务平台(Cloud Service Platform):部署在远程的一个或多个服务器上,提供强大的计算能力，可与车辆进行通信，收发消息。
LKA结合ACC(Adaptive Cruise Control，主动定速巡航)，是实现ADAS(Advanced Driving Assistance System，高级辅助驾驶)和L2-L5自动驾驶的关键技术，更是自动驾驶车辆安全行驶的重要保证。如图1所示，车辆通过自身的传感器以及控制系统实现车道保持，包括如下具体步骤：(1)车载感知设备获取道路图像；(2)图像处理，提取并分析车道线信息；(3)根据计算得的车道线信息决定车辆的车道保持控制策略。然而，上述车道保持控制方法主要依赖于车载设备进行实时感知，感知范围较小，且易受光线、天气等影响。当遇到炫光、大雾等能见度低场景时，车载感知设备往往无法识别出有效的车道线信息。除了车载传感器自身的限制以外，真实的道路情况往往也会影响车道保持功能的实现。例如，一些车道线因为年代久远而导致磨损缺失，或者在一些交通路口原本就没有车道线，又或者车道线被其他车辆所遮挡。在上述场景下，车载感知设备无法感知到真实的车道线信息，进而导致车道保持功能失效，需要提醒人工进行干预。
本申请实施例提供了一种车辆的车道保持控制方法，可以帮助车辆在车道线异常(例如车道线缺失、磨损、被遮挡等等)的情况下也能正常使用车道保持的功能，进而提高行车安全。
参见图2，本申请提供了一种系统架构，包括云服务平台21、车辆22、路侧系统23。
云服务平台21部署在云端服务器上，可以提供各种计算服务，并且可以与车辆22和路侧系统23进行通信。在一种可能的实现方式中，云服务平台21主要用于计算生成虚拟地理信息并且发送给车辆22，或者通过路侧系统23将计算生成的虚拟地理信息转发给车辆22。
车辆22包括通信模块221、定位模块222、处理模块223。其中，定位模块222用于获取车辆22自身的位置信息。通信模块221用于与云服务平台21或者路侧系统23进行通信，上报车辆的位置信息，以及接收虚拟地理信息。处理模块223主要用于根据接收到的虚拟地理信息进行处理计算，以控制车辆实现车道保持功能。需要说明的是，虚拟地理信息可以是云服务平台计算或者路侧系统计算后发送给车辆的，也可以是车辆自己处理计算生成的。
路侧系统23包括路侧感知系统231、路侧单元RSU232以及边缘计算节点MEC233。路 侧感知系统231包括一个或多个感知设备，例如摄像机或者雷达。在本申请提供的实施例中，路侧感知系统231采集的数据(视频图像数据或者雷达数据)用于实现两种功能，一个用于是探测真实道路元素的位置信息，另一个是用于计算道路交通信息。路侧单元RSU232是一个通信模块，采用车路通信(包括但不限于LVE-V，4G/5G,DSRC)技术与车载单元(OBU，On Board Unit)进行通讯，采用有线网络或移动蜂窝网络等无线网络与云端服务器进行通讯。边缘计算节点MEC233提供计算能力，用于根据感知系统采集的数据计算道路交通信息。在一种可能的实现方式中，边缘计算节点MEC还用于根据真实道路元素的位置信息计算生成虚拟地理信息。
图2A以及图2B展示了本申请实施例提供的两种具体应用场景。
图2A展示了本申请实施例在高速公路场景下的应用场景示意图。系统中主要包括了车辆A、车辆B、云服务平台21以及路侧系统23。车辆A前方的车道线被磨损，在这样的场景下，车道保持功能可能会失效进而引发一系列的安全问题。而本申请实施例提供的方法可以根据真实道路元素(例如241中的车道线)的位置信息生成虚拟地理信息(242)，并且向车辆A发送该虚拟地理信息，以使得车辆的车道保持功能可以正常使用。同样的，车辆B前方是高速路口区域，本身不存在车道线，车辆在快要到达该区域时往往会提醒用户提前介入控制或者监控车辆，以顺利通过高速出口。而本申请提供的方法可以根据251中的车道线或者高速路边缘线的位置信息提前生成虚拟地理信息252，并且推送给车辆B,使得车辆可以控制自身运行在合理的车道范围内。需要说明的是，241、242、251、252中的车辆只是用来标识车辆行驶的位置，不属于虚拟地理信息。
图2B展示了本申请实施例在城市道路场景下的系统示意图。与图2A类似，当遇到车道线磨损或者在城市路口本身不存在车道线时，可以由云服务平台21或者路侧系统中的计算单元233等可提供计算能力的设备根据真实道路元素的位置信息生成虚拟地理信息261、271。生成的虚拟地理信息中填补了缺失的车道线信息，从而确保车辆在行驶的过程中能够全程控制自身行驶在合理的车道范围内，减少了人工干预，大大提高安全性。
结合图3介绍本申请实施例提供的车辆的车道保持控制方法的整体流程,方法可以应用于云服务平台或者路侧系统中的计算单元233等各种可提供计算能力的设备。如图所示，所述方法包括步骤S32-S36：
S32:获取真实道路元素的位置信息。真实道路元素为行车环境中影响车辆行驶的道路元素。示例性的，真实道路元素包括道路上的车道线、道路中央分隔线/中央分隔带、道路边缘线、停车线等真实存在的道路元素。示例性的,获取真实道路元素位置信息的方式有如下两种:
方式一：通过路侧感知系统231获取真实道路元素的位置信息。利用摄像机、激光雷达等传感器获取真实的车道线、道路中央分隔线/中央分隔带、真实道路边缘线、真实停车线的位置数据。示例性的，将路侧摄像机拍摄的图像输入至预训练的神经网络(例如VGG-16、Faster-RCNN等)提取各线在图像中的坐标，然后根据该摄像机的标定关系将上述各真实道路元素的图像坐标转换成世界坐标，最终可以得到真实道路元素的全局位置 信息。
