Early warning control method for traffic information in roadway based on perception fusion technology

Abstract

The invention relates to an early warning control method for traffic information in a roadway based on a perception fusion technology, belongs to the technical field of traffic information in the roadway, and solves the problem that in the prior art, traffic information is difficult to effectively obtain in an underground roadway in a mine scene, and accidents are frequently caused. Identifying a sub-scene in a well mining scene where the vehicle is located; s2, identifying traffic lights and traffic signboards; step S3, vehicle-road perception fusion: obtaining a vehicle-road perception fusion result list and sending the vehicle-road perception fusion result list to a decision module of the vehicle-mounted electronic unit; s4, dangerous area situation assessment: performing dangerous area situation assessment according to the received vehicle and road perception fusion result list to obtain a dangerous area situation assessment result; and S5, a decision module of the vehicle-mounted electronic unit performs decision processing and early warning control according to the obtained dangerous area situation assessment result to obtain a decision result or an early warning result.

Description

Technical Field

[0001] The present invention relates to the technical field of in-lane traffic information technology, and in particular to an in-lane traffic information early warning control method based on perception fusion technology. Background Art

[0002] At present, safe driving in underground mining scenarios in coal mine operations is the first rigid demand of underground vehicle users. In the process of driving in underground mining scenarios, rear-end collisions are the main factors causing vehicle accidents. Especially in underground tunnels, due to the environment and low visibility in the tunnels, most vehicle accidents are caused by rear-end collisions, and often multiple vehicles collide in a chain, causing serious casualties and losses. The cause is often too close to the car or the driver cannot grasp more observation information. Especially in underground tunnels in underground mining scenarios, the above accidents are very likely to occur.

[0003] When operating vehicles in underground mining scenarios are driving, it is common for them to encounter situations such as when passing through intersections, where a large vehicle in front blocks the traffic lights and traffic signs in the underground mining scenario, where visibility underground is low and the traffic lights and traffic signs in front cannot be identified, or where the traffic lights and traffic signs in front are blurred and confused, and where the traffic lights and traffic sign information (TSP) in front cannot be accurately obtained due to limited perception capabilities. The above situations will seriously hinder the driver's ability to identify the road environment ahead and may cause traffic collisions.

[0004] Therefore, the driving environment in the underground tunnels of underground mines poses a severe challenge to intelligent driving technology. Although some solutions have been proposed in this field, such as the single-vehicle TSR (Traffic Sign Recognition System) information broadcast function that obtains traffic information ahead in advance. However, this method cannot timely integrate the information of the intersection in the real underground mining scene with the data on the vehicle side, resulting in the inability to obtain accurate and effective status information, and even obtaining erroneous information, which makes the TSR function invalid.

[0005] There is a need in this field for an intelligent driving method that can obtain accurate environmental perception and make decisions in underground mining scenarios. Summary of the invention

[0006] In order to solve the above problems, the embodiments of the present invention provide a method for early warning and control of traffic information in lanes based on perception fusion technology. By combining vehicle-mounted sensing and V2X (vehicle-to-external information exchange) technology (V2V vehicle-to-vehicle information exchange or V2I vehicle-to-road information exchange), the perception precision and accuracy are improved, the missed judgment rate and the misjudgment rate are reduced, and better results are achieved compared to the single-vehicle TSR function. By integrating vehicle-mounted sensing and V2X technology, more types of information can be obtained, effectively solving problems such as blurry and easily confused traffic lights and traffic signs in underground lanes, thereby improving driving safety. Compared with traditional methods, V2X perception technology can provide more accurate and comprehensive intersection information, allowing drivers to make better decisions and reduce the possibility of misjudgment in complex environments.

[0007] According to an embodiment of the present invention, a method for early warning control of traffic information in a lane based on perception fusion technology is provided, comprising the following steps:

[0008] Step S1, scene recognition, identifying the sub-scenes in the underground mining scene where the vehicle is located through the vehicle-side information obtained in real time by the vehicle-mounted sensor, and judging one or more of the following situations: scene 1 that the traffic light at the intersection in the underground tunnel is blocked, scene 2 that the traffic light at the intersection in the underground tunnel is blurred or easily confused, and scene 3 that the traffic sign in the underground tunnel is blurred or easily confused. The vehicle-side information includes the vehicle position detected by the vehicle-mounted sensor, the vehicle speed, the traffic light status of the road ahead, the traffic light position, the traffic sign position, and the traffic sign information;

[0009] Step S2, traffic light and traffic sign recognition, obtaining the traffic light status and traffic sign information of the vehicle's current lane as roadside information through the roadside unit, and sending the roadside information as roadside target information to the vehicle's onboard electronic unit after roadside sensing fusion processing;

[0010] Step S3, vehicle-road perception fusion, the on-board electronic unit synchronizes the roadside target information received from the roadside unit with the vehicle-side information obtained in real time by the on-board sensor in time and space, and performs fusion processing to obtain vehicle-road matching target information, unmatched vehicle-side target information and unmatched roadside target information, which are sent to the decision module of the on-board electronic unit as a vehicle-road perception fusion result list;

[0011] Step S4, dangerous area situation assessment, in step S1, when it is determined that the vehicle is in one or more of the scenarios 1, 2 and 3, the decision module of the on-board electronic unit performs dangerous area situation assessment according to the received vehicle-road perception fusion result list to obtain the dangerous area situation assessment result;

[0012] Step S5: The decision module of the vehicle-mounted electronic unit performs decision processing and early warning control according to the obtained dangerous area situation assessment result to obtain a decision result or an early warning result.

[0013] Optionally, step S3 further comprises:

[0014] Step S3.1, obtaining the input of vehicle-road perception fusion, including the roadside target information received by the vehicle-mounted electronic unit from the roadside unit, and the vehicle-side target information, wherein the vehicle-side target information is the vehicle-side information acquired by the vehicle-mounted sensor in real time and outputted after the vehicle-side perception fusion processing is performed by the vehicle-mounted electronic unit;

[0015] Step S3.2, a time-space synchronization step, in which the onboard electronic unit performs trend extrapolation prediction on the received roadside target information, compensates for the delay of the roadside target information relative to the vehicle-side target information, and obtains the roadside target information after time-space synchronization;

[0016] Step S3.3, a fusion processing step, sequentially performing GNN association matching on the roadside target information and the vehicle-side information after time-space synchronization, to obtain successfully matched vehicle-road matching target information, as well as unmatched vehicle-side target information and unmatched roadside target information;

[0017] Step S3.4, performing Kalman filter fusion and GNN multi-target matching tracking on the vehicle-road matching target information to obtain fused vehicle-road matching target information, and sending the fused vehicle-road matching target information together with the unmatched vehicle-side target information and the unmatched roadside target information as a vehicle-road perception fusion result list to the decision module of the on-board electronic unit;

[0018] Among them, the list of vehicle-road perception fusion results includes the vehicle position, the vehicle initial speed v SV , the traffic light status of the road ahead, the location of the intersection where the traffic light is located, the location of the traffic sign, the traffic sign information, the road condition information ahead, and one or more of the remaining time of the green light countdown.

[0019] Optionally, the dangerous area situation assessment in step S4 includes traffic light status assessment, traffic sign assessment, and dangerous area judgment, and further includes the following steps:

[0020] Step S4.1, when it is determined according to step S1 that the vehicle is in one or more of the scenarios 1, 2 and 3, the decision module of the onboard electronic unit evaluates the traffic light status according to the received vehicle-road perception fusion result list, including: when the vehicle-road perception fusion result list shows that the traffic light status of the road ahead is red, yellow or green, the vehicle position and the initial speed v of the vehicle are obtained. SV, traffic light status, intersection location of the traffic light, and remaining time of the green light countdown, and use this information as the result of the dangerous area situation assessment;

[0021] Step S4.2, the decision module of the on-board electronic unit performs a traffic sign evaluation according to the received vehicle-road perception fusion result list, including: when the vehicle-road perception fusion result list shows that the traffic sign information ahead includes speed limit sign information, obtain the traffic sign position, traffic sign information, maximum speed limit v lim information, and use this information as the result of the dangerous area situation assessment;

[0022] Step S4.3, the decision module of the on-board electronic unit makes a dangerous area judgment based on the received vehicle-road perception fusion result list, including: when the road condition information ahead in the vehicle-road perception fusion result list shows the existence of a construction area, obtains the safety line position and the boundary information of the dangerous area, and uses this information as the dangerous area situation assessment result.

[0023] Optionally, the step S5 further comprises the following steps:

[0024] Step S5.1, in the obtained dangerous area situation assessment result, when the traffic light status is displayed as red light or yellow light, decision processing and early warning control are performed to obtain a decision result or an early warning result;

[0025] Step S5.2, in the obtained dangerous area situation assessment result, the traffic light status is displayed as a green light countdown, and decision processing and early warning control are performed to obtain a decision result or an early warning result;

[0026] Step S5.3: when the traffic sign information in the obtained dangerous area situation assessment result includes a speed limit sign, decision processing and early warning control are performed to obtain a decision result or an early warning result;

[0027] Step S5.4: When there is a construction area in the obtained dangerous area situation assessment results, decision processing and early warning control are performed to obtain a decision result or an early warning result.

[0028] Optionally, the step S5.1 specifically includes the following steps:

[0029] In the obtained dangerous area situation assessment results, when the traffic light status is red or yellow, the distance traveled by the vehicle when it is gently decelerated to a stop at the preset gentle acceleration a1 is calculated as

[0030]

[0031] Among them, v SV is the initial speed of the vehicle, t svd is the preset driver reaction time, t rbris the preset braking system response time, d pre To preset a warning safety distance;

[0032] Compare the distance S0 that the vehicle travels when gently decelerating to a stop with the distance from the vehicle's current position to the stop line corresponding to the traffic light: when S0 is less than the distance from the vehicle's current position to the stop line, gentle deceleration is performed as a decision result; when S0 is greater than the distance from the vehicle's current position to the stop line, the driver is prompted to take over as a warning result.

[0033] Optionally, the step S5.2 specifically includes the following steps:

[0034] In the obtained dangerous area situation assessment results, the traffic light status is displayed as a green light countdown, which further determines whether the vehicle can pass through the intersection;

[0035] The vehicle has an initial speed v SV The time to pass the intersection with traffic lights is

[0036]

[0037] Among them, d SV is the distance from the vehicle to the intersection where the traffic light is located, v SV is the initial speed of the vehicle, t sv Count down the remaining time for the green light;

[0038] Compare the time TTC required for the vehicle to reach the intersection at the current speed and driving direction with the remaining time t of the green light countdown sv , if TTC≤t sv , the vehicle passes normally as the decision result; if TTC>t sv , the driver is prompted to take over as a warning result.

[0039] Optionally, the step S5.3 specifically includes the following steps:

[0040] In the obtained dangerous area situation assessment results, when the traffic sign information includes a speed limit sign, compare the vehicle initial speed v SV Maximum speed limit with traffic signs lim , if the vehicle's initial speed is less than the maximum speed limit v on the traffic sign lim Then normal driving is taken as the decision result, otherwise the following steps are performed;

[0041] When the vehicle's initial speed is greater than the maximum speed limit v on the traffic sign lim In the case of a vehicle with a moderate acceleration, the distance required to decelerate gently to the maximum speed limit of the traffic sign is calculated as follows:

[0042]

[0043] The distance S required for the vehicle to decelerate from the initial speed to the maximum speed limit with a gentle acceleration is compared. lim , and the distance from the vehicle’s current position to the traffic sign. When S lim When S is less than the distance from the current position of the vehicle to the traffic sign, a gentle deceleration is performed as the decision result; when S lim When it is greater than the distance from the vehicle's current position to the traffic sign, it indicates that it is not possible to pass with gentle deceleration, and the driver is prompted to take over as a warning result.

[0044] Optionally, the step S5.4 specifically includes the following steps:

[0045] In the obtained dangerous area situation assessment results, when there is a construction area, the location of the construction area is obtained as the safety line location, and the distance traveled by the vehicle when it is gently decelerated to a stop at the preset gentle acceleration a1 is calculated as

[0046]

[0047] Compare the distance S0 that the vehicle travels when gently decelerating to a stop with the distance from the vehicle's current position to the safety line: when S0 is less than the distance from the vehicle's current position to the safety line, gentle deceleration is performed as a decision result; when S0 is greater than the distance from the vehicle's current position to the safety line, the driver is prompted to take over as a warning result.

[0048] The embodiments of the present invention provide a method for early warning and control of traffic information in a lane based on perception fusion technology. Through the perception fusion of vehicle-mounted sensors and V2X technology, it brings higher accuracy and reliability to the intelligent driving system in the underground mining scene, making driving in the underground lanes safer and more efficient. The embodiments of the present invention provide a method for early warning and control of traffic information in a lane based on perception fusion technology. By utilizing vehicle-road collaboration and using V2X technology, information interaction and sharing are achieved between vehicles and road infrastructure, and single-vehicle perception and vehicle-road collaboration are combined, effectively improving the intelligent driving system solution and technical path, enhancing the environmental perception ability of automobile intelligent driving, greatly enhancing the perception ability of automobiles, and ultimately improving the functions, performance and safety and reliability of automobile intelligent driving. In addition, the widespread application of vehicle-road collaboration can also reduce the cost of single-vehicle intelligent perception and improve safety in underground lanes of underground mines.

[0049] An embodiment of the present invention provides a method for early warning and control of traffic information in a tunnel based on perception fusion technology. By closely combining single-vehicle perception and vehicle-road collaboration technology in underground mining scenarios, accurate perception of the environment and better control decisions can be achieved. Even if the traffic lights and / or traffic signs in the underground mining scenario are blurred or confused, or there are fault lights on the road, accurate decisions and warnings can be made in real time, bringing a more efficient and safer intelligent driving experience to underground mines, and facilitating the widespread application of intelligent driving technology in this scenario. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. The features and advantages of the present invention can be more clearly understood by referring to the drawings. The drawings are schematic and should not be understood as limiting the present invention in any way. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0051] Figure 1 This is a schematic diagram of a TSR+ operation vehicle driving through an intersection and a large truck in front blocking the traffic light.

[0052] Figure 2 This is a schematic diagram of a scenario where there is a traffic light ahead of the TSR+ operation vehicle.

[0053] Figure 3 This is a schematic diagram of a scenario where there is a traffic sign in front of the TSR+ operation vehicle.

[0054] Figure 4 A schematic diagram of the longitudinal coverage area of ​​a scene blocked and / or blurred and / or confused by a traffic light in an embodiment of an in-lane traffic information warning control method based on perception fusion technology according to an embodiment of the present invention.

[0055] Figure 5 A schematic diagram of the longitudinal coverage area of ​​traffic sign confusion in an embodiment of a method for early warning control of traffic information in a lane based on perception fusion technology according to an embodiment of the present invention.

[0056] Figure 6 The present invention is a flowchart of vehicle-road perception fusion processing of a method for early warning control of traffic information in a lane based on perception fusion technology according to an implementation manner of the present invention.

[0057] Figure 7 The present invention is a flowchart of decision-making processing and warning control of a method for warning control of traffic information in a lane based on perception fusion technology according to an implementation manner of the present invention. DETAILED DESCRIPTION

[0058] In order to more clearly understand the above-mentioned purpose, features and advantages of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be noted that the embodiments of the present invention and the features in the embodiments can be combined with each other without conflict.

[0059] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein. Therefore, the protection scope of the present invention is not limited to the specific embodiments disclosed below.

[0060] Glossary of relevant technical terms

[0061] TSP: Traffic lights and traffic sign information ahead;

[0062] V2V: Vehicle to Vehicle (V2V);

[0063] V2I: Vehicle to Infrastructure (V2I);

[0064] V2P: Vehicle to Pedestrian (V2P) communication;

[0065] V2N: Vehicle to Network (V2N);

[0066] TSR: Traffic Sign Recognition System;

[0067] TSR+: Intelligent connected traffic sign recognition system;

[0068] OBU: On-board electronic unit;

[0069] RSU: Roadside Unit.

[0070] The present invention realizes the integration of vehicle-mounted perception and roadside perception technology in the tunnel (including with other vehicles) in the underground mining scene through the application of V2X and TSR+ technology, and solves the problems that cannot be solved by vehicle-mounted perception on the basis of obtaining more reliable and accurate environmental perception information. Specifically, after solving the TSR information obtained in front of the underground tunnel road in the underground mining scene, the TSR information obtained in front is integrated through single-vehicle perception and V2V technology, and the TSR information in front is perceived or predicted as early as possible, thereby increasing the available reaction time of the vehicle, and more effectively realizing the decision-making of the vehicle, thereby realizing the functions, performance and reliability that cannot be obtained by traditional TSR information.

[0071] A method for early warning control of traffic information in a lane based on perception fusion technology provided according to an embodiment of the present invention is described in detail below with reference to the accompanying drawings.

[0072] According to an embodiment of the present invention, a method for early warning control of traffic information in a lane based on perception fusion technology includes the following steps.

[0073] Step S1, scene recognition, through the vehicle-side information obtained in real time by the vehicle-mounted sensor, the sub-scene in the underground mining scene where the vehicle is located is identified, and one or more of the following situations are judged: scene 1 that the traffic light at the intersection in the underground tunnel is blocked, scene 2 that the traffic light at the intersection in the underground tunnel is blurred or easily confused, and scene 3 that the traffic sign in the underground tunnel is blurred or easily confused. The vehicle-side information may include the following detected by the vehicle-mounted sensor: vehicle position, vehicle speed, traffic light status on the road ahead, traffic light position, traffic sign position, and traffic sign information light. In addition, vehicle-side information can also include: vehicle posture information: such as the pitch angle and roll angle of the vehicle body, to help understand the dynamic situation of the vehicle in the lane; environmental perception information: such as data from vehicle-mounted sensors such as self-mounted cameras, lidar, radar, etc., used to detect obstacles and pedestrians in the surrounding environment; road condition information: such as road friction coefficient, slope, etc., which can be obtained through the vehicle's sensors; vehicle power information: such as the current engine speed, torque, throttle and brake status, etc., which can be used to judge the dynamic state of the vehicle; weather and lighting information: if the vehicle is equipped with corresponding sensors, it can obtain environmental information such as the current light intensity, temperature, humidity, etc.; vehicle communication information: communication status information with other vehicles or infrastructure, such as V2V (vehicle-to-vehicle communication) or V2I (vehicle-to-infrastructure communication) signal quality, delay, etc. These additional vehicle-side information can further enhance the vehicle's perception of underground mining scenes and help identify complex scenes more accurately.

[0074] Obtain vehicle and front traffic information including traffic lights and traffic signs in front, and predict the road traffic information environment ahead. The embodiment of the present invention provides the following three sub-scenario TSR information acquisition of TSR+operating vehicles as a corresponding control decision technology for predicting.

[0075] Scenario 1: TSR is applied in underground mining scenarios where traffic lights at intersections in underground tunnels are blocked. Since the traffic lights are blocked, the traffic light status information cannot be obtained, and the TSR function fails. TSR+ is applied in scenarios where traffic lights at intersections are blocked. Through the integration of single-vehicle perception and V2X perception, compared with the single-vehicle TSR function, the monitoring distance and area are increased, and the changing status of the traffic lights ahead can be perceived in advance; the perception precision and accuracy are improved, and the missed judgment rate and misjudgment rate are reduced; more types of information perceived by V2X can effectively solve the problem of traffic lights being blocked, and improve driving safety and traffic efficiency.

[0076] Scenario 2: TSR is applied in underground mining scenarios where traffic lights at intersections in underground tunnels are blurred or easily confused. Due to blurred or easily confused traffic lights, traffic light status information cannot be obtained, or the wrong traffic light status is obtained, making the TSR function invalid. The TSR+ function is applied to scenarios where traffic lights are blurred or easily confused. Through the fusion of single-vehicle perception and V2X perception, it improves the perception precision and accuracy compared to the single-vehicle TSR function, and reduces the missed judgment rate and false judgment rate; through V2X perception of more types of information, it effectively solves problems such as blurred and easily confused traffic lights and improves driving safety.

[0077] Scenario 3: TSR is applied in underground mining scenarios where the signs in the underground tunnels are blurred or easily confused. Due to the blurred or easily confused traffic signs, the traffic sign information cannot be obtained, or the wrong traffic sign information is obtained, and the TSR function fails. The TSR+ function is applied to the scene where the traffic signs are blurred or easily confused. Through the fusion of single-vehicle perception and V2X perception, compared with the single-vehicle TSR function, it improves the perception precision and accuracy, reduces the missed judgment rate and the false judgment rate; through V2X perception of more types of information, it effectively solves the problems of blurred and easily confused traffic lights, and improves driving safety.

[0078] like Figure 1 As shown in the figure, TSR is applied to the scene where the traffic light at the intersection in the tunnel of the underground mining scene is blocked. Since the traffic light is blocked, the traffic light status information cannot be obtained, and the TSR function fails. TSR+ is applied to the scene where the traffic light at the intersection is blocked. Through the fusion of single-vehicle perception and V2X perception, compared with the single-vehicle TSR function, the monitoring distance and area are increased, and the change status of the traffic light ahead can be perceived in advance; the perception precision and accuracy are improved, and the missed judgment rate and misjudgment rate are reduced; through more types of information perceived by V2X, the problem of traffic light obstruction is effectively solved, and driving safety and traffic efficiency are improved.

[0079] like Figure 2As shown in the figure, TSR is applied in underground mining scenarios where traffic lights at intersections in underground tunnels are blurred or easily confused. Due to blurred or easily confused traffic lights, traffic light status information cannot be obtained, or the wrong traffic light status is obtained, and the TSR function fails. The TSR+ function is applied in scenarios where traffic lights are blurred or easily confused. Through the fusion of single-vehicle perception and V2X perception, it improves the perception precision and accuracy compared to the single-vehicle TSR function, and reduces the missed judgment rate and false judgment rate; through V2X perception of more types of information, it effectively solves problems such as blurred and easily confused traffic lights, and improves driving safety.

[0080] like Figure 3 As shown in the figure, TSR is applied in underground mining scenarios where traffic signs in underground tunnels are blurred or easily confused. Due to the blurred or easily confused traffic signs, the traffic sign information cannot be obtained, or the wrong traffic sign information is obtained, and the TSR function fails. The TSR+ function is applied to scenarios where traffic signs are blurred or easily confused. Through the fusion of single-vehicle perception and V2X perception, compared with the single-vehicle TSR function, it improves the perception precision and accuracy, reduces the missed judgment rate and the false judgment rate; through V2X perception of more types of information, it effectively solves the problems of blurred and easily confused traffic lights, and improves driving safety.

[0081] The following is an explanation of environmental perception and conditions.

[0082] Braking response time: Article 7.14.1.2 of GB7258 stipulates that "the time from pressing the brake pedal to the most unfavorable brake chamber response should be less than or equal to 0.6s". In order to ensure maximum safety, the braking system response time is set to 0.6s in this system.

[0083] In the implementation mode of the present invention, the working vehicle in the underground mining scene passes through the roadside unit RSU of the lane, and through V2X communication, realizes the communication with the roadside sensor equipment equipped in front through the roadside unit RSU, and collaboratively perceives the TSR information in front, and then fuses it with the vehicle-mounted perception information of the vehicle. On the one hand, the vehicle perceives the traffic information in front through its own perception system, and at the same time, obtains the behavior information of the vehicle on the road ahead in time through the front TSR obtained by V2V. If the traffic information in front is blurred or blocked and the scene occurs, after the vehicle senses the roadside unit information through the on-board perception device, the system can obtain the traffic information status in front in advance, perceive and judge the environment in advance, make decisions in advance, and take necessary measures in advance, such as alarm, active deceleration or emergency braking. The ultimate goal is to make the vehicle drive safely and obtain TSR information in advance. To achieve the above system technical goals, the following technical contents are mainly solved: real-time communication of V2V, perception fusion based on V2X, prediction and behavior judgment after obtaining TSR information, and vehicle control decision module. The present invention mainly includes an early warning control method based on a perception fusion system.

[0084] Step S2, based on the scene recognition result of step S1, obtain information about the traffic lights and traffic signs in the identified scene. The vehicle-side information is obtained through the on-board sensor and transmitted to the on-board electronic unit of the vehicle, and the traffic lights and traffic signs in the current lane of the vehicle are obtained as roadside information through the roadside unit RSU, and the roadside information is fused and processed and sent to the on-board electronic unit OBU of the vehicle as roadside target information RSI. Among them, the vehicle obtains the traffic lights and traffic signs in the current lane through the roadside unit RSU, including the location, display information, etc. of the traffic lights and traffic signs. The on-board electronic unit OBU (On-Board Unit) may include on-board sensors, processing modules, decision modules, and multiple functional modules for execution control.

[0085] Step S3: Vehicle-road perception fusion, the RSI received in step S2 is used as the roadside target information to synchronize with the vehicle-side target information in time and space, and then fused to obtain vehicle-road matching target information, unmatched vehicle-side target information, and unmatched roadside target information, which are sent to the decision module as a list of vehicle-road perception fusion results. The vehicle-road matching target information is the roadside target information RSI and vehicle-side target information that are matched after time and space synchronization and fusion processing; the unmatched vehicle-side target information is the vehicle-side target information that fails to match after fusion processing; and the unmatched roadside target information is the roadside target information RSI that fails to match after fusion processing. This step is described in detail below.

[0086] refer to Figure 6 , showing the structural block diagram of the vehicle-road perception fusion system of the vehicle-side electronic unit.

[0087] Step S3.1, obtaining the input of vehicle-road perception fusion, including the vehicle-side information acquired by the vehicle-mounted sensor in real time, which is output as the vehicle-side target information after the vehicle-mounted electronic unit performs perception fusion processing, and the road-side information initially acquired by the road-side unit is transmitted to the vehicle-mounted electronic unit after the fusion processing by the road-side unit as the RSI (i.e., road-side target information). The vehicle-side information acquired by the vehicle-mounted sensor in real time includes the status of traffic lights, traffic signs, vehicle speed, vehicle acceleration, vehicle position, etc. The vehicle-side electronic unit performs vehicle-side perception fusion processing on the vehicle-side information to obtain the vehicle-side perception fusion result including: vehicle-side timestamp, target ID, target position, target speed, target acceleration, etc. The road-side unit performs road-side perception fusion processing on the road-side information to obtain the road-side perception fusion result including: road-side timestamp, target ID, target position, target speed, target acceleration, etc.

[0088] In the vehicle-side target information and roadside target information after perception fusion processing respectively, the initially obtained vehicle-side information and roadside information have been processed by the on-board electronic unit and the roadside unit respectively into data such as absolute longitude and latitude coordinates and speed in the world coordinate system, that is, spatial alignment has been completed.

[0089] Step S3.2, time-space synchronization, the time-space synchronization step, performs trend extrapolation prediction on the received roadside target information RSI, compensates for the delay of the roadside target information RSI relative to the vehicle-side target information, and obtains the roadside target information RSI after time-space synchronization. For the roadside target information transmitted to the on-board electronic unit in the above step S3.1, after the on-board electronic unit unlocks the roadside target information, the vehicle-side target information and roadside target information that are time-space synchronized are obtained by compensating for the impact of the roadside target information delay. As mentioned above, the vehicle-side target information and the roadside target information have been unified into absolute longitude and latitude coordinates, and no additional spatial synchronization is required. The following explains the synchronization and compensation of time. In terms of time, in theory, the two parts can be synchronized only after the two parts are synchronized using Lagrangian system interpolation, and then the two parts of the perception can be fused. However, in practical applications, due to the time consumption of multiple processes such as calculation and signal transmission, the roadside target information of the lane will inevitably have a certain delay relative to the vehicle-side target information. The reasons for the delay of roadside target information relative to vehicle-side target information include: signal transmission delay (due to distance, obstacles, interference, etc.), data processing delay (decoding, synchronization, fusion process), network delay (network instability), multipath propagation and interference (complex environment leads to increased complexity of signal reception) and system processing load (processing speed slows down under high load). For the perception fusion results of the on-board electronic unit and the roadside unit, due to the different transmission topologies, and considering the many links involved in the image processing, perception fusion, RSU-OBU data communication and transparent data analysis of the lane roadside camera, a large time delay will inevitably occur (for example, but not limited to, up to 0.2s). For the lane roadside target information that always has a certain lag, the influence of the roadside information delay can be compensated by testing and statistics of the time delay, according to the specific time delay length, using trend extrapolation prediction, and obtaining the spatiotemporally synchronized vehicle-side target information and roadside target information, and then the subsequent vehicle-road perception fusion processing can be carried out. The length of the time delay is tested and counted through real-time measured data. It is necessary to obtain delay data through testing and statistics according to the specific length of the time delay, and then use the trend extrapolation prediction method to compensate for the impact of roadside information delay. Therefore, the processing of the delay is based on real-time measurement and statistical results.

[0090] Step S3.3, perform GNN association matching on the vehicle-side target information and roadside target information RSI after time and space synchronization, and obtain successfully matched vehicle-road matching target information, unmatched vehicle-side target information, and unmatched roadside target information. Unmatched vehicle-side target information may include: vehicle-mounted target ID, timestamp, vehicle-mounted target speed, vehicle-mounted target position, vehicle-mounted target acceleration, etc. Unmatched roadside target information may include: roadside target ID, timestamp, roadside target speed, roadside target position, roadside target acceleration, etc.

[0091] Step S3.4, continue to perform Kalman filter fusion processing and GNN multi-target matching tracking processing on the successfully matched vehicle-road matching target information to obtain the fused vehicle-road matching target information. The fused vehicle-road matching target information and the unmatched vehicle-side target information and the unmatched roadside target information are sent to the decision module of the on-board electronic unit as a vehicle-road perception fusion result list. In other words, after the above fusion processing, the vehicle-road matching target information, the unmatched vehicle-side target information and the unmatched roadside target information can be obtained, and all are retained. The three parts of information are superimposed as a vehicle-road perception fusion result list, and sent to the decision module in the on-board electronic unit for target behavior prediction and decision-making. The fused vehicle-road matching target information may include: fused ID, timestamp, fused target speed, fused target position, fused target acceleration, etc. The vehicle-road perception fusion result list includes vehicle position, vehicle initial speed v SV , the traffic light status of the road ahead, the location of the intersection where the traffic light is located, the location of the traffic sign, the traffic sign information, the road condition information ahead, the remaining time of the green light countdown, the road condition information ahead, and information of other related targets.

[0092] The fusion system in the on-board electronic unit is consistent with the vehicle-side and road-side perception, and uses three steps of GNN association matching, Kalman filter fusion and GNN multi-target matching tracking to fuse the spatiotemporally synchronized vehicle-side target information and road-side target information. The accurate vehicle-road matching target information, unmatched vehicle-side target information and unmatched road-side target information obtained will be sent to the decision module in the on-board electronic unit as part of the vehicle-road perception fusion result list for subsequent control and decision-making. This fusion processing step solves the problem in the prior art that the information of the intersection and the roadside in the real underground mining scene cannot be timely integrated with the data on the vehicle side, resulting in the inability to obtain accurate and effective status information, and even erroneous information may be obtained, thereby rendering the TSR function invalid.

[0093] In addition, the list of vehicle-road perception fusion results provides basic data for the following dangerous area situation assessment. These data include information such as traffic light status, distance between vehicles and traffic facilities, etc., which are used for analysis of dangerous areas and calculation of warning distances.

[0094] Step S4 is a dangerous area situation assessment step. When step S1 determines that the vehicle is in one or more of the scenarios 1, 2 and 3, the decision module of the onboard electronic unit performs a dangerous area situation assessment based on the received vehicle-road perception fusion result list to obtain a dangerous area situation assessment result. The dangerous area situation assessment may include: traffic light status assessment, traffic sign assessment, and dangerous area judgment. Step S4 specifically includes the following steps.

[0095] Step S4.1: When it is determined in step S1 that the vehicle is in one or more of the scenarios 1, 2 and 3, the decision module of the onboard electronic unit evaluates the traffic light status according to the received vehicle-road perception fusion result list: when the vehicle-road perception fusion result list shows that the traffic light status of the road ahead is red, yellow or green, the vehicle position and initial vehicle speed v are obtained. SV , traffic light status, intersection location of traffic light, remaining time of green light countdown, etc., and use this information as the result of dangerous area situation assessment, otherwise the vehicle passes normally. In the dangerous area situation assessment result, the safety line position is the intersection location of the traffic light. When the vehicle-road perception fusion result list shows that the traffic light status of the road ahead is green, continue to judge whether the green light is in the countdown state based on the V2X information. If the green light has started counting down, it will be used as the result of dangerous area situation assessment.

[0096] Step S4.2, the decision module of the on-board electronic unit evaluates the traffic sign according to the received vehicle-road perception fusion result list: when the vehicle-road perception fusion result list shows that the traffic sign information ahead includes speed limit information, obtain the traffic sign location, traffic sign content and other information, and use this information as the dangerous area situation assessment result, otherwise the vehicle passes normally. The traffic sign content may include the maximum speed limit v lim In the dangerous area situation assessment result, the safety line position is the traffic sign position.

[0097] Step S4.3, the decision module of the on-board electronic unit makes a dangerous area judgment based on the received vehicle-road perception fusion result list: when the road condition information ahead in the vehicle-road perception fusion result list shows that there is a construction area or other dangerous area, obtain the safety line position, the boundary of the dangerous area and other information, and use this information as the dangerous area situation assessment result, otherwise the vehicle passes normally. In the dangerous area situation assessment result, the safety line position is the construction area position.

[0098] According to the information in the vehicle-road perception fusion result list, if the traffic light status is red, yellow, green, or cannot be passed before the green light ends, the on-board electronic unit will remind the driver to brake to ensure that the vehicle stops before the safety line; if the traffic sign is a speed limit sign, the vehicle should slow down to the speed limit when passing the traffic sign; if it is a dangerous area, the driver should be reminded and pass at the speed limit (for example, 5km / h); if it is a construction area, the driver should be reminded and pass according to the actual situation. For both dangerous areas and construction areas, when the on-board electronic unit determines that an early warning is needed, the most stringent vehicle action should be performed to remind the driver.

[0099] The following describes the perception process of the perception system of the vehicle-mounted electronic unit.

[0100] Scene 1: Traffic lights at intersections in underground tunnels are blocked and / or Scene 2: Traffic lights at intersections in underground tunnels are blurred or easily confused (the above Figure 1 and Figure 2 In the scenario described above, the ROI (region of interest) analysis considers the situation where the intersection lights cannot be directly observed, that is, the TSR+ intersection traffic lights are blocked and / or the traffic lights are blurred. In this scenario, the analysis of the ROI area considers two aspects, one is the traffic light status information, and the other is the warning distance. The status information of the traffic light can be sent to the on-board electronic unit OBU through the roadside unit RSU. At this stage, the roadside communication coverage range, the range within which the roadside unit can communicate, can reach 200 meters to 300 meters. The consideration of the warning distance is mainly to ensure that the vehicle passes through the traffic light intersection in compliance with traffic regulations, and to issue driving warning information when there is a driving risk.

[0101] The following describes the steps for assessing the situation in a dangerous area.

[0102] Scenario judgment: For scenarios 1, 2, and 3 of the TSR+ scenario, when the vehicle obtains the traffic light or traffic sign information at the intersection ahead from the road attributes in the RSI information sent by the roadside unit RSU received from the OBU, the scenario flag can be activated, and the scenario module can be activated at the same time to determine whether it is in one or more of the above scenarios 1, 2, and 3, and activate the scenario flag. The scenario manager schedules the scenario to make relevant decisions.

[0103] Situation assessment and decision-making: Figure 7 As shown, the scene traffic decision mainly includes two parts. One part is whether the vehicle can safely pass the traffic light intersection according to the current speed under the green light state, and the other part is to gently brake and stop in front of the safety line under the red or yellow light state.

[0104] By analyzing the RSI (Roadside Real-time Traffic Information) information received from the OBU and sent by the Roadside Unit RSU, the information and location of the traffic light or traffic sign ahead are obtained. Corresponding reminders are given in different ranges for different traffic lights and traffic sign information to ensure that the driver takes over the vehicle in time and passes safely.

[0105] To facilitate the explanation of the situation assessment and decision logic of the second scene of intersection occlusion, the relevant parameters can be defined as follows: The initial speed of the vehicle is v SV , the traffic sign shows that the maximum speed limit is v lim ; The speed limit area can be set to, for example, 30 km / h; Set the driver's reaction time to tsvd ; Set the vehicle braking system reaction lag time to t rbr ; The preset safety distance constant for early warning is d pre (For example but not limited to, 25m, the specific value can be adjusted and optimized as needed).

[0106] The roadside unit RSU monitors the status of traffic lights in real time and publishes traffic light information and traffic sign information (including, for example, the content and location of traffic lights and traffic signs). It determines whether the vehicle's current driving status can safely pass the traffic light or traffic sign position, and then sends reminder information at the appropriate location to achieve the purpose of early warning.

[0107] Step S5, the decision module of the on-board electronic unit performs decision processing and early warning control according to the obtained dangerous area situation assessment results, obtains decision results or early warning results, and issues corresponding early warning reminders to the driver of the vehicle. According to the boundary of the dangerous area obtained from the dangerous area situation assessment results, combined with information such as the safety line position, vehicle position, vehicle speed, traffic light status, traffic sign position, and traffic sign, the time for the vehicle to pass the safety line and / or traffic sign is calculated, so as to perform normal driving, driving warning and other operations on the vehicle. The triggering principle of driving warning is that when the vehicle enters the communication range of the roadside unit and is still a certain distance away from the traffic light or traffic sign, it can ensure that after the warning is issued, the driver can make the vehicle safely pass the traffic sign or traffic light through appropriate operations.

[0108] Step S5.1: When the traffic light status is displayed as red or yellow in the obtained dangerous area situation assessment result, decision processing and early warning control are performed to obtain a decision result or an early warning result, and the driver is reminded. This step specifically includes the following steps.

[0109] The vehicle has an initial speed v SV When the system detects that the traffic light ahead is red or yellow, it automatically controls the vehicle to decelerate at a preset gentle acceleration to determine whether it can stop before the stop line. The distance traveled by the vehicle during this gentle deceleration to stop is:

[0110]

[0111] Among them, v SV is the initial speed of the vehicle, t svd is the preset driver reaction time, t rbr is the preset braking system response time, a1 is the set mild braking acceleration (vector value, positive when accelerating and negative when decelerating); d pre is the preset warning safety distance. Since the vehicle decelerates from the initial speed to a stop (speed is 0km / h), the formula for Simplified.

[0112] The distance S0 traveled by the vehicle when it is gently decelerated to a stop is compared with the distance from the current position of the vehicle to the stop line. When S0 is less than the distance from the current position of the vehicle to the stop line, gentle deceleration is performed as a decision result; when S0 is greater than the distance from the current position of the vehicle to the stop line, the driver is prompted to take over as a warning result. When the driver receives the warning result, he can brake in time.

[0113] In one example, for example, take v SV =32km / h, t svd =0.5s, t rbr =0.6s, a1=-0.2g,d pre = 25m, and the distance traveled by the vehicle from the initial speed to the stop is s0 = 67m. The above is only an example, and the value can be set according to the specific application requirements.

[0114] like Figure 4 The vehicle shown enters the communication range of the lane roadside unit (the dotted and double-dash line box range in the figure), and then makes a comprehensive judgment on the vehicle's motion state and traffic light information, and issues a reminder to the driver after the vehicle enters the warning distance (the dotted line box range in the figure). After receiving the system prompt, the driver takes the initiative to take braking measures to achieve deceleration measures.

[0115] Step S5.2: In the obtained dangerous area situation assessment result, the traffic light status is displayed as a green light countdown for decision processing and early warning control, a decision result or early warning result is obtained, and the driver is reminded. This step specifically includes the following steps.

[0116] The traffic light status is displayed as a green light countdown to further determine whether the vehicle can pass through the intersection; if it is no longer possible to pass, the driver needs to be prompted to take over; otherwise, the vehicle passes normally.

[0117] When the vehicle cannot pass the intersection before the green light ends, a judgment should be made based on the vehicle's driving information. This judgment is consistent with the above processing process for the red or yellow light state, so ROI analysis is also included when the traffic light shows green. The vehicle is at the current initial speed v SV The time to pass through the intersection with traffic lights is as follows:

[0118]

[0119] Among them, d SV is the distance from the vehicle to the intersection where the traffic light is located, v SV is the initial speed of the vehicle, t svCount down the remaining time for the green light. Here, TTC is the time it takes the vehicle to reach the intersection at its current speed and direction of travel.

[0120] Compare the time TTC required for the vehicle to reach the intersection at the current speed and driving direction with the remaining time t of the green light countdown sv , if TTC≤t sv , the vehicle passes normally as the decision result; if TTC>t sv , the driver is prompted to take over as a warning result. The driver can directly control the vehicle to brake or slow down, and the control method is similar to the red or yellow light state. See the steps above for reminding the driver when the traffic light shows red or yellow light state. Figure 4 Execution of early warning and control: Based on the above evaluation results, the decision module will send corresponding control signals to remind the driver or directly control the vehicle (such as braking or deceleration).

[0121] Step S5.3: in the obtained dangerous area situation assessment result, if the traffic sign information includes a speed limit sign, decision processing and early warning control are performed to obtain a decision result or an early warning result, and the driver is reminded. This step specifically includes the following steps.

[0122] Compare the vehicle initial speed v SV Maximum speed limit with traffic signs lim , and then issue corresponding instructions to the vehicle. If there is no additional speed limit sign in the area, it will pass through according to the set safety speed limit (for example, 30km / h). If the initial speed of the vehicle is greater than the maximum speed limit v lim It is then determined whether it is possible to decelerate to the maximum speed limit with gentle acceleration, otherwise normal driving is the decision result.

[0123] If the vehicle's initial speed is greater than the maximum speed limit v on the traffic sign lim The calculation formula for the distance required for a vehicle to decelerate gently to the maximum speed limit indicated by the traffic sign with gentle acceleration is as follows:

[0124]

[0125] Among them, S lim v is the distance required for the vehicle to decelerate from its initial speed to the maximum speed limit with gentle acceleration; lim This is the maximum speed limit indicated on the traffic sign.

[0126] The distance S required for the vehicle to decelerate from the initial speed to the maximum speed limit with a gentle acceleration is compared. lim , and the distance from the vehicle’s current position to the traffic sign. When S limWhen S is less than the distance from the current position of the vehicle to the traffic sign, a gentle deceleration is performed as the decision result; when S lim When the distance is greater than the distance from the vehicle's current position to the traffic sign, it means that it is impossible to pass through with gentle deceleration, and the driver is prompted to take over as a warning result. When the driver receives the warning result, he can perform emergency deceleration.

[0127] In one example, for example, v SV =32km / h, t svd =0.5s, t rbr =0.6s, a1=0.2g, v lim =30km / h, according to the above formula, S lim =59m. That is, at this distance, the driver can take the initiative to brake after receiving the system prompt, and achieve safe passage by slowing down to the specified speed. The above is only an example, and the value can be set according to specific application requirements.

[0128] Step S5.4, when there is a construction area in the obtained dangerous area situation assessment result, decision processing and early warning control are performed to obtain a decision result or an early warning result, and the driver is reminded. The step S5.4 specifically includes the following steps.

[0129] In the obtained dangerous area situation assessment results, when there is a construction area, the location of the construction area is obtained as the safety line location, and the distance traveled by the vehicle when it is gently decelerated to a stop at the preset gentle acceleration a1 is calculated as

[0130]

[0131] Compare the distance S0 traveled by the vehicle when it is gently decelerated to a stop with the distance from the current position of the vehicle to the safety line: when S0 is less than the distance from the current position of the vehicle to the safety line, perform gentle deceleration as a decision result; when S0 is greater than the distance from the current position of the vehicle to the safety line, prompt the driver to take over as a warning result. Optionally, in the dangerous area situation assessment result, when there are other dangerous conditions, such as obstacles, damaged roads, etc., this step can also be performed similarly.

[0132] In short, the triggering condition of the warning is that the vehicle enters the communication range of the roadside unit (i.e. Figure 7 ) when the vehicle is within the perception area shown in the figure: the current vehicle speed cannot safely pass the traffic light and / or traffic sign; or the distance between the vehicle and the traffic light and / or traffic sign is less than the calculated distance above.

[0133] like Figure 5The vehicle shown enters the communication range of the roadside unit (the dotted and double-dash line box range in the figure), and then the vehicle's motion state and traffic sign information are comprehensively judged, and the driver is reminded after the vehicle enters the warning distance (the dotted line box range in the figure). After receiving the system prompt, the driver takes the initiative to take braking measures to achieve the vehicle deceleration to the speed limit and pass safely.

[0134] For the processing of the TSR+traffic sign confusion scenario (scenario 3 above), it can also include firstly measuring and locating the traffic sign in advance, and then entering the traffic sign information and location of the traffic sign into the roadside unit and broadcasting it. In this scenario, the analysis of the ROI area considers two aspects, one is the traffic sign information, and the other is the warning distance. The status information of the traffic light is mainly sent to the OBU through the roadside unit RSU. At present, the roadside communication coverage range of the lane can reach 300 meters. The warning distance needs to ensure that the vehicle passes the traffic sign in compliance with traffic regulations, and that driving warning information is issued when there is a driving risk. The traffic sign information in the dangerous area situation assessment results can also include speed limit signs, construction signs, school area signs, etc. Decision-making and early warning control can be carried out for the situation where the above traffic sign information appears based on the information in the vehicle-road perception fusion result list and the dangerous area situation assessment results.

[0135] The control module (ie, the decision-making module of the vehicle-mounted electronic unit) is described below.

[0136] According to the function and decision system design, the control in the method provided in the embodiment of the present invention is mainly the control of HMI (human-machine interface). The control module performs the control function including the following steps.

[0137] HMI control steps: When the system decides that a driving warning is needed, the corresponding control signal is sent to the HMI interface according to the functional design to warn the driver in the form of sound and image.

[0138] Control function takeover and exit steps: Based on the above basic warning functions, the driver takeover and exit mechanism of related functions can also be considered. For example, in a mild braking state, when the driver performs a braking operation but the deceleration is insufficient, the system will continue to remind the driver and continue to calculate the time to pass the road ahead to determine whether it can meet the safety requirements.

[0139] According to the embodiment of the present invention, a method for early warning and control of traffic information in a lane based on perception fusion technology has at least the following advantages: through V2V (or V2I) technology, the fusion of vehicle-mounted perception and roadside perception (including other vehicle driving status information sent from other vehicles) in underground mining scenarios is realized, and the front TSR information is obtained to make the vehicle more accurate and reliable, and the motion status information of the vehicle is obtained around the clock; scene judgment in scenes and situations where TSR information is blurred or blocked in underground mining scenarios; based on the perception fusion of the vehicle and the front TSR information in underground mining scenarios, control decisions are provided for the vehicle; the control decision function in underground mining scenarios takes safety and collision avoidance into consideration, and is suitable for the application of ADAS (Advanced Driver Assistance Systems) and automatic driving systems; the vehicle early warning control strategy in underground mining scenarios can obtain the front TSR information in advance to make prompts and warnings; using perception information, a braking pre-control strategy is applied to improve the system response speed, reduce braking delay, and improve collision avoidance performance.

[0140] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by any technician familiar with the technical field within the technical scope disclosed by the present invention should be covered within the protection scope of the present invention.

Claims

1. A method for early warning control of traffic information in a lane based on perception fusion technology, characterized in that: The method comprises the following steps: Step S1, scene recognition, identifying the sub-scenes in the underground mining scene where the vehicle is located through the vehicle-side information obtained in real time by the vehicle-mounted sensor, and judging one or more of the following situations: scene 1 that the traffic light at the intersection in the underground tunnel is blocked, scene 2 that the traffic light at the intersection in the underground tunnel is blurred or easily confused, and scene 3 that the traffic sign in the underground tunnel is blurred or easily confused. The vehicle-side information includes the vehicle position detected by the vehicle-mounted sensor, the vehicle speed, the traffic light status of the road ahead, the traffic light position, the traffic sign position, and the traffic sign information; Step S2, traffic light and traffic sign recognition, obtaining the traffic light status and traffic sign information of the vehicle's current lane as roadside information through the roadside unit, and sending the roadside information as roadside target information to the vehicle's onboard electronic unit after roadside sensing fusion processing; Step S3, vehicle-road perception fusion, the on-board electronic unit synchronizes the roadside target information received from the roadside unit with the vehicle-side information obtained in real time by the on-board sensor in time and space, and performs fusion processing to obtain vehicle-road matching target information, unmatched vehicle-side target information and unmatched roadside target information, which are sent to the decision module of the on-board electronic unit as a vehicle-road perception fusion result list; Step S4, dangerous area situation assessment, in step S1, when it is determined that the vehicle is in one or more of the scenarios 1, 2 and 3, the decision module of the on-board electronic unit performs dangerous area situation assessment according to the received vehicle-road perception fusion result list to obtain the dangerous area situation assessment result; Step S5: The decision module of the vehicle-mounted electronic unit performs decision processing and early warning control according to the obtained dangerous area situation assessment result to obtain a decision result or an early warning result.

2. The method for early warning control of traffic information in a lane based on perception fusion technology as claimed in claim 1, characterized in that: The step S3 further comprises: Step S3.1, obtaining the input of vehicle-road perception fusion, including the roadside target information received by the vehicle-mounted electronic unit from the roadside unit, and the vehicle-side target information, wherein the vehicle-side target information is the vehicle-side information acquired by the vehicle-mounted sensor in real time and outputted after the vehicle-side perception fusion processing is performed by the vehicle-mounted electronic unit; Step S3.2, a time-space synchronization step, in which the onboard electronic unit performs trend extrapolation prediction on the received roadside target information, compensates for the delay of the roadside target information relative to the vehicle-side target information, and obtains the roadside target information after time-space synchronization; Step S3.3, a fusion processing step, sequentially performing GNN association matching on the roadside target information and the vehicle-side information after time-space synchronization, to obtain successfully matched vehicle-road matching target information, as well as unmatched vehicle-side target information and unmatched roadside target information; Step S3.4, performing Kalman filter fusion and GNN multi-target matching tracking on the vehicle-road matching target information to obtain fused vehicle-road matching target information, and sending the fused vehicle-road matching target information together with the unmatched vehicle-side target information and the unmatched roadside target information as a vehicle-road perception fusion result list to the decision module of the on-board electronic unit; Among them, the list of vehicle-road perception fusion results includes the vehicle position, the vehicle initial speed v SV , the traffic light status of the road ahead, the location of the intersection where the traffic light is located, the location of the traffic sign, the traffic sign information, the road condition information ahead, and one or more of the remaining time of the green light countdown.

3. The method for early warning control of traffic information in a lane based on perception fusion technology as claimed in claim 2, characterized in that: The dangerous area situation assessment in step S4 includes traffic light status assessment, traffic sign assessment, and dangerous area judgment, and further includes the following steps: Step S4.1, when it is determined according to step S1 that the vehicle is in one or more of the scenarios 1, 2 and 3, the decision module of the onboard electronic unit evaluates the traffic light status according to the received vehicle-road perception fusion result list, including: when the vehicle-road perception fusion result list shows that the traffic light status of the road ahead is red, yellow or green, the vehicle position and the initial speed v of the vehicle are obtained. SV , traffic light status, intersection location of the traffic light, and remaining time of the green light countdown, and use this information as the result of the dangerous area situation assessment; Step S4.2, the decision module of the on-board electronic unit performs a traffic sign evaluation according to the received vehicle-road perception fusion result list, including: when the vehicle-road perception fusion result list shows that the traffic sign information ahead includes speed limit sign information, obtain the traffic sign position, traffic sign information, maximum speed limit v lim and use this information as the result of the dangerous area situation assessment; Step S4.3, the decision module of the on-board electronic unit makes a dangerous area judgment based on the received vehicle-road perception fusion result list, including: when the road condition information ahead in the vehicle-road perception fusion result list shows the existence of a construction area, obtains the safety line position and the boundary information of the dangerous area, and uses this information as the dangerous area situation assessment result.

4. The method for early warning control of traffic information in a lane based on perception fusion technology as claimed in claim 3, characterized in that: The step S5 further comprises the following steps: Step S5.1, in the obtained dangerous area situation assessment result, when the traffic light status is displayed as red light or yellow light, decision processing and early warning control are performed to obtain a decision result or an early warning result; Step S5.2, in the obtained dangerous area situation assessment result, the traffic light status is displayed as a green light countdown, and decision processing and early warning control are performed to obtain a decision result or an early warning result; Step S5.3: when the traffic sign information in the obtained dangerous area situation assessment result includes a speed limit sign, decision processing and early warning control are performed to obtain a decision result or an early warning result; Step S5.4: When there is a construction area in the obtained dangerous area situation assessment results, decision processing and early warning control are performed to obtain a decision result or an early warning result.

5. The method for early warning control of traffic information in a lane based on perception fusion technology as claimed in claim 4, characterized in that: The step S5.1 specifically includes the following steps: In the obtained dangerous area situation assessment results, when the traffic light status is red or yellow, the distance traveled by the vehicle when it is gently decelerated to a stop at the preset gentle acceleration a1 is calculated as Among them, v SV is the initial speed of the vehicle, t svd is the preset driver reaction time, t rbr is the preset braking system response time, d pre To preset a warning safety distance; Compare the distance S0 that the vehicle travels when gently decelerating to a stop with the distance from the vehicle's current position to the stop line corresponding to the traffic light: when S0 is less than the distance from the vehicle's current position to the stop line, gentle deceleration is performed as a decision result; when S0 is greater than the distance from the vehicle's current position to the stop line, the driver is prompted to take over as a warning result.

6. The method for early warning control of traffic information in a lane based on perception fusion technology as claimed in claim 5, characterized in that: The step S5.2 specifically includes the following steps: In the obtained dangerous area situation assessment results, the traffic light status is displayed as a green light countdown, which further determines whether the vehicle can pass through the intersection; The vehicle has an initial speed v SV The time to pass the intersection with traffic lights is Among them, d SV is the distance from the vehicle to the intersection where the traffic light is located, v SV is the initial speed of the vehicle, t sv Count down the remaining time for the green light; Compare the time TTC required for the vehicle to reach the intersection at the current speed and driving direction with the remaining time t of the green light countdown sv , if TTC≤t sv , the vehicle passes normally as the decision result; if TTC>t sv , the driver is prompted to take over as a warning result.

7. The method for early warning control of traffic information in a lane based on perception fusion technology as claimed in claim 6, characterized in that: The step S5.3 specifically includes the following steps: In the obtained dangerous area situation assessment results, when the traffic sign information includes a speed limit sign, compare the vehicle initial speed v SV Maximum speed limit with traffic signs lim , if the vehicle's initial speed is less than the maximum speed limit v on the traffic sign lim Then normal driving is taken as the decision result, otherwise the following steps are performed; When the vehicle's initial speed is greater than the maximum speed limit v on the traffic sign lim In the case of a vehicle with a moderate acceleration, the distance required to decelerate gently to the maximum speed limit of the traffic sign is calculated as follows: The distance S required for the vehicle to decelerate from the initial speed to the maximum speed limit with a gentle acceleration is compared. lim , and the distance from the vehicle’s current position to the traffic sign. When S lim When S is less than the distance from the current position of the vehicle to the traffic sign, