Vehicle-road collaborative calibration system, method, electronic equipment and vehicle

Abstract

The present invention relates to a vehicle-road collaborative calibration system, method, electronic device and vehicle. The vehicle-road collaborative calibration system includes a vehicle-side sensing device, a road-side sensing device and a beacon. The vehicle-side sensing device is installed on the vehicle, and the road-side sensing device and the beacon are installed on the road side, wherein: the vehicle-side sensing device is used to collect beacon sensing information and first environmental characteristic data of the beacon, and obtain the real beacon sensing information of the beacon, so as to calibrate according to the beacon sensing information, the first environmental characteristic data and the real beacon sensing information; and / or, the road-side sensing device is used to collect vehicle sensing information and second environmental characteristic data of the vehicle, and obtain the real vehicle sensing information of the vehicle collected by the vehicle-side sensing device, so as to calibrate according to the vehicle sensing information, the second environmental characteristic data and the real vehicle sensing information. The present invention enables the vehicle-side sensing device and the road-side sensing device to meet the vehicle-road collaborative calibration requirements in a dynamic environment.

Description

Technical Field

[0001] The present invention relates to the field of autonomous driving and intelligent transportation technology, and in particular to a vehicle-road collaborative calibration system, method, electronic equipment, and vehicle. Background Art

[0002] With the rapid development of autonomous driving technology, the coordinated work of vehicle-side perception equipment and road-side perception equipment has become the key to improving the safety and reliability of autonomous driving.

[0003] In practical applications, it is necessary to calibrate the vehicle-side perception devices and the road-side perception devices. Specifically, calibration refers to the process of measuring, calculating, and adjusting the output values ​​of the vehicle-side perception devices and the road-side perception devices to ensure that the output values ​​are consistent with the true values ​​(or reference standards), thereby ensuring that the measurement results of the vehicle-side perception devices and the road-side perception devices are accurate and reliable, and then providing autonomous driving services based on accurate and reliable measurement results.

[0004] However, the calibration methods for vehicle-side and road-side sensing devices in related technologies primarily rely on static calibration, which is difficult to meet the requirements for coordinated calibration of vehicle-side and road-side sensing devices in dynamic environments. Static calibration relies on calibration parameters provided by devices in a fixed state, meaning devices that are stationary or in a stable environment. Summary of the Invention

[0005] One of the purposes of the present invention is to provide a vehicle-road collaborative calibration system to solve the problem that the calibration method of vehicle-side perception equipment and road-side perception equipment in related technologies mainly relies on static calibration, which is difficult to meet the collaborative calibration requirements of vehicle-side perception equipment and road-side perception equipment in dynamic environments; the second purpose is to provide a vehicle-road collaborative calibration method; the third purpose is to provide an electronic device; and the fourth purpose is to provide a vehicle.

[0006] In order to achieve the above object, the technical solution adopted by the present invention is as follows:

[0007] A vehicle-road cooperative calibration system includes a vehicle-side sensing device, a road-side sensing device, and a beacon. The vehicle-side sensing device is installed on the vehicle, and the road-side sensing device and the beacon are installed on the road side, wherein:

[0008] The vehicle-side sensing device is used to collect beacon sensing information of the beacon, and to obtain real beacon sensing information of the beacon, so as to perform calibration based on the beacon sensing information and the real beacon sensing information; specifically, the beacon is used to broadcast an encrypted positioning signal, and the broadcast encrypted positioning signal includes a beacon identifier, a location fingerprint and an environmental compensation parameter, the environmental compensation parameter is generated based on the multimodal environmental data collected around the beacon, and the location fingerprint is generated based on the real beacon sensing information and the multimodal environmental data. The vehicle-side sensing device is used to perform calibration based on the beacon sensing information, the location fingerprint and the environmental compensation parameter, and the beacon is also used to actively adjust the generating device for generating a prompt signal to inform the vehicle-side sensing device that a beacon exists nearby based on the multimodal environmental data, so that the vehicle-side sensing device determines that a beacon exists nearby based on the prompt signal sent by the generating device, and collects the beacon sensing information of the beacon; and / or,

[0009] The road-side perception device is used to collect vehicle perception information of the vehicle, and to obtain real vehicle perception information of the vehicle collected by the vehicle-side perception device, so as to perform calibration based on the vehicle perception information and the real vehicle perception information.

[0010] A vehicle-road collaboration calibration method involves a vehicle-side sensing device and a beacon, wherein the vehicle-side sensing device is installed on the vehicle and the beacon is installed on the road side. The method includes:

[0011] collecting beacon sensing information of the beacon;

[0012] Acquiring real beacon perception information of the beacon;

[0013] The vehicle-side perception device is calibrated according to the beacon perception information and the real beacon perception information; specifically, the beacon is used to broadcast an encrypted positioning signal, and the broadcast encrypted positioning signal includes a beacon identifier, a location fingerprint and an environmental compensation parameter. The environmental compensation parameter is generated based on the multimodal environmental data around the beacon, and the location fingerprint is generated based on the real beacon perception information and the multimodal environmental data. The vehicle-side perception device is used to calibrate according to the beacon perception information, the location fingerprint and the environmental compensation parameter. The beacon is also used to actively adjust the generating device for generating a prompt signal to inform the vehicle-side perception device that a beacon exists nearby according to the multimodal environmental data, so that the vehicle-side perception device determines that a beacon exists nearby according to the prompt signal sent by the generating device, and collects the beacon perception information of the beacon.

[0014] A vehicle-road collaboration calibration method involves a vehicle-side sensing device and a road-side sensing device, wherein the vehicle-side sensing device is installed on a vehicle and the road-side sensing device is installed on the road side. The method includes:

[0015] collecting vehicle perception information of the vehicle;

[0016] Acquiring real vehicle perception information of the vehicle collected by the vehicle-side perception device;

[0017] The road-side perception device is calibrated according to the vehicle perception information and the real vehicle perception information; specifically, the beacon is used to broadcast an encrypted positioning signal, and the broadcast encrypted positioning signal includes a beacon identifier, a location fingerprint and an environmental compensation parameter. The environmental compensation parameter is generated based on the multimodal environmental data around the beacon, and the location fingerprint is generated based on the real beacon perception information and the multimodal environmental data. The vehicle-side perception device is used to calibrate according to the beacon perception information, the location fingerprint and the environmental compensation parameter. The beacon is also used to actively adjust the generating device for generating a prompt signal to inform the vehicle-side perception device that a beacon exists nearby according to the multimodal environmental data, so that the vehicle-side perception device determines that a beacon exists nearby according to the prompt signal sent by the generating device, and collects the beacon perception information of the beacon.

[0018] An electronic device comprising: a processor; a memory for storing instructions executable by the processor;

[0019] The processor is configured to execute the instructions to implement the above-mentioned vehicle-road collaborative calibration method.

[0020] A computer-readable storage medium, when the instructions in the storage medium are executed by the processor of a mobile terminal, enables the mobile terminal to execute the above-mentioned vehicle-road cooperative calibration method.

[0021] A vehicle comprises the electronic device described above.

[0022] Beneficial effects of the present invention:

[0023] In an embodiment of the present invention, a vehicle-side perception device is used to collect beacon perception information of beacons on the road side, and obtain real beacon perception information of the beacons, so as to perform calibration based on the beacon perception information and the real beacon perception information, and / or a road-side perception device is used to collect vehicle perception information of a vehicle, and obtain real vehicle perception information of the vehicle collected by the vehicle-side perception device, so as to perform calibration based on the vehicle perception information and the real vehicle perception information. In an embodiment of the present invention, in a dynamic environment, such as when the vehicle is traveling, in a collision, or in a complex environment, the vehicle-side perception device can use the beacon perception information of the beacons on the road side and the real beacon perception information of the beacons collected in real time for calibration. In addition, the road-side perception device can also use the vehicle perception information provided by the vehicle-side perception device of the vehicle on the road and the real vehicle perception information of the vehicle collected in real time for calibration, so that the vehicle-side perception device and the road-side perception device can achieve the vehicle-road collaborative calibration requirements in a dynamic environment, and then provide accurate and reliable autonomous driving services based on the measurement results provided by the calibrated vehicle-side perception device and the road-side perception device. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 This is a schematic structural diagram of a vehicle-road cooperative calibration system provided in an embodiment of the present invention;

[0025] Figure 2 This is a flowchart of a high-precision perception calibration method based on vehicle-road collaboration provided in an embodiment of the present invention;

[0026] Figure 3 This is a flow chart of a dynamic adjustment compensation algorithm logic based on vehicle-road collaboration provided in an embodiment of the present invention;

[0027] Figure 4 This is a flow chart of a method for analyzing the association between environmental feature data and perception data provided in an embodiment of the present invention;

[0028] Figure 5 A flowchart of constructing a machine learning model for dynamic adjustment compensation provided in an embodiment of the present invention;

[0029] Figure 6 A flowchart of a feedback optimization method for dynamically adjusting a compensation model provided in an embodiment of the present invention;

[0030] Figure 7 A flowchart of the steps of a vehicle-road cooperative calibration method provided in an embodiment of the present invention;

[0031] Figure 8 This is a flowchart of another vehicle-road collaboration calibration method provided in an embodiment of the present invention;

[0032] Figure 9A schematic structural diagram of an electronic device provided by an embodiment of the present invention.

[0033] Description of reference numerals:

[0034] 100: Road-side communication unit; 101: Road-side sensing device; 102: Signal transmission line connecting beacon and road-side communication unit; 103: Distance between beacons; 104: Communication network distance between vehicle-side communication unit and road-side communication unit; 110, 111…, N: Beacons; 120 and 121: Ordinary vehicles; 122: Intelligent connected vehicle. DETAILED DESCRIPTION

[0035] The following describes the embodiments of the present invention with reference to the accompanying drawings and preferred embodiments. Those skilled in the art will readily appreciate the other advantages and benefits of the present invention from the disclosure herein. The present invention may also be implemented or applied through various other specific embodiments, and the various details in this specification may be modified or altered based on different viewpoints and applications without departing from the spirit of the present invention. It should be understood that the preferred embodiments are intended only to illustrate the present invention and are not intended to limit the scope of protection of the present invention.

[0036] With the rapid development of autonomous driving technology, the coordinated operation of vehicle-side and road-side sensing devices has become the key to improving the safety and reliability of autonomous driving. However, the following problems exist in related technologies:

[0037] Perception error problem: Vehicle-side perception devices (such as lidar and cameras) and road-side perception devices (such as radar and cameras) are easily affected by environmental interference in complex environments (such as rain, snow, fog, etc.), resulting in reduced perception accuracy.

[0038] Insufficient calibration accuracy: Existing calibration methods mainly rely on static calibration, which is difficult to meet the needs of coordinated calibration of vehicle-side and road-side equipment in dynamic environments, especially the error compensation problem in dynamic environments such as high-speed vehicle movement.

[0039] Therefore, there is an urgent need for a high-precision perception system and method that can combine vehicle-road collaborative technology to achieve dynamic calibration and error compensation of vehicle-side perception devices and / or road-side perception devices.

[0040] To address the above issues, embodiments of the present invention propose a high-precision perception calibration system and method based on vehicle-infrastructure collaboration. Through vehicle-infrastructure collaboration, this system utilizes precise roadside beacons to calibrate vehicle-side perception devices; and / or utilizes intelligent connected vehicles traveling on the road to calibrate road-side perception devices. Furthermore, by combining environmental perception, data association, machine learning, and feedback optimization technologies, and through real-time analysis of environmental characteristic data (such as traffic volume, speed, and weather), the system dynamically adjusts the parameters or models of the compensation algorithm, effectively improving the perception accuracy of the vehicle-side perception devices and / or road-side perception devices of intelligent connected vehicles, enhancing the adaptability and reliability of the perception system, and ultimately providing better autonomous driving services for autonomous driving and intelligent transportation systems.

[0041] Reference Figure 1 , showing a structural diagram of a vehicle-road cooperative calibration system provided in an embodiment of the present invention, the vehicle-road cooperative calibration system includes a vehicle-side sensing device, a road-side sensing device 101 and beacons (110, 111..., N), the vehicle-side sensing device is installed on a vehicle, for example, it can be installed on an intelligent connected vehicle 122, the intelligent connected vehicle 122 can exchange data with the road-side sensing device, wherein the road-side sensing device 101 and the beacon are installed on the road side, the road-side sensing device 101 and the beacon can be one, usually multiple.

[0042] The vehicle-side sensing device is used to collect beacon sensing information of the beacon and obtain real beacon sensing information of the beacon, so as to perform calibration based on the beacon sensing information and the real beacon sensing information; and / or,

[0043] The road-side perception device 101 is used to collect vehicle perception information of the vehicle, and to obtain real vehicle perception information of the vehicle collected by the vehicle-side perception device, so as to perform calibration based on the vehicle perception information and the real vehicle perception information.

[0044] Reference Figure 1As shown, 100 is the Roadside Unit (RSU), 101 is a roadside sensing device (such as a roadside camera or radar), 102 is the signal transmission line connecting the roadside beacon and the roadside communication unit 100; 103 is the distance between roadside beacons, which is typically 500 meters apart; 104 is the communication network distance between the On-Board Unit (OBU) and the roadside communication unit 100; the effective communication distance between the OBU and the roadside communication unit 100 is within 500 meters; 110, 111, ..., N are roadside beacons, 120 and 121 are ordinary vehicles on the road, and 122 is an intelligent connected vehicle on the road, which can be used to calibrate the roadside sensing device 101. Vehicle-side sensing devices may include sensor devices such as cameras and radars, and roadside sensing devices may include sensor devices such as cameras and radars.

[0045] It should be added that the road-side communication unit 100 can be integrated into the road-side perception device 101 or be independent of the road-side perception device 101, and the vehicle-side communication unit can be integrated into the vehicle-side perception device or be independent of the vehicle-side perception device. The embodiments of the present invention do not need to impose any restrictions on this.

[0046] In a specific implementation, the beacon perception information and the real beacon perception information may include perception information such as the location of the beacon; the vehicle perception information and the real vehicle perception information may include perception information such as the location of the vehicle.

[0047] Among them, the vehicle-side perception device calibration uses the precise beacons 110, 111,...N (high-precision GPS (Global Positioning System) equipment, radar reflectors, etc.) deployed on the road side to collect real beacon perception information, such as the precise location of the beacon. The real beacon perception information of the beacon can be sent out through V2X (vehicle to everything) communication technology. The intelligent connected vehicles 122 around the beacon, such as the intelligent connected vehicles 122 within 500m, receive the real beacon perception information sent by the road-side communication unit 100 through the vehicle-side communication unit, and the vehicle-side perception equipment of the intelligent connected vehicle 122 itself also perceives the beacon perception information of the surrounding beacons in real time. Combined with the received real beacon perception information, the perception error of the vehicle-side perception equipment is calibrated in real time by dynamically adjusting the compensation algorithm model to ensure the accuracy of the vehicle-side perception system of the vehicle.

[0048] Among them, the road-side perception device calibration is carried out by using the intelligent connected vehicle 122 traveling on the road, using its high-precision positioning GPS device and IMU (Inertial Measurement Unit) to obtain real vehicle perception information, such as the precise location of the vehicle. The intelligent connected vehicle 122 sends the real vehicle perception information to the road-side communication unit 100 through the vehicle-side communication unit. The road-side communication unit 100 then combines the vehicle perception data obtained by its own perception of the vehicle, and dynamically adjusts the compensation algorithm model to calibrate the perception error of the road-side perception device in real time to ensure the accuracy of the road-side perception system.

[0049] It should be noted that the accuracy of the real beacon perception information of road-side beacons or the real vehicle perception information of intelligent connected vehicles, that is, the real data value, is higher than the accuracy perceived by vehicle-side perception devices or road-side perception devices. The perception accuracy difference is at least about 10 times. This is mainly ensured by the following aspects:

[0050] Sources of Real Data Values: Real data values ​​(such as the location information of roadside beacons or the high-precision location information of connected vehicles) typically come from higher-precision devices or systems. First, through high-precision sensors. Connected vehicles are often equipped with high-precision sensors (such as high-precision GPS, IMUs, and LiDAR). These sensors offer far higher measurement accuracy than standard on-board sensors (such as standard cameras and millimeter-wave radars). For example, high-precision GPS can achieve centimeter-level positioning accuracy, while standard sensors may only have decimeter-level or lower accuracy. Furthermore, the location information of roadside beacons is typically pre-calibrated under static conditions using high-precision measurement equipment (such as total stations and high-precision RTK-GPS (Real-Time Kinematic Global Positioning System)). This accuracy far exceeds the perception accuracy of sensors in dynamic environments. Second, through redundancy and fusion technologies, real data values ​​are typically derived through the redundancy and data fusion of multiple high-precision sensors. For example, position information calculated by fusing multiple sensors such as GPS, IMUs, and LiDAR is significantly improved by this fusion approach.

[0051] Transmission method of real data values: The real data values ​​of vehicle-side perception devices or road-side perception devices are mainly transmitted through PC5 end-to-end communication, with a transmission delay of less than 20ms (milliseconds), ensuring low latency and high reliability performance of real data values.

[0052] Applying the calibration method of the embodiment of the present invention can improve the reliability of vehicle-side perception devices or road-side perception devices. Affected by factors such as sensor errors, environmental interference and insufficient system calibration, the reliability of vehicle-side perception devices or road-side perception devices is usually around 60%-70%. However, after calibrating the vehicle-side perception devices or road-side perception devices using the high-precision perception calibration method for vehicle-road collaboration based on the embodiment of the present invention, the reliability is significantly improved to 85%-95%.

[0053] In actual applications, during vehicle driving, collision (for example, a collision occurs with the vehicle in which the vehicle-side sensing device is located, or the road-side sensing device is hit by a vehicle or animal, etc.), or in dynamic environments such as complex environments, the vehicle-side sensing device and / or the road-side sensing device are prone to perception errors, resulting in inaccurate measurement results. The vehicle-side sensing device and / or the road-side sensing device of an embodiment of the present invention can use the perception information of the devices around it on the road for real-time calibration. In this way, the vehicle-side sensing device and / or the road-side sensing device can also be calibrated in a dynamic environment, realizing the vehicle-road collaborative calibration requirements in a dynamic environment. In this way, the vehicle-side sensing device and / or the road-side sensing device of an embodiment of the present invention can still provide accurate measurement results in a dynamic environment.

[0054] Furthermore, in actual applications, vehicle-side or road-side sensing devices may experience false triggering during autonomous driving. False triggering refers to the system incorrectly triggering a safety response (such as emergency braking or collision warning) when there is no real risk. Without calibration, the safety of the vehicle-road cooperative system is affected by the perception errors of the vehicle-side or road-side sensing devices. The false triggering probability of the vehicle-side or road-side sensing devices may be between 5% and 10%. However, the embodiments of the present invention, through calibration of the vehicle-side or road-side sensing devices, can improve the safety of vehicle-road system scenarios and reduce the false triggering of the vehicle-side or road-side sensing devices in collision-related scenarios in the vehicle-road cooperative system. In practice, the false triggering probability of the system is reduced to below 1%.

[0055] In an embodiment of the present invention, specifically, the beacon is used to broadcast an encrypted positioning signal, and the broadcast encrypted positioning signal includes a beacon identifier (beacon ID), a location fingerprint and an environmental compensation parameter. The environmental compensation parameter is generated based on the multimodal environmental data around the beacon, and the location fingerprint is generated based on the real beacon perception information and the multimodal environmental data. The vehicle-side perception device is used to calibrate according to the beacon perception information, the location fingerprint and the environmental compensation parameter. The beacon is also used to actively adjust the generating device for generating a prompt signal to inform the vehicle-side perception device that a beacon exists nearby based on the multimodal environmental data, so that the vehicle-side perception device determines the presence of a beacon nearby based on the prompt signal sent by the generating device, and collects the beacon perception information of the beacon.

[0056] It should be noted that the beacon of the embodiment of the present invention has a built-in multimodal sensor, wherein the multimodal sensor may include a millimeter-wave radar, an infrared transmitter, a radar reflector and an environmental sensor, etc. Therefore, multimodal environmental data can be collected through the beacon. The multimodal environmental data may include but is not limited to: 1) optical parameters: light intensity, rain and fog concentration; 2) electromagnetic parameters: signal interference intensity; 3) spatial parameters: distance to surrounding obstructions.

[0057] In an embodiment of the present invention, a plurality of beacons are provided on the road side, and the plurality of beacons can be combined into a signal cluster. The beacon cluster can exchange multimodal environmental data and other data through other networks such as a Mesh network to generate a dynamic error compensation model, and generate environmental compensation parameters according to the dynamic error compensation model. Dynamic compensation can be performed based on the environmental compensation parameters. For example, a Z-axis offset of 0.25m is added to the camera (vehicle-side perception device) in a dense fog environment. In this way, the camera can more accurately collect the beacon perception information of the beacon after the offset is increased, thereby avoiding the problem of calibration error caused by environmental factors.

[0058] Each beacon broadcasts locally collected multimodal environmental data every 5 seconds. The multimodal environmental data may include temperature and humidity values, optical transmittance, electromagnetic interference intensity and other data. The beacon cluster can use a distributed hash table to build an environmental situation map of the entire road network based on the multimodal environmental data collected by each beacon. Specifically, the full road network environmental situation map constructs a fully digitalized environmental model by integrating multimodal environmental data, providing real-time decision support for autonomous vehicles and roadside equipment. For example, if the full road network environmental situation map determines that visibility in a certain area is relatively low (e.g., on rainy or foggy days), the beacons in that area can actively transmit infrared light through their infrared transmitters, allowing the vehicle-side sensing equipment to determine the presence of beacons through the infrared light and then collect the beacon's sensing information for calibration. For another example, if the full road network environmental situation map determines that visibility in a certain area is relatively high (e.g., on sunny or daytime days), the beacons in that area can actively reflect the signals emitted by the vehicle-side sensing equipment through radar reflectors, allowing the vehicle-side sensing equipment to determine the presence of beacons through the reflected signals from the radar reflectors and then collect the beacon's sensing information for calibration. This shows that beacons have been upgraded from passive positioning markers to active calibration nodes.

[0059] Among them, the location fingerprint is a unique location fingerprint generated by the beacon each time it broadcasts, and the location fingerprint = Hash (beacon ID + timestamp + multimodal environment data + real beacon perception information).

[0060] In a specific example, the process of implementing calibration based on beacons in an embodiment of the present invention includes:

[0061] 1. Beacon environment perception and self-calibration (preliminary preparation)

[0062] 1.1. Beacon built-in multimodal sensor (millimeter wave radar + infrared camera + infrared transmitter + environmental sensor)

[0063] 1.2. Real-time collection of multimodal environmental data:

[0064] 1) Optical parameters: light intensity, rain and fog concentration

[0065] 2) Electromagnetic parameters: signal interference intensity

[0066] 3) Spatial parameters: distance from surrounding obstructions

[0067] 3. The beacon cluster exchanges environmental data through the Mesh network to generate a dynamic error compensation model.

[0068] 3.1 Mesh Network Collaborative Perception Network

[0069] A multi-hop communication network is automatically established by self-organizing adjacent beacons (distance ≤ 300m), using a dual-frequency redundant transmission mechanism:

[0070] Main link: 60GHz millimeter wave (10Gbps high-speed transmission, suitable for use on sunny days)

[0071] Backup link: Sub-1GHz low frequency (automatically switches in rainy and foggy weather to ensure connectivity)

[0072] Multimodal environmental data synchronization. For example, each beacon broadcasts locally collected triplet data every 5 seconds: [temperature and humidity values, optical transmittance, electromagnetic interference intensity]. The cluster constructs a full network environmental situation map using a distributed hash table, with a latency of <50ms.

[0073] 3.2 Spatial Modeling and Error Tracing

[0074] Environmental field reconstruction:

[0075] The discrete data can be converted into a continuous field model based on the Kriging spatial interpolation algorithm: the input is the measured multimodal environmental data of each beacon location, and the output is the predicted environmental value at any coordinate (x, y).

[0076] Device drift detection:

[0077] By comparing the multimodal sensor readings of multiple beacons in the same environment, if a beacon consistently deviates from the cluster median by more than 15%, it is marked as a suspected faulty device and its data weight is automatically reduced to 30%. This prevents contamination of the global model.

[0078] 3.3 Vehicle-Road-Cloud Closed-Loop Verification

[0079] The model upgrades the beacon cluster from a "passive positioning marker" to an "active error correction node" through the four-step logic of environmental field reconstruction → device status diagnosis → dynamic compensation generation → closed-loop verification.

[0080] 2. Vehicle-beacon collaborative calibration

[0081] 2.1. Beacon active wake-up

[0082] The vehicle enters the 50-meter range of the beacon → The beacon broadcasts an encrypted positioning signal through PC5. The broadcast encrypted positioning signal can contain: location fingerprint = Hash (beacon ID + timestamp + multimodal environment data + real beacon perception information).

[0083] 2.2 Real-time calibration on the vehicle side

[0084] The vehicle-side perception device obtains the beacon broadcast encrypted positioning signal including the beacon identifier, location fingerprint and environmental compensation parameters, wherein the environmental compensation parameters are generated based on the multimodal environmental data around the beacon collected by the multimodal sensor, wherein the location fingerprint is generated based on the real beacon perception information and multimodal environmental data, and the vehicle-side perception device can adjust its parameter settings for collecting beacon perception information based on the beacon perception information collected, thereby improving the measurement accuracy, and then it can be calibrated based on the collected beacon perception information and the real location fingerprint in the location fingerprint.

[0085] The actual beacon perception information of the beacon in the embodiments of the present invention can be pre-measured using static measurement methods to obtain accurate and highly precise actual beacon perception information, thereby providing an absolute benchmark for the system. Compared to traditional dynamic calibration, this avoids the problem of cumulative error. Furthermore, the beacon in the embodiments of the present invention can also use multimodal sensors to collect multimodal environmental data in real time and generate environmental compensation parameters (such as camera Z-axis offset), reducing calibration errors in dynamic environments and further improving calibration accuracy.

[0086] The embodiments of the present invention solve the problems of traditional calibration methods in calibration accuracy, calibration reliability and calibration environment adaptability through static benchmark real beacon perception information, dynamic compensation vehicle-side perception equipment, and beacon cluster data sharing. The calibration accuracy and precision of the embodiments of the present invention are higher.

[0087] In the above-mentioned vehicle-road cooperative system, the vehicle-side perception device is used to collect beacon perception information of beacons on the road side, and obtain the real beacon perception information of the beacons, so as to perform calibration based on the beacon perception information and the real beacon perception information, and / or the road-side perception device is used to collect vehicle perception information of the vehicle, and obtain the real vehicle perception information of the vehicle collected by the vehicle-side perception device, so as to perform calibration based on the vehicle perception information and the real vehicle perception information. In an embodiment of the present invention, in a dynamic environment, such as when the vehicle is driving, in a collision, or in a complex environment, the vehicle-side perception device can use the beacon perception information of the beacons on the road side collected in real time and the real beacon perception information of the beacons to perform calibration. In addition, the road-side perception device can also use the vehicle perception information provided by the vehicle-side perception device of the vehicle on the road collected in real time and the real vehicle perception information of the vehicle to perform calibration, so that the vehicle-side perception device and the road-side perception device can realize the vehicle-road cooperative calibration requirements in a dynamic environment, and then provide accurate and reliable autonomous driving services based on the measurement results provided by the calibrated vehicle-side perception device and the road-side perception device.

[0088] In one embodiment of the present invention, the vehicle-side perception device is used to collect the beacon perception information and the first environmental characteristic data of the beacon, and to obtain the real beacon perception information of the beacon, so as to perform calibration based on the beacon perception information, the first environmental characteristic data and the real beacon perception information; and / or, the road-side perception device is used to collect the vehicle perception information and the second environmental characteristic data of the vehicle, and to obtain the real vehicle perception information of the vehicle collected by the vehicle-side perception device, so as to perform calibration based on the vehicle perception information, the second environmental characteristic data and the real vehicle perception information.

[0089] Among them, environmental characteristic data may include traffic flow, speed, weather, etc.

[0090] In a specific implementation, the measurement accuracy of the vehicle-side perception device or the road-side perception device will drop to about 70%-80% in complex environments (such as rain, snow, fog and other weather conditions). To address the above problems, the vehicle-side perception device of the embodiment of the present invention can collect the beacon perception information and first environmental data of the surrounding beacons, as well as the real beacon perception data of the beacons in real time, so that it can be calibrated based on multi-source information such as beacon perception information, first environmental feature data and real beacon perception information, thereby improving the calibration accuracy of the vehicle-side perception device and further improving the measurement accuracy of the vehicle-side perception device; the road-side perception device of the embodiment of the present invention can collect the vehicle perception information and second environmental data of the surrounding vehicles, as well as the real vehicle perception data of the vehicles, so that it can be calibrated based on multi-source information such as vehicle perception information, second environmental feature data and real vehicle perception information, thereby improving the calibration accuracy of the road-side perception device and further improving the measurement accuracy of the road-side perception device.

[0091] The vehicle-side perception device or road-side perception device of the embodiment of the present invention is calibrated in combination with environmental feature data, so that the measurement accuracy of the vehicle-side perception device or road-side perception device can be improved in complex environments. The measurement accuracy of the vehicle-side perception device or road-side perception device can be improved by about 10%-20%, so that the measurement accuracy of the vehicle-side perception device or road-side perception device in complex environments can reach about 90%, thereby ensuring the measurement accuracy of the vehicle-side perception device or road-side perception device in complex environments.

[0092] In one embodiment of the present invention, the vehicle collaborative calibration system also includes a vehicle-side computing unit and a road-side computing unit: the vehicle-side computing unit is used to obtain the first perception information of the beacon collected by the vehicle-side perception device under the first sample environment characteristic data, and the first real perception information of the beacon under the first sample environment characteristic data, so as to establish a vehicle-side relationship model based on the first sample environment characteristic data, the first perception information and the first real perception information; and / or, the road-side computing unit is used to obtain the second perception information of the vehicle collected by the road-side perception device under the second sample environment characteristic data, and the second real perception information of the vehicle under the second sample environment characteristic data, so as to establish a road-side relationship model based on the second sample environment characteristic data, the second perception information and the second real perception information.

[0093] Specifically, in one embodiment of the present invention: the vehicle-side computing unit is used to divide the first sample environmental feature data, the first perception information and the first real perception information into a first training set and a first verification set, and use the first training set to train the vehicle-side relationship model to be trained, and when the vehicle-side relationship model meets the preset first convergence condition, use the first verification set to verify the vehicle-side relationship model, and when the verification passes, obtain the trained vehicle-side relationship model; and / or, the road-side computing unit is used to divide the second sample environmental feature data, the second perception information and the second real perception information into a second training set and a second verification set, and use the second training set to train the road-side relationship model to be trained, and when the road-side relationship model meets the preset second convergence condition, use the second verification set to verify the road-side relationship model, and when the verification passes, obtain the trained road-side relationship model.

[0094] The perception information (the first perception information and the second perception information) and the real perception information (the first real perception information and the second real perception information) may include perception information such as position.

[0095] In an embodiment of the present invention, a vehicle-side relationship model or a road-side relationship model can be established in advance. Based on the vehicle-side relationship model, the correlation between the perception information of the beacon collected by the road-side perception device and the actual perception information of the beacon under different environmental characteristic data can be determined; based on the road-side relationship model, the correlation between the perception information of the vehicle collected by the road-side perception device and the actual perception information of the vehicle can be determined under different environmental characteristic data.

[0096] In a specific implementation, the vehicle-road cooperative system may also include a computing unit, which may include an edge computing unit on the roadside (road-side computing unit) and a computing unit on the vehicle (vehicle-side computing unit).

[0097] Reference Figure 2 , is a flow chart of a high-precision perception calibration method based on vehicle-road collaboration provided by an embodiment of the present invention. In the vehicle-road collaboration system, since the vehicle-side perception equipment and / or the road-side perception equipment are easily affected by environmental interference in complex environments (such as rain, snow, fog, etc.), resulting in a decrease in perception accuracy, it is necessary to calibrate the vehicle-side perception equipment and / or the road-side perception equipment separately; wherein, the vehicle-side perception equipment uses the roadside beacon as the true value (real data value) for calibration, and the road-side perception equipment uses the intelligent connected vehicle on the road as the true value for calibration. The calibration process is implemented by using a dynamic adjustment compensation model. The implementation process, specifically, refers to Figure 3 , is a flowchart of the logic of a dynamic adjustment compensation algorithm based on vehicle-road collaboration provided by an embodiment of the present invention. The specific process of the dynamic adjustment compensation algorithm is: performing environmental perception and data correlation analysis, identifying key environmental features that affect the error of the perception device, and using them to build a model, specifically including: performing environmental feature data and perception information correlation analysis, analyzing the relationship between the error of environmental feature data and perception information through statistical methods or machine learning technology, and identifying features that have a significant impact on the perception error; constructing a machine learning model to predict the error changes of the road-side or vehicle-side perception device under different environmental conditions, specifically including: using the identified environmental feature data that has a significant impact on the perception error to construct a machine learning model to predict the error changes of the road-side perception device or the vehicle-side perception device under different environmental conditions; performing feedback optimization based on the true value, dynamically adjusting the parameters of the compensation model, and calibrating the deviation of the vehicle-side or road-side perception device, specifically including:, based on the real data value fed back in real time by the road-side perception device or the vehicle-side perception device, dynamically adjusting the parameters or model of the compensation algorithm through the feedback mechanism, calibrating the deviation of the vehicle-side or road-side perception device, and ensuring that the vehicle-side or road-side perception device of the perception system can maintain high precision under different environmental conditions.

[0098] Specifically, when establishing the vehicle-side relationship model or the road-side relationship model, that is, adjusting the compensation model, specifically, referring to Figure 4 , is a flow chart of a method for associating and analyzing environmental feature data with perception data, provided by an embodiment of the present invention. Environmental perception and data association analysis specifically includes data collection, data preprocessing, and data association analysis to identify features that significantly affect errors. This embodiment of the present invention uses data association technology to associate these environmental feature data with perception data or perception errors (the deviation between the perception information collected by the perception device and the actual perception information) from perception devices (vehicle-side perception devices or road-side perception devices). This primarily identifies key environmental features that influence perception device errors, thereby providing a data foundation and model basis for subsequent error compensation. The implementation logic can be as follows:

[0099] Data collection: Use multiple sensors to collect environmental characteristic data in real time as sample environmental characteristic data, such as traffic flow, speed, weather conditions, etc.

[0100] Data preprocessing: Clean and standardize the collected environmental feature data to extract useful environmental feature data.

[0101] Data association analysis to identify features that have a significant impact on errors: Using statistical methods or machine learning techniques, the relationship between environmental feature data and perception errors is analyzed to identify environmental feature data that has a significant impact on the perception errors of perception devices.

[0102] Algorithm derivation:

[0103] Assume that the error of the sensing device is , the vector of environmental characteristic data is in Indicates the i-th environmental feature data at time Through statistical analysis or machine learning methods, a relationship model between perception error and environmental feature data can be established:

[0104]

[0105] Among them, f is a function that represents the relationship between environmental feature data and perception error, is a random error term.

[0106] Through correlation analysis, regression analysis or causal inference, the feature subset that has the greatest impact on the perception error e(t) in the environmental feature data can be identified.

[0107] Reference Figure 5 , is a flowchart of a machine learning model for dynamically adjusting compensation provided by an embodiment of the present invention. The machine learning model (vehicle-side relationship model or road-side relationship model) is constructed based on historical data and real-time data to predict the error changes of road-side perception devices or vehicle-side perception devices under different environmental conditions, and implement the following logic:

[0108] Data preparation: Integrate historical data and real-time data to construct training and validation sets. For vehicle-side perception devices, historical data and real-time data may include the first perception information of the beacon collected by the vehicle-side perception device under the first sample environmental feature data, as well as the first real perception information of the beacon under the first sample environmental feature data. For road-side devices, historical data and real-time data may include the second perception information of the vehicle collected by the road-side perception device under the second sample environmental feature data, as well as the second real perception information of the vehicle under the second sample environmental feature data.

[0109] Model selection: Choose an appropriate machine learning algorithm based on the data characteristics, such as support vector regression, random forest, or neural network.

[0110] Model training: Use the training set to train the machine learning model and optimize the parameters to minimize the prediction error.

[0111] Model validation and optimization: Evaluate the performance of the trained machine learning model through the validation set, perform hyperparameter tuning and feature selection, and improve model accuracy and generalization ability.

[0112] Algorithm derivation:

[0113] Assume that the machine learning model is , by minimizing the loss function:

[0114]

[0115] represents the model parameters of the optimized machine learning model, Represents the model parameters of the machine learning model before optimization, where N is the number of training samples.

[0116] In one embodiment of the present invention:

[0117] The vehicle-side computing unit is configured to determine a first loss function gradient based on the beacon perception information, the first environmental feature data, and the real beacon perception information, and update the vehicle-side relationship model according to a first preset learning rate and the first loss function gradient; and / or,

[0118] The road-end calculation unit is used to determine the second loss function gradient based on the vehicle perception information, the second environmental feature data and the real vehicle perception information, and update the road-end relationship model according to the second preset learning rate and the second loss function gradient.

[0119] Reference Figure 6 , is a flow chart of a feedback optimization method for dynamically adjusting a compensation model provided by an embodiment of the present invention. In this embodiment of the present invention, based on the real data values ​​fed back in real time by road-side or vehicle-side sensing devices, the parameters or model of the compensation algorithm are dynamically adjusted through a feedback mechanism to ensure that the perception system maintains high accuracy under different environmental conditions. The implementation logic is as follows:

[0120] Real-time monitoring and data collection: In practical applications, the performance of the perception system is continuously monitored to collect new environmental features and perception errors. For vehicle-side perception devices, the vehicle-side computing unit can collect beacon perception information, first environmental feature data, and actual beacon perception information as new environmental features and perception errors. For road-side perception devices, the road-side computing unit can collect vehicle perception information, second environmental feature data, and actual vehicle perception information as new environmental features and perception errors.

[0121] Model update: Using new environmental features and perception errors, the parameters of the machine learning model or compensation algorithm are updated through online learning methods to adapt to environmental changes.

[0122] Performance evaluation and adjustment: Utilizing the real data values ​​of road-side or vehicle-side perception devices, that is, new environmental features and perception errors obtained through real-time monitoring and data collection, regularly evaluate the performance of compensation algorithms or models and adjust algorithm parameters to maintain high accuracy.

[0123] Algorithm derivation:

[0124] In feedback optimization, the model parameters of a machine learning model are updated using online learning methods. :

[0125]

[0126] in, is the learning rate, ∇L is the gradient of the loss function. ,The compensation algorithm adapts to environmental changes and dynamically adjusts the error compensation amount.

[0127] In one embodiment of the present invention,

[0128] The vehicle-side sensing device is configured to collect beacon sensing information and first environmental characteristic data of the beacon, and call the vehicle-side relationship model to output a first compensation value according to the first environmental characteristic data to adjust the beacon sensing information, and obtain actual beacon sensing information of the beacon, so as to perform calibration based on the adjusted beacon sensing information and the actual beacon sensing information; and / or,

[0129] The road-side perception device is used to collect the vehicle perception information and second environmental characteristic data of the vehicle, and call the road-side relationship model to output a second compensation value according to the second environmental characteristic data to adjust the vehicle perception information, and obtain the real vehicle perception information of the vehicle collected by the vehicle-side perception device to calibrate according to the adjusted vehicle perception information and the real vehicle perception information.

[0130] In an embodiment of the present invention, when the vehicle-side relationship model training is completed and passed the verification, the vehicle-side perception device can collect the beacon perception information and the first environmental feature data of the surrounding beacons, and call the vehicle-side relationship model deployed on the vehicle to output the first compensation value according to the first environmental feature data to adjust the beacon perception information, and obtain the real beacon perception information of the beacon to calibrate according to the adjusted beacon perception information and the real beacon perception information; when the road-side relationship model training is completed and passed the verification, the road-side perception device can collect the vehicle perception information and the second environmental feature data of the surrounding vehicles, and call the road-side relationship model deployed on the road to output the second compensation value according to the second environmental feature data to adjust the vehicle perception information, and obtain the real vehicle perception information of the vehicle to calibrate according to the adjusted vehicle perception information and the real vehicle perception information.

[0131] An embodiment of the present invention provides a high-precision perception calibration system based on vehicle-road collaboration. By utilizing roadside beacons and intelligent connected vehicles, combined with environmental perception, data association, machine learning, and feedback optimization technologies, the compensation algorithm is dynamically adjusted to improve the accuracy of vehicle-side perception devices or road-side perception devices as well as the system safety and reliability, providing innovative solutions for autonomous driving and intelligent transportation.

[0132] Reference Figure 7 , shows a flowchart of the steps of a vehicle-road collaboration calibration method provided in an embodiment of the present invention, involving a vehicle-side sensing device, a road-side sensing device, and a beacon. The vehicle-side sensing device is installed on the vehicle, and the road-side sensing device and the beacon are installed on the road side. Specifically, the method includes the following steps:

[0133] Step 701: Collect beacon sensing information of the beacon;

[0134] Step 702: Acquire real beacon perception information of the beacon;

[0135] Step 703: calibrate the vehicle-side perception device according to the beacon perception information and the real beacon perception information; specifically, the beacon is used to broadcast an encrypted positioning signal, and the broadcast encrypted positioning signal includes a beacon identifier, a location fingerprint and an environmental compensation parameter. The environmental compensation parameter is generated based on the multimodal environmental data collected around the beacon, and the location fingerprint is generated based on the real beacon perception information and the multimodal environmental data. The vehicle-side perception device is used to calibrate according to the beacon perception information, the location fingerprint and the environmental compensation parameter. The beacon is also used to actively adjust the generating device for generating a prompt signal to inform the vehicle-side perception device that a beacon exists nearby according to the multimodal environmental data, so that the vehicle-side perception device determines that a beacon exists nearby according to the prompt signal sent by the generating device, and collects the beacon perception information of the beacon.

[0136] In one embodiment of the present invention, step 703 includes: collecting beacon perception information and first environmental characteristic data of the beacon, and obtaining real beacon perception information of the beacon to calibrate according to the beacon perception information, the first environmental characteristic data and the real beacon perception information.

[0137] In one embodiment of the present invention, the accuracy of the beacon perception information is lower than the accuracy of the real beacon perception information; and the generating device includes at least an infrared transmitter and a radar reflector.

[0138] Acquiring vehicle perception information collected by the vehicle-side perception device of the vehicle, and sending the vehicle perception information to the road-side communication unit, so as to send the vehicle perception information to the road-side perception device through the road-side communication unit;

[0139] In one embodiment of the present invention, the method comprises:

[0140] Obtain the first perception information of the beacon collected by the vehicle-side perception device under the first sample environmental characteristic data, and the first real perception information of the beacon under the first sample environmental characteristic data, so as to establish a vehicle-side relationship model based on the first sample environmental characteristic data, the first perception information and the first real perception information.

[0141] In one embodiment of the present invention, the method comprises:

[0142] The first sample environmental feature data, the first perception information and the first real perception information are divided into a first training set and a first verification set. The first training set is used to train the vehicle-side relationship model to be trained. When the vehicle-side relationship model meets the preset first convergence condition, the first verification set is used to verify the vehicle-side relationship model. When the verification passes, a trained vehicle-side relationship model is obtained.

[0143] In one embodiment of the present invention, the method comprises:

[0144] The beacon perception information and the first environmental characteristic data of the beacon are collected, and the vehicle-side relationship model is called to output a first compensation value according to the first environmental characteristic data to adjust the beacon perception information, and the real beacon perception information of the beacon is obtained to calibrate according to the adjusted beacon perception information and the real beacon perception information.

[0145] In one embodiment of the present invention, the method comprises:

[0146] A first loss function gradient is determined based on the beacon perception information, the first environmental feature data, and the real beacon perception information, and the vehicle-side relationship model is updated according to a first preset learning rate and the first loss function gradient.

[0147] Reference Figure 8 , shows a flowchart of another vehicle-road collaboration calibration method provided in an embodiment of the present invention, involving a vehicle-side sensing device, a road-side sensing device, and a beacon. The vehicle-side sensing device is installed on the vehicle, and the road-side sensing device and the beacon are installed on the road side. Specifically, the method includes the following steps:

[0148] Step 801: Collect vehicle perception information of the vehicle;

[0149] Step 802: Acquire real vehicle perception information of the vehicle collected by the vehicle-side perception device;

[0150] Step 803: calibrate the road-side perception device according to the vehicle perception information and the real vehicle perception information; specifically, the beacon is used to broadcast an encrypted positioning signal, and the broadcast encrypted positioning signal includes a beacon identifier, a location fingerprint and an environmental compensation parameter. The environmental compensation parameter is generated based on the multimodal environmental data collected around the beacon, and the location fingerprint is generated based on the real beacon perception information and the multimodal environmental data. The vehicle-side perception device is used to calibrate according to the beacon perception information, the location fingerprint and the environmental compensation parameter. The beacon is also used to actively adjust the generating device for generating a prompt signal to inform the vehicle-side perception device that a beacon exists nearby according to the multimodal environmental data, so that the vehicle-side perception device determines that a beacon exists nearby according to the prompt signal sent by the generating device, and collects the beacon perception information of the beacon.

[0151] In one embodiment of the present invention, step 803 includes: collecting vehicle perception information and second environmental characteristic data of the vehicle, and obtaining real vehicle perception information of the vehicle collected by the vehicle-side perception device, so as to perform calibration based on the vehicle perception information, the second environmental characteristic data and the real vehicle perception information.

[0152] In one embodiment of the present invention, the accuracy of the vehicle perception information is lower than the accuracy of the real vehicle perception information; and the generating device includes at least an infrared emitter and a radar reflector.

[0153] In one embodiment of the present invention, the method includes: obtaining real beacon perception information of the beacon, and sending the real beacon perception information to the vehicle-side communication unit, so as to send the real beacon perception information to the vehicle-side perception device through the vehicle-side communication unit.

[0154] In one embodiment of the present invention, the method includes: obtaining second perception information of the vehicle collected by the road-end perception device under second sample environmental characteristic data, and second real perception information of the vehicle under the second sample environmental characteristic data, so as to establish a road-end relationship model based on the second sample environmental characteristic data, the second perception information and the second real perception information.

[0155] In one embodiment of the present invention, the method includes: dividing the second sample environmental feature data, the second perception information and the second real perception information into a second training set and a second verification set, using the second training set to train the road-end relationship model to be trained, and when the road-end relationship model meets a preset second convergence condition, using the second verification set to verify the road-end relationship model, and when the verification passes, a trained road-end relationship model is obtained.

[0156] In one embodiment of the present invention, the method includes: collecting vehicle perception information and second environmental characteristic data of the vehicle, and calling the road-side relationship model to output a second compensation value according to the second environmental characteristic data to adjust the vehicle perception information, and obtaining the real vehicle perception information of the vehicle collected by the vehicle-side perception device to calibrate according to the adjusted vehicle perception information and the real vehicle perception information.

[0157] In one embodiment of the present invention, the method includes: determining a second loss function gradient based on the vehicle perception information, the second environmental feature data and the real vehicle perception information, and updating the road-end relationship model according to a second preset learning rate and the second loss function gradient.

[0158] It should be noted that for the sake of simplicity, the method embodiments are described as a series of actions. However, those skilled in the art should be aware that the embodiments of the present invention are not limited by the order of the actions described, because according to the embodiments of the present invention, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in this specification are all preferred embodiments, and the actions involved are not necessarily required for the embodiments of the present invention.

[0159] As for the method embodiment, since it is basically similar to the system embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the system embodiment.

[0160] The embodiment of the present invention further provides an electronic device, such as Figure 9 As shown, it includes a processor 1001, a device interface 1002, a memory 1003 and a bus 1004;

[0161] Memory 1003, used for storing computer programs;

[0162] The processor 1001 is configured to implement the above steps when executing the program stored in the memory 1003 .

[0163] The bus mentioned in the terminal above can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus. This bus can be divided into address buses, data buses, control buses, etc. For ease of illustration, the figure shows only one thick line, but this does not mean that there is only one bus or only one type of bus.

[0164] The memory may include random access memory (RAM) or non-volatile memory, such as at least one disk storage. Alternatively, the memory may be at least one storage device located away from the processor.

[0165] The above-mentioned processor can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.

[0166] The present invention also provides a storage medium, which, when the instructions in the storage medium are executed by a processor of an electronic device, enables the electronic device to execute the vehicle-road cooperative calibration method of the aforementioned embodiment.

[0167] As for the device embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0168] The algorithm and display provided herein are not inherently related to any particular computer, virtual device or other equipment. According to the above description, it is obvious that the structure required for constructing this type of device is suitable. In addition, the present invention is not directed to any specific programming language. It should be understood that various programming languages ​​can be utilized to realize the content of the present invention described herein, and the above description of specific languages ​​is for the purpose of disclosing the best mode of the present invention.

[0169] In the description provided herein, numerous specific details are described. However, it is understood that embodiments of the present invention may be practiced without these specific details. In some instances, well-known methods, structures, and techniques are not shown in detail so as not to obscure the understanding of this description.

[0170] Similarly, it should be understood that in order to streamline the present invention and aid in understanding one or more of the various inventive aspects, in the above description of exemplary embodiments of the invention, various features of the invention are sometimes grouped together into a single embodiment, figure, or description thereof. However, this disclosed method should not be interpreted as reflecting an intention that the claimed invention requires more features than are expressly recited in each claim. Rather, as reflected in the claims below, inventive aspects lie in less than all the features of the individual embodiments disclosed above. Accordingly, the claims following the detailed description are hereby expressly incorporated into this detailed description, with each claim standing on its own as a separate embodiment of the invention.

[0171] Those skilled in the art will appreciate that the modules in the devices of the embodiments can be adaptively modified and installed in one or more devices different from the embodiments. The modules, units, or components in the embodiments can be combined into a single module, unit, or component, and furthermore, they can be divided into multiple sub-modules, sub-units, or sub-components. All features disclosed in this specification (including the accompanying claims, abstract, and drawings), and all processes or units of any method or device disclosed therein, can be combined in any combination, unless at least some of such features and / or processes or units are mutually exclusive. Unless expressly stated otherwise, each feature disclosed in this specification (including the accompanying claims, abstract, and drawings) may be replaced by an alternative feature that serves the same, equivalent, or similar purpose.

[0172] The various component embodiments of the present invention may be implemented in hardware, or as software modules running on one or more processors, or a combination thereof. Those skilled in the art will appreciate that in practice, a microprocessor or digital signal processor (DSP) may be used to implement some or all of the functions of some or all of the components of the sorting device according to the present invention. The present invention may also be implemented as an apparatus or device program for performing part or all of the methods described herein. Such a program implementing the present invention may be stored on a computer-readable medium or in the form of one or more signals. Such signals may be downloaded from an Internet website, provided on a carrier signal, or provided in any other form.

[0173] It should be noted that the above embodiments illustrate rather than limit the invention, and that those skilled in the art may devise alternative embodiments without departing from the scope of the appended claims. In the claims, any reference signs placed between brackets should not be construed as limiting the claims. The word "comprising" does not exclude the presence of elements or steps not listed in the claims. The word "a" or "an" preceding an element does not exclude the presence of a plurality of such elements. The present invention may be implemented by means of hardware comprising several different elements and by means of appropriately programmed computers. In a unit claim enumerating several means, several of these means may be embodied by the same item of hardware. The use of the words first, second, and third etc. does not indicate any order. These words may be interpreted as names.

[0174] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described devices, apparatuses and units can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0175] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

[0176] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any modifications or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be based on the scope of protection of the claims.

[0177] It should be noted that the various data-related processes in the embodiments of the present application are all carried out in compliance with the corresponding data protection laws and policies of the country where they are located, and with the authorization given by the owner of the corresponding device.

Claims

1. A vehicle-road cooperative calibration system, characterized in that: The vehicle-road cooperative calibration system includes a vehicle-side sensing device, a road-side sensing device, and a beacon. The vehicle-side sensing device is installed on the vehicle, and the road-side sensing device and the beacon are installed on the road side. In a dynamic environment: The vehicle-side sensing device is used to collect beacon sensing information and first environmental feature data of the beacon, and to obtain real beacon sensing information of the beacon, so as to perform calibration based on the beacon sensing information, the first environmental feature data and the real beacon sensing information; specifically, the beacon is used to broadcast an encrypted positioning signal, and the broadcast encrypted positioning signal includes a beacon identifier, a location fingerprint and an environmental compensation parameter, the environmental compensation parameter is generated based on the multimodal environmental data collected around the beacon, and the location fingerprint is generated based on the real beacon sensing information and the multimodal environmental data. The vehicle-side sensing device is used to perform calibration based on the beacon sensing information, the location fingerprint and the environmental compensation parameter. The beacon is also used to actively adjust the generating device for generating a prompt signal to inform the vehicle-side sensing device that a beacon exists nearby based on the multimodal environmental data, so that the vehicle-side sensing device determines that a beacon exists nearby based on the prompt signal sent by the generating device, and collects the beacon sensing information of the beacon; The road-side perception device is used to collect the vehicle perception information and the second environmental characteristic data of the vehicle, and to obtain the real vehicle perception information of the vehicle collected by the vehicle-side perception device, so as to perform calibration based on the vehicle perception information, the second environmental characteristic data and the real vehicle perception information.

2. The vehicle-road cooperative calibration system according to claim 1, characterized in that: The accuracy of the beacon perception information is lower than the accuracy of the real beacon perception information; the accuracy of the vehicle perception information is lower than the accuracy of the real vehicle perception information; the generating device includes at least an infrared transmitter and a radar reflector.

3. The vehicle-road cooperative calibration system according to claim 1, characterized in that: The vehicle-road cooperative calibration system further includes a vehicle-side communication unit and a road-side communication unit: The vehicle-side communication unit is configured to obtain vehicle perception information collected by the vehicle-side perception device of the vehicle and send the vehicle perception information to the road-side communication unit, so as to send the vehicle perception information to the road-side perception device via the road-side communication unit; and / or, The road-side communication unit is used to obtain the real beacon perception information of the beacon and send the real beacon perception information to the vehicle-side communication unit, so as to send the real beacon perception information to the vehicle-side perception device through the vehicle-side communication unit.

4. The vehicle-road cooperative calibration system according to claim 1, characterized in that: The vehicle collaborative calibration system also includes a vehicle-side computing unit and a road-side computing unit: The vehicle-side computing unit is configured to obtain first perception information of the beacon collected by the vehicle-side perception device under first sample environmental characteristic data, and first real perception information of the beacon under the first sample environmental characteristic data, so as to establish a vehicle-side relationship model based on the first sample environmental characteristic data, the first perception information, and the first real perception information; and / or, The road-end computing unit is used to obtain the second perception information of the vehicle collected by the road-end perception device under the second sample environmental characteristic data, as well as the second real perception information of the vehicle under the second sample environmental characteristic data, so as to establish a road-end relationship model based on the second sample environmental characteristic data, the second perception information and the second real perception information.

5. The vehicle-road cooperative calibration system according to claim 4, characterized in that: The vehicle-side computing unit is configured to divide the first sample environmental feature data, the first perception information, and the first real perception information into a first training set and a first validation set, use the first training set to train the vehicle-side relationship model to be trained, and when the vehicle-side relationship model satisfies a preset first convergence condition, use the first validation set to validate the vehicle-side relationship model, and obtain a trained vehicle-side relationship model when the validation passes; and / or, The road-end calculation unit is used to divide the second sample environmental feature data, the second perception information and the second real perception information into a second training set and a second verification set, use the second training set to train the road-end relationship model to be trained, and when the road-end relationship model meets the preset second convergence condition, use the second verification set to verify the road-end relationship model. When the verification passes, a trained road-end relationship model is obtained.

6. The vehicle-road cooperative calibration system according to claim 5, characterized in that: The vehicle-side sensing device is configured to collect beacon sensing information and first environmental characteristic data of the beacon, and call the vehicle-side relationship model to output a first compensation value according to the first environmental characteristic data to adjust the beacon sensing information, and obtain actual beacon sensing information of the beacon, so as to perform calibration based on the adjusted beacon sensing information and the actual beacon sensing information; and / or, The road-side perception device is used to collect the vehicle perception information and second environmental characteristic data of the vehicle, and call the road-side relationship model to output a second compensation value according to the second environmental characteristic data to adjust the vehicle perception information, and obtain the real vehicle perception information of the vehicle collected by the vehicle-side perception device to calibrate according to the adjusted vehicle perception information and the real vehicle perception information.

7. The vehicle-road cooperative calibration system according to claim 4, characterized in that: The vehicle-side computing unit is configured to determine a first loss function gradient based on the beacon perception information, the first environmental feature data, and the real beacon perception information, and update the vehicle-side relationship model according to a first preset learning rate and the first loss function gradient; and / or, The road-end calculation unit is used to determine the second loss function gradient based on the vehicle perception information, the second environmental feature data and the real vehicle perception information, and update the road-end relationship model according to the second preset learning rate and the second loss function gradient.

8. A vehicle-road collaboration calibration method, characterized in that: The invention relates to a vehicle-side sensing device and a beacon, wherein the vehicle-side sensing device is installed on the vehicle and the beacon is installed on the road side. In a dynamic environment, the method includes: The vehicle-side sensing device is used to collect beacon sensing information and first environmental characteristic data of the beacon; Acquiring real beacon perception information of the beacon; The vehicle-side perception device is calibrated according to the beacon perception information, the first environmental feature data and the real beacon perception information; wherein, the beacon is used to broadcast an encrypted positioning signal, and the broadcast encrypted positioning signal includes a beacon identifier, a location fingerprint and an environmental compensation parameter. The environmental compensation parameter is generated based on the multimodal environmental data around the beacon, and the location fingerprint is generated based on the real beacon perception information and the multimodal environmental data. The vehicle-side perception device is used to calibrate according to the beacon perception information, the location fingerprint and the environmental compensation parameter. The beacon is also used to actively adjust the generating device for generating a prompt signal to inform the vehicle-side perception device that a beacon exists nearby according to the multimodal environmental data, so that the vehicle-side perception device determines that a beacon exists nearby according to the prompt signal sent by the generating device, and collects the beacon perception information of the beacon.

9. A vehicle-road collaboration calibration method, characterized in that: The method involves a beacon, a vehicle-side sensing device, and a road-side sensing device, wherein the vehicle-side sensing device is installed on a vehicle, the beacon is installed on the road side, and the road-side sensing device is installed on the road side. In a dynamic environment, the method includes: The roadside sensing device is used to collect vehicle sensing information of the vehicle and second environmental feature data; Acquiring real vehicle perception information of the vehicle collected by the vehicle-side perception device; The road-side perception device is calibrated according to the vehicle perception information, the second environmental feature data and the real vehicle perception information; the vehicle-side perception device is used to collect the beacon perception information and the first environmental feature data of the beacon; obtain the real beacon perception information of the beacon; and calibrate the vehicle-side perception device according to the beacon perception information, the first environmental feature data and the real beacon perception information. Specifically, the beacon is used to broadcast an encrypted positioning signal, and the broadcast encrypted positioning signal includes a beacon identifier, a location fingerprint and an environmental compensation parameter. The environmental compensation parameter is generated based on the multimodal environmental data collected around the beacon, and the location fingerprint is generated based on the real beacon perception information and the multimodal environmental data. The vehicle-side perception device is used to calibrate according to the beacon perception information, the location fingerprint and the environmental compensation parameter. The beacon is also used to actively adjust the generating device for generating a prompt signal to inform the vehicle-side perception device that a beacon exists nearby based on the multimodal environmental data, so that the vehicle-side perception device determines the presence of a beacon nearby based on the prompt signal sent by the generating device and collects the beacon perception information of the beacon.

10. An electronic device, characterized in that: include: processor; a memory for storing processor-executable instructions; The processor is configured to execute the instructions to implement the vehicle-road cooperative calibration method as described in any one of claim 8 or claim 9.

11. A computer-readable storage medium, characterized in that When the instructions in the storage medium are executed by the processor of the mobile terminal, the mobile terminal is enabled to execute the vehicle-road cooperative calibration method according to any one of claims 8 or 9.

12. A vehicle, characterized in that: The vehicle includes the electronic device according to claim 10.

Images