方式二：从地图(可以是高精度地图)中直接提取真实道路元素的位置信息。当所述方法应用于云服务平台时，可以直接调用预存的高精地图，根据地图数据得到车道线的位置信息。当所述方法应用于路侧系统中的计算单元时，可以先搜索本地存储设备中是否存有地图数据，如果存在则直接调用，如果不存在则向管理平台请求获取地图数据。
S34：根据真实道路元素的位置信息生成虚拟地理信息。其中，虚拟地理信息包括真实道路元素对应的第一虚拟地理信息以及缺失的道路元素对应的第二虚拟地理信息。根据上述步骤S31可以得到真实道路元素的全局位置信息，进而可以确定道路元素的缺失部分(例如磨损的车道线或者路口区域本身不存在的车道线)。然后可以使用基本的数理方法或者机器学习等方法生成完整的虚拟地理信息。
S36：向车辆发送所述虚拟地理信息以使所述车辆根据所述虚拟地理信息进行车道保持。车辆驶入预设区域并且上报位置，然后根据车辆的位置发送虚拟地理信息。出于数据脱敏的考虑，可以只向车辆发送其预计行驶范围内的虚拟地理信息。示例性的，向发送以车辆位置为圆心，前方半径500米的半圆大小范围内的虚拟地理信息。若虚拟地理信息由路侧系统计算，则路侧系统可以直接向进入服务范围内的车辆推送虚拟地理信息。若虚拟地理信息由云服务平台计算，则云服务平台可以直接通过移动蜂窝数据网络或者WI-FI等无线网络向车辆推送虚拟地理信息，也可以根据各个路侧系统的位置先向不同区域的路侧系统推送对应区域的虚拟地理信息，再由路侧系统转发给车辆。总之，本申请对于计算虚拟地理信息的设备以及发送虚拟地理信息的设备不做具体限定，视实际情况而定。
在一种可能的实现方式中，向车辆推送虚拟地理信息(步骤S36)之前还包括步骤S35：
S35：校正生成的虚拟地理信息。实际的道路交通环境较为复杂，交通事故、道路施工等事件的发生会占用原本可通行的车道，此时如果仅仅使用理论上计算出来的虚拟地理信息的话，与真实的道路情况不同步，依旧可能会导致车道保持功能失效。因此，在向车辆推送虚拟地理信息之前，可以根据道路交通信息校正步骤S33生成的虚拟地理信息。其中，道路交通信息用于指示车道是否可以正常使用，示例性的，包括车流量信息、道路施工信息、道路损坏信息、交通事故信息等等。示例性的，获取道路交通信息的方式可以有如下几种：
方式一：通过路侧感知系统获取。监控摄像机采集道路上的视频数据，可以通过深度学习或者轨迹拟合等方法分析得到道路的车流量数据，或者交通事故数据。
方式二：交通管理平台推送交通事故信息或者车流量信息。
方式三：施工单位上报施工时间以及区域范围。
需要说明的是，校正虚拟地理信息的设备与计算生成虚拟地理信息的设备不一定是同一个设备。示例性的，可以是云服务平台计算生成虚拟地理信息，然后云服务平台将该信息推送给路侧系统，路侧系统根据实时道路交通信息对虚拟地理信息校正后发送给车辆。还可以是云服务平台计算生成了虚拟地理信息，然后根据路侧感知系统发送的实时道路交通信息进行校正，再发送给车辆。
步骤S32-S36生成了虚拟地理信息之后发送给车辆，车辆结合车载感知系统采集的数据以及接收到的虚拟地理信息实现车道保持。在行驶的过程中，车辆可以采用多种策略实现车道保持。示例性的，可以以车载感知系统采集的数据为主，当采集的数据置信度低于预设阈值时，再采用接收到的虚拟地理信息进行车道保持。
可选的，当车辆的车道保持系统检测到车辆偏离车道线(真实车道线或者虚拟车道线)时，车道偏离预警系统将发出警告，提醒驾驶员进行人工干预；或者，根据用户原本的设置，直接控制车辆回到车道线范围内。此外，对于运营车辆，当车道保持系统检测到车辆处于异常行驶状况并严重偏离行驶车道时，还可将异常偏离情况上报给车辆管理平台，车辆管理平台可根据车辆行驶状况和异常事件紧急程度对运营车辆采取预警措施或接管措施，避免运营车辆因异常驾驶发生驶入对向车道、坠湖等事件导致严重交通事故。
本申请实施例生成的虚拟地理信息包括了真实车道线对应的虚拟地理信息，以及缺失的车道线对应的虚拟地理信息。推送给车辆，可以使得车辆在车道线异常(车道线磨损、路口无车道线、车道线被遮挡)的情况下依然能正常实现车道保持的功能；同时，车辆将云端系统或者路侧系统等计算设备生成虚拟地理信息与车载感知系统采集的数据相结合，有效地避免了极端天气等对车载感知系统的影响，拓展了车道保持系统的适用区域和适用场景。可选的，本申请实施例还根据道路交通信息对生成的虚拟地理信息进行校正，大大提高了虚拟地理信息的准确性，减少了交通事故、道路施工等对车辆车道保持功能的影响
结合附图4和附图6，分别介绍通过路侧系统以及云服务平台生成虚拟地理信息并且推送给车辆以辅助车辆实现车道保持的具体流程步骤。需要说明的是，除了可以通过路侧系统或者云服务平台生成虚拟地理信息以外，任何可以提供计算能力的设备也同样可以获取数据生成虚拟地理信息并且推送给车辆,本发明对此不做具体限定,在本申请实施例中仅以路侧系统以及云服务平台为例。
路侧系统中的计算单元233生成虚拟地理信息辅助车辆进行车道保持
在本申请实施例中，虚拟地理信息由路侧系统中的计算单元(例如MEC)计算生成并且推送给车辆，具体包括如下步骤：
S41：车辆进入路侧系统服务区域，与路侧设备单元RSU建立通讯连接并鉴权后，上报车辆位置信息并请求服务。
需要说明的是，这里可以是车辆主动请求路侧系统推送对应位置的虚拟地理信息，也可以是路侧系统检测到车辆后，根据车辆位置主动向车辆推送对应位置的虚拟地理信息。
S42：路侧系统中的计算单元获取该车辆行驶区域中真实道路元素的位置信息。
同上面所描述的一样，真实道路元素包括真实的车道线、道路中央分隔线/中央分隔带、道路边缘线、停车线等等。正如上面所描述的那样，路侧系统中的计算单元可以有如下两种方式获取真实道路元素对应的位置信息：
方式一：路侧系统本地的存储单元存储有对应于车辆行驶区域的地图数据。地图可以是高精度地图也可以是普通的电子地图。路侧系统中计算单元可以直接从本地存储 的地图数据中提取第一虚拟地理信息(如图5A所示)。
方式二：通过路侧感知系统上传的图像数据生成该区域真实道路元素对应的位置信息。示例性的，路侧感知系统可以是摄像机，将拍摄到的画面数据输入进预训练的神经网络可以提取其中真实道路元素的特征信息，将真实道路元素的图像坐标转换成世界坐标，即可生成如图5A所示的第一虚拟地理信息。
S43：根据真实道路元素的位置信息生成虚拟地理信息。
图5A是根据真实道路元素的位置信息生成的第一虚拟地理信息。车道区域车道线磨损以及路口区域本身就不存在车道线，极易导致车辆的车道保持功能失效。因此本申请实施例生成了对应于真实道路元素以及缺失道路元素的虚拟地理信息，以确保车道保持功能的正常使用。可以使用基本的数理方法或者深度学习等方法生成虚拟地理信息。示例性的，以基本的数理方法为例，根据真实道路元素的位置信息生成虚拟地理信息包括如下步骤：
1)根据真实道路元素的位置信息生成如图5A所示的第一虚拟地理信息。
2)如图5B所示，根据第一虚拟地理信息生成各车道线以及停车线的交点或者端点。
3)如图5C所示，若道路元素的缺失部分在车道区域，则使用直线直接相连；若道路元素的缺失部分在路口区域，则使用曲线相连，曲线的曲率可以根据驾驶员的行车习惯所确定。需要说明的是，虚拟地理信息不仅包括车道线，还可以包括行车引导线(如5D所示)以及车辆可行驶区域(如5E所示)。
除了基本的数理方法之外，还可以直接将真实道路元素的位置信息输入至预训练的机器学习模型生成如图5C、5D、5E所示的虚拟地理信息。该机器学习模型可以是ResNet、GoogleNet等深度学习神经网路，也可以是支持向量机等机器学习分类模型。
S44：路侧感知系统上报道路交通数据。
在城市道路施工或者发生交通事故时，车道线被破坏或者遮挡往往会影响车道保持功能的正常使用，因此需要获取最新的道路交通信息对虚拟地理信息进行校正。道路交通数据可以是拍摄的道路图像数据也可以是雷达的探测数据。示例性的，摄像机持续采集道路车流，并且拟合生成车流轨迹，当发现原本有车流的车道上没有车流轨迹，则说明该车道异常，并且将该信息发送给路侧计算单元。
S45：路侧计算单元生成校正后的虚拟地理信息。
路侧计算单元根据路侧感知系统上传的道路交通信息对S53步骤生成的虚拟地理信息进行校正。如图6所示，车E在通过路口时原本有1、2、3条车道可以选，即原本可以生成三组虚拟车道线，但是由于车道2、3因为各种原因已经被占用，所以路侧计算单元最终只保留对应于车道1的虚拟车道线。
S46：路侧RSU将校正后的虚拟地理信息推送给车辆。
S47:车辆结合车载感知系统采集的数据以及接收到的虚拟地理信息实现车道保持。虚拟地理信息是全局数据，车道线也是全局坐标系下的坐标，所以当车辆接收到虚拟地理信息时，首先需要将虚拟地理信息转换成以车为中心的相对坐标，进而实现车道保持。
可选的，车辆在接收到虚拟地理信息之后，将该虚拟地理信息显示在用户界面上，以便于驾驶员或者乘客实时关注车辆行驶状态，改善用户体验。
需要说明的是，本申请对于车辆将车载感知数据与虚拟地理信息数据相结合的车道保持策略不做具体限定，视情况而定。示例性的，车辆系统可以预设一个阈值，当车载感知系统探测周围车道线信息时的置信度低于这个阈值时就使用虚拟地理信息提供的车道线数据，当置信度高于阈值时就采用车载感知系统探测的数据。当虚拟地理信息的置信度较高时，也可以行车全程只采用虚拟地理信息。
在上述实施例中，步骤S41车辆先上报了位置请求服务，然后路侧系统根据车辆的位置生成了包含该位置的预设大小范围内的虚拟地理信息。但是在实际的应用中，车辆上报位置的时间(步骤S41)并不限定在生成虚拟地理信息之前，也可以当路侧系统生成了自身服务区域范围内所有的虚拟地理信息之后车辆再上报位置信息请求服务。总之，本申请实施例对于车辆上报位置请求服务的时间不做具体限定。此外，对于虚拟地理信息的校正可以是不断进行的，可以是当车辆请求服务的时候再根据最新的路况数据进行校正。
本申请实施例通过生成虚拟车道线、引导线等虚拟地理信息并推送给车辆的技术方法，辅助车载感知更全面地获取道路车道信息，在无车道线区域/车道被遮挡环境下也可确保车道保持功能正常使用，减少了天气环境、车载感知系统本身的限制对车道保持功能的影响，大大扩展了应用场景，提高了行车安全。此外，路侧感知系统架设在道路高处，视角更远更好，有效地提升了车辆的感知距离和准确性。传统的车道保持功能未对车辆感知到的或从地图获取到的车道信息进行修正，本申请实施例根据实时路况修正虚拟地理信息，帮助车辆减少交通事故、道路施工等事件对车辆的车道保持系统功能的影响。
云服务平台生成虚拟地理信息辅助车辆进行车道保持
在本申请实施例中，虚拟地理信息由云服务平台计算生成并且推送给车辆，具体包括如下步骤：
S71：车辆与云服务平台进行连接，鉴权后上报位置信息，请求相关服务。
S72：云服务平台将车辆的位置信息与后台预存储的地图进行匹配，确定车辆的行驶区域。
其中，区域的大小可以是预设范围的大小。示例性的，车辆的行驶区域可以是以车为圆心，车前方半径500米的半圆。
S73：云服务平台从地图中获取该区域中真实道路元素的位置信息。
通常，云服务平台会存储有全局的地图数据，可以是普通的电子地图也可以是高精度地图。云服务平台可以直接从地图中提取车辆行驶区域的真实道路元素对应的第一虚拟地理信息。同样的，从地图中提取的虚拟地理信息仍然有缺失的部分(比如路口)，所以需要补足虚拟地理信息。示例性的，可以通过最基本的几何方法补足，路口弯道的曲率根据驾驶员的行车习惯所确定。
需要说明的是，当云服务平台无法获取地图数据时，无需执行步骤S82-S83，而是向服务范围覆盖车辆行驶区域的路侧系统发送请求以获取道路图像，进而计算出真实道路元素的位置信息(见步骤S32中的方式一)。
S74：路侧感知系统向云端服务平台上报道路交通信息。
路侧感知系统(比如摄像机或者雷达)可以根据实时探测的数据确定道路异常区域。除了通过路侧感知系统以外，公安的公共平台也可以上报事故信息或者道路施工单位直接上报的道路施工情况。总之，本申请实施例对于实时路况信息的来源不做具体限定。
S75：云服务平台根据道路交通信息校正虚拟地理信息。
同路侧系统一样，如图6所示，车E通过路口时即将直行，对面有三个车道1、2、3可以选择(暂时先不考虑对端车道方向，等过了路口车辆可以自行调整)。根据路况数据可以确定车道2、3由于事故或者施工而被占用，因此只保留了通向车道1的虚拟车道线。
S76：云服务平台向车辆推送虚拟地理信息。
云服务平台可以直接向车辆推送也可以经由路侧系统转发。
S77：车辆结合车载感知系统采集的数据以及接收到的虚拟地理信息实现车道保持。
虚拟地理信息由全局坐标系下的坐标点构成，车辆在使用虚拟地理信息之前需要首先将虚拟地理信息在全局坐标系下的坐标转换成以车辆为中心的相对坐标，便于使用。本申请实施例对于车辆基于车载感知以及虚拟地理信息的车道保持策略不做具体限定，示例性的，车辆系统预存一个阈值，车载感知系统探测车道线时会生成置信度，当置信度低于上述阈值时，则使用虚拟地理信息进行车道保持，当置信度高于上述阈值时则使用车载感知数据进行车道保持。
需要说明的是，本申请实施例对于各个步骤的顺序不做具体限定。在上述实施例中，云端系统是根据车辆上报的位置来确定区域，进而生成该区域的虚拟地理信息的。在实际的应用场景中，也可以云端系统将全局的虚拟地理信息都提前计算好存储在存储单元中，当车辆上报位置请求服务的时候，获取对应区域的道路交通数据对预存的该区域的虚拟地理信息进行校正后再推送给车辆。
本申请实施例通过云服务平台生成虚拟车道线、引导线等虚拟地理信息并推送给车辆，帮助车辆更全面地获取车道信息，在车道线异常(无车道线区域、车道线磨损、车道线被遮挡等)的场景下也可确保车道保持功能可用，扩展了应用场景；比起仅仅使用车载感知进行车道保持，本申请实施例提供的车载感知结合虚拟地理信息的车道保持策略大大提高了车辆的感知距离，提升了行车安全性；通过路况数据对生成的虚拟地理信息进行校正，减少了交通事故、道路施工等事件对车道保持功能的影响。
除了路侧系统中的计算单元以及云服务平台以外，车辆也可以作为计算设备生成虚拟地理信息，并且根据该虚拟地理信息实现车道保持功能，方法包括：车辆根据路侧感知系统采集的数据或者地图数据生成真实道路元素的位置信息，然后根据该位置信息生成虚拟地理信息，其中，虚拟地理信息包括真实道路元素对应的第一虚拟地理信息和缺失道路元素对应的第二虚拟地理信息；然后车辆根据车载感知系统采集的数据以及虚拟地理信息实现车道保持。在一种可能的实现方式中，该车辆将计算后的虚拟地理信息发送给路侧系统或者云服务系统，当有其他车辆进入该区域时，路侧系统或者云服务系统先对虚拟地理信息进行校正，然后再将校正后的虚拟地理信息发送给其他车辆。
上文结合图1-图7描述了本申请实施例提供的车辆的车道保持控制方法，下面将描述本申请实施例提供的一种车道保持控制装置800。可以包括通信模块801和处理模块 802：
处理模块802用于获取真实道路元素的位置信息，所述真实道路元素为车辆行车环境中影响所述车辆行驶的真实存在的道路元素，还用于根据所述位置信息生成虚拟地理信息，所述虚拟地理信息包括所述真实的道路元素对应的第一虚拟地理信息以及缺失的道路元素对应的第二虚拟地理信息；
通信模块801用于向所述车辆发送所述虚拟地理信息以指示所述车辆根据所述虚拟地理信息进行车道保持。
可选的，所述通信模块801还用于接收路侧感知系统发送的感知数据，所述路侧感知系统包括摄像机、毫米波雷达、激光雷达中的至少一种；所述处理模块802还用于根据所述感知数据生成所述真实道路元素的位置信息。
可选的，处理模块802还用于根据地图数据生成所述真实道路元素的位置信息，所述地图数据包括所述车辆的行驶区域的道路信息。
可选的，通信模块801还用于接收道路交通信息，所述道路交通信息用于指示道路是否可用；所述处理模块802还用于根据所述道路交通信息校正所述虚拟地理信息。
可选的，处理模块802还用于：根据所述位置信息生成所述第一虚拟地理信息，根据所述第一虚拟地理信息生成所述第二虚拟地理信息。
可选的，处理模块802还用于：将所述位置信息输入至预训练的机器学习模型生成所述虚拟地理信息。
结合图9，介绍一种具有车道保持控制功能的车辆。该车辆包括：通信模块901、处理模块902以及车载传感器903。其中通信模块901用于，接收虚拟地理信息，所述虚拟地理信息包括真实的道路元素对应的第一虚拟地理信息以及缺失的道路元素对应的第二虚拟地理信息；所述处理模块902用于根据所述虚拟地理信息以及所述车载传感器903采集的数据进行车道保持。
可选的，所述处理模块902还用于，根据所述车载感知系统采集的数据或者所述虚拟地理信息确定偏离车道；所述通信模块901还用于向用户或者车辆管理平台推送告警提示以控制所述车辆驶回所述车道范围内。
可选的，车辆还包括显示模块904，所述显示模块用于向用户显示所述虚拟地理信息。
可选的，处理模块902还用于根据所述车载感知系统采集的数据确定所述车辆前方预设范围内的道路元素在相对坐标系下的坐标，所述相对坐标系的原点为所述车辆；将所述虚拟地理信息的坐标转换成所述相对坐标系下的坐标，所述相对坐标系的原点为所述车辆；当所述车辆前方预设范围内的道路元素的坐标的置信度低于预设阈值时，根据所述虚拟地理信息的相对坐标进行车道保持。
图10为本申请实施例提供的一种车辆的车道保持控制设备1000的示意图，如图所示，所述设备1000包括处理器1001、存储介质1002、通信接口1003和内存单元1004。其中，处理器1001、存储介质1002、通信接口1003、内存单元1004通过总线1005进行通信，也可以通过无线传输等其他手段实现通信。该存储器1002用于存储指令，该处理器1001用于执行该存储器1002存储的指令。该存储器1002存储程序代码，且处理器 1001可以调用存储器1002中存储的程序代码执行以下操作：
获取真实道路元素的位置信息；
根据真实道路元素的位置信息生成虚拟地理信息；
向车辆发送所述虚拟地理信息以使车辆根据车载感知系统采集的数据和虚拟地理信息进行车道保持。
应理解，在本申请实施例中，该处理器1001可以是CPU，该处理器1001还可以是其他通用处理器、数字信号处理器(DSP)、专用集成电路(ASIC)、现场可编程门阵列(FPGA)或者其他可编程逻辑器件、分立门或者晶体管逻辑器件、分立硬件组件等。通用处理器可以是微处理器或者是任何常规的处理器等。
该存储器1002可以包括只读存储器和随机存取存储器，并向处理器1001提供指令和数据。存储器1002还可以包括非易失性随机存取存储器。例如，存储器1002还可以存储设备类型的信息。
该存储器1002可以是易失性存储器或非易失性存储器，或可包括易失性和非易失性存储器两者。其中，非易失性存储器可以是只读存储器(read-only memory,ROM)、可编程只读存储器(programmable ROM,PROM)、可擦除可编程只读存储器(erasable PROM,EPROM)、电可擦除可编程只读存储器(electrically EPROM,EEPROM)或闪存。易失性存储器可以是随机存取存储器(random access memory,RAM)，其用作外部高速缓存。通过示例性但不是限制性说明，许多形式的RAM可用，例如静态随机存取存储器(static RAM,SRAM)、动态随机存取存储器(DRAM)、同步动态随机存取存储器(synchronous DRAM,SDRAM)、双倍数据速率同步动态随机存取存储器(double data date SDRAM,DDR SDRAM)、增强型同步动态随机存取存储器(enhanced SDRAM,ESDRAM)、同步连接动态随机存取存储器(synchlink DRAM,SLDRAM)和直接内存总线随机存取存储器(direct rambus RAM,DR RAM)。
该总线1005除包括数据总线之外，还可以包括电源总线、控制总线和状态信号总线等。但是为了清楚说明起见，在图中将各种总线都标为总线105。
应理解，根据本申请实施例的设备1000可对应于本申请实施例中的管理装置800，并可以对应于执行根据本申请实施例中图4中的路侧MEC或图7中的云服务平台，为了简洁，在此不再赘述。
作为一种可能的实施例，本申请还提供一种具有车道保持控制功能的车辆，该车辆的结构可以参考图10，包括处理器、存储介质、通信接口和内存单元。其中，处理器、存储介质、通信接口、内存单元通过总线进行通信，也可以通过无线传输等其他手段实现通信。该存储器用于存储指令，该处理器用于执行该存储器存储的指令。该存储器存储程序代码，且处理器1001可以调用存储器1002中存储的程序代码执行图4或者图7所示的方法中的车辆所执行的操作步骤，为了简洁，在此不再赘述。
作为一种可能的实施例，本申请还提供一种车道保持控制系统。该系统包括如图4所示的路侧系统以及车辆。或者，该系统包括如图7所示的云服务平台和车辆。系统中的各个装置执行如各图中所示的方法，为了简洁，在此不再赘述。
上述实施例，可以全部或部分地通过软件、硬件、固件或其他任意组合来实现。当使 用软件实现时，上述实施例可以全部或部分地以计算机程序产品的形式实现。所述计算机程序产品包括一个或多个计算机指令。在计算机上加载或执行所述计算机程序指令时，全部或部分地产生按照本申请实施例所述的流程或功能。所述计算机可以为通用计算机、专用计算机、计算机网络、或者其他可编程装置。所述计算机指令可以存储在计算机可读存储介质中，或者从一个计算机可读存储介质向另一个计算机可读存储介质传输，例如，所述计算机指令可以从一个网站站点、计算机、服务器或数据中心通过有线(例如同轴电缆、光纤、数字用户线(DSL))或无线(例如红外、无线、微波等)方式向另一个网站站点、计算机、服务器或数据中心进行传输。所述计算机可读存储介质可以是计算机能够存取的任何可用介质或者是包含一个或多个可用介质集合的服务器、数据中心等数据存储设备。所述可用介质可以是磁性介质(例如，软盘、硬盘、磁带)、光介质(例如，DVD)、或者半导体介质。半导体介质可以是固态硬盘(solid state drive,SSD)。
以上所述，仅为本申请的具体实施方式。熟悉本技术领域的技术人员根据本申请提供的具体实施方式，可想到变化或替换，都应涵盖在本申请的保护范围之内。

Claims (28)

  1. 一种车辆的车道保持控制方法，其特征在于，所述方法包括：
    获取真实道路元素的位置信息，所述真实道路元素为车辆行车环境中影响所述车辆行驶的真实存在的道路元素；
    根据所述位置信息生成虚拟地理信息，所述虚拟地理信息包括所述真实的道路元素对应的第一虚拟地理信息以及缺失的道路元素对应的第二虚拟地理信息；
    向所述车辆发送所述虚拟地理信息以指示所述车辆根据所述虚拟地理信息以及车载感知系统采集的数据进行车道保持。
  2. 根据权利要求1所述的方法，其特征在于，所述虚拟地理信息包括虚拟车道线、虚拟行车引导线、车辆可行驶区域中的至少一种。
  3. 根据权利要求1或2所述的方法，其特征在于，当所述缺失的道路元素位于车道区域时，所述缺失的道路元素为被磨损的车道线；当所述缺失的道路元素位于路口区域时，所述缺失的道路元素为所述路口缺失的车道线。
  4. 根据权利要求1-3所述的方法，其特征在于，所述获取真实道路元素的位置信息包括：
    接收路侧感知系统发送的感知数据，所述路侧感知系统包括摄像机、毫米波雷达、激光雷达中的至少一种；
    根据所述感知数据生成所述真实道路元素的位置信息。
  5. 根据权利要求1-3所述的方法，其特征在于，所述获取真实道路元素的位置信息包括：
    根据地图数据生成所述真实道路元素的位置信息，所述地图数据包括所述车辆的行驶区域的道路信息。
  6. 根据权利要求1-5所述的方法，其特征在于，所述方法还包括：
    接收道路交通信息，所述道路交通信息用于指示道路是否可用；
    根据所述道路交通信息校正所述虚拟地理信息。
  7. 根据权利要求6所述的方法，其特征在于，所述道路交通信息包括：道路施工信息、交通事故信息、道路损坏信息和车流信息中的至少一种。
  8. 根据权利要求1-7所述的方法，其特征在于，所述根据所述位置信息生成虚拟地理信息包括：
    根据所述位置信息生成所述第一虚拟地理信息，
    根据所述第一虚拟地理信息生成所述第二虚拟地理信息。
  9. 根据权利要求1-7所述的方法，其特征在于，所述根据所述位置信息生成虚拟地理信息包括：
    将所述位置信息输入至预训练的机器学习模型生成所述虚拟地理信息。
  10. 根据权利要求1-9所述的方法，其特征在于，所述真实道路元素包括真实的车道线、道路中央分隔线、道路中央绿化带、真实的道路边缘线、真实的停车线中的一种或多种。
  11. 一种车辆的车道保持控制方法，其特征在于，所述方法包括：
    车辆接收虚拟地理信息，所述虚拟地理信息包括真实的道路元素对应的第一虚拟地理信息以及缺失的道路元素对应的第二虚拟地理信息；
    所述车辆根据所述虚拟地理信息以及车载感知系统采集的数据进行车道保持。
  12. 根据权利要求11所述的方法，其特征在于，所述方法还包括：
    所述车辆根据所述车载感知系统采集的数据或者所述虚拟地理信息确定偏离车道；所述车辆向用户或者车辆管理平台推送告警提示以控制所述车辆驶回所述车道范围内。
  13. 根据权利要求11或12所述的方法，其特征在于，所述方法还包括：
    所述车辆向用户显示所述虚拟地理信息。
  14. 根据权利要求11-13所述的方法，其特征在于，所述虚拟地理信息包括虚拟车道线、虚拟行车引导线、车辆可行驶区域中的至少一种。
  15. 根据权利要求11-14所述的方法，其特征在于，所述根据所述虚拟地理信息以及车载感知系统采集的数据进行车道保持包括：
    所述车辆根据所述车载感知系统采集的数据确定所述车辆前方预设范围内的道路元素在相对坐标系下的坐标，所述相对坐标系的原点为所述车辆；
    所述车辆将所述虚拟地理信息的坐标转换成所述相对坐标系下的坐标，所述相对坐标系的原点为所述车辆；
    当所述车辆前方预设范围内的道路元素在相对坐标系下的坐标的置信度低于预设阈值时，所述车辆根据所述虚拟地理信息的相对坐标进行车道保持。
  16. 一种车辆的车道保持控制装置，其特征在于，所述装置包括通信模块和处理模块，
    所述处理模块用于获取真实道路元素的位置信息，所述真实道路元素为车辆行车环境中影响所述车辆行驶的真实存在的道路元素；
    所述处理模块还用于根据所述位置信息生成虚拟地理信息，所述虚拟地理信息包括所述真实的道路元素对应的第一虚拟地理信息以及缺失的道路元素对应的第二虚拟地理信息；
    所述通信模块用于向所述车辆发送所述虚拟地理信息以指示所述车辆根据所述虚拟地理信息进行车道保持。
  17. 根据权利要求16所述的装置，其特征在于，
    所述通信模块还用于接收路侧感知系统发送的感知数据，所述路侧感知系统包括摄像机、毫米波雷达、激光雷达中的至少一种；
    所述处理模块还用于根据所述感知数据生成所述真实道路元素的位置信息。
  18. 根据权利要求16所述的装置，其特征在于，所述处理模块还用于根据地图数据生成所述真实道路元素的位置信息，所述地图数据包括所述车辆的行驶区域的道路信息。
  19. 根据权利要求16-18所述的装置，其特征在于，
    所述通信模块还用于接收道路交通信息，所述道路交通信息用于指示道路是否可用；
    所述处理模块还用于根据所述道路交通信息校正所述虚拟地理信息。
  20. 根据权利要求16-19所述的装置，其特征在于，所述处理模块还用于：
    根据所述位置信息生成所述第一虚拟地理信息，
    根据所述第一虚拟地理信息生成所述第二虚拟地理信息。
  21. 根据权利要求16-19所述的装置，其特征在于，所述处理模块还用于：
    将所述位置信息输入至预训练的机器学习模型生成所述虚拟地理信息。
  22. 一种具有车道保持控制功能的装置，所述装置放置于车辆中，其特征在于，所述装置包括通信模块、处理模块、车载传感器，
    所述通信模块用于，接收虚拟地理信息，所述虚拟地理信息包括真实的道路元素对应的第一虚拟地理信息以及缺失的道路元素对应的第二虚拟地理信息；
    所述处理模块用于根据所述虚拟地理信息以及所述车载传感器采集的数据进行车道保持。
  23. 根据权利要求22所述的装置，其特征在于，
    所述处理模块还用于，根据所述车载感知系统采集的数据或者所述虚拟地理信息确定偏离车道；
    所述通信模块还用于向用户或者车辆管理平台推送告警提示以控制所述车辆驶回所述车道范围内。
  24. 根据权利要求22或23所述的装置，其特征在于，所述车辆还包括显示模块，
    所述显示模块用于向用户显示所述虚拟地理信息。
  25. 根据权利要求22-24所述的装置，其特征在于，所述处理模块还用于：
    根据所述车载感知系统采集的数据确定所述车辆前方预设范围内的道路元素在相对坐标系下的坐标，所述相对坐标系的原点为所述车辆；
    将所述虚拟地理信息的坐标转换成所述相对坐标系下的坐标，所述相对坐标系的原点为所述车辆；
    当所述车辆前方预设范围内的道路元素的坐标的置信度低于预设阈值时，根据所述虚拟地理信息的相对坐标进行车道保持。
  26. 一种控制车辆实现车道保持的系统，其特征在于，所述系统包括如权利要求16-21任一项所述的控制装置以及如权利要求22-25任一项所述的车辆。
  27. 一种控制车辆实现车道保持的设备，其特征在于，所述设备包括：存储器和处理器，
    所述存储器用于存储程序指令，
    所述处理器用于调用存储器中的程序指令执行如权利要求1-15任一项所述方法。
  28. 一种计算机可读存储介质，其特征在于，所述可读存储介质上存储有计算机程序；
    所述计算机程序在被执行时，实现如权利要求1-15任一项所述的通信方法。
PCT/CN2021/099499 2020-07-01 2021-06-10 一种车辆的车道保持控制方法、装置及系统 WO2022001618A1 (zh)

Applications Claiming Priority (4)

Application Number Priority Date Filing Date Title
CN202010621623.6 2020-07-01
CN202010621623 2020-07-01
CN202010943544.7 2020-09-09
CN202010943544.7A CN113879298A (zh) 2020-07-01 2020-09-09 一种车辆的车道保持控制方法、装置及系统

Publications (1)

Publication Number Publication Date
WO2022001618A1 true WO2022001618A1 (zh) 2022-01-06

Family

ID=79013361

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2021/099499 WO2022001618A1 (zh) 2020-07-01 2021-06-10 一种车辆的车道保持控制方法、装置及系统

Country Status (2)

Country Link
CN (1) CN113879298A (zh)
WO (1) WO2022001618A1 (zh)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115909779A (zh) * 2022-10-31 2023-04-04 海信集团控股股份有限公司 一种车辆托底预警方法及车辆
CN116534021A (zh) * 2023-06-26 2023-08-04 小米汽车科技有限公司 虚拟车道生成方法、装置、车辆及存储介质
CN117452392A (zh) * 2023-12-26 2024-01-26 哈尔滨工业大学(深圳)(哈尔滨工业大学深圳科技创新研究院) 一种用于车载辅助驾驶系统的雷达数据处理系统和方法

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116929326A (zh) * 2022-04-07 2023-10-24 华为技术有限公司 一种数据采集方法及装置

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105551307A (zh) * 2016-02-25 2016-05-04 江苏大学 一种基于车路通信的车道识别系统及其方法
CN108509832A (zh) * 2017-02-28 2018-09-07 三星电子株式会社 用于产生虚拟车道的方法和装置
CN111192466A (zh) * 2018-11-15 2020-05-22 长城汽车股份有限公司 一种路口引导方法和装置
KR20200077485A (ko) * 2020-06-19 2020-06-30 현대모비스 주식회사 가상차선 유지 장치 및 그 방법
CN111710159A (zh) * 2020-05-29 2020-09-25 同济大学 一种基于虚拟车道线的交叉路口车辆路径规划方法和装置

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105551307A (zh) * 2016-02-25 2016-05-04 江苏大学 一种基于车路通信的车道识别系统及其方法
CN108509832A (zh) * 2017-02-28 2018-09-07 三星电子株式会社 用于产生虚拟车道的方法和装置
CN111192466A (zh) * 2018-11-15 2020-05-22 长城汽车股份有限公司 一种路口引导方法和装置
CN111710159A (zh) * 2020-05-29 2020-09-25 同济大学 一种基于虚拟车道线的交叉路口车辆路径规划方法和装置
KR20200077485A (ko) * 2020-06-19 2020-06-30 현대모비스 주식회사 가상차선 유지 장치 및 그 방법

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115909779A (zh) * 2022-10-31 2023-04-04 海信集团控股股份有限公司 一种车辆托底预警方法及车辆
CN116534021A (zh) * 2023-06-26 2023-08-04 小米汽车科技有限公司 虚拟车道生成方法、装置、车辆及存储介质
CN116534021B (zh) * 2023-06-26 2023-12-12 小米汽车科技有限公司 虚拟车道生成方法、装置、车辆及存储介质
CN117452392A (zh) * 2023-12-26 2024-01-26 哈尔滨工业大学(深圳)(哈尔滨工业大学深圳科技创新研究院) 一种用于车载辅助驾驶系统的雷达数据处理系统和方法
CN117452392B (zh) * 2023-12-26 2024-03-08 哈尔滨工业大学(深圳)(哈尔滨工业大学深圳科技创新研究院) 一种用于车载辅助驾驶系统的雷达数据处理系统和方法

Also Published As

Publication number Publication date
CN113879298A (zh) 2022-01-04

Similar Documents

Publication Publication Date Title
WO2022001618A1 (zh) 一种车辆的车道保持控制方法、装置及系统
US11958504B2 (en) Method and apparatus for controlling a vehicle's driving operation using advance information
WO2021077809A1 (zh) 一种信息融合的方法和系统
US11685360B2 (en) Planning for unknown objects by an autonomous vehicle
CN110920611B (zh) 基于邻车的车辆控制方法和装置
WO2021155685A1 (zh) 一种更新地图的方法、装置和设备
CN112033425B (zh) 车辆辅助驾驶方法、装置、计算机设备和存储介质
KR20220047732A (ko) 차량 감시 방법 및 장치, 전자 기기, 저장 매체 및 컴퓨터 프로그램, 클라우드 제어 플랫폼 및 차량 도로 협조 시스템
US10369995B2 (en) Information processing device, information processing method, control device for vehicle, and control method for vehicle
CN111524362A (zh) 基于多数据融合的车辆安全行驶保障系统与方法
KR20140112171A (ko) 위치기반 실시간 차량정보 표시시스템
EP4160346A1 (en) Method and device for identifying vehicle lane changing tendency
CN112712729B (zh) 预测运动轨迹的方法和系统
CN113997950A (zh) 车辆控制装置和车辆控制方法
WO2023092451A1 (zh) 预测可行驶车道的方法和装置
CN113748316A (zh) 用于车辆遥测的系统和方法
US20230018996A1 (en) Method, device, and computer program for providing driving guide by using vehicle position information and signal light information
US20230406357A1 (en) Autonomous driving control device and autonomous driving control method
CN113140050A (zh) 分流管控方法、装置、计算机设备和介质
DE102021132096A1 (de) Fahrzeugortsbestimmung unter verwendung kombinierter eingaben redundanter ortsbestimmungs-pipelines
CN111104824A (zh) 车道偏离的检测方法、电子设备及计算机可读存储介质
EP3896639A1 (en) Methods and systems for managing an automated driving system of a vehicle
JP2019146012A (ja) 撮像装置
US20230242145A1 (en) Mobile object control device, mobile object control method, and storage medium
CN114333390A (zh) 共享车辆停放事件的检测方法、装置及系统

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 21832860

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 21832860

Country of ref document: EP

Kind code of ref document: A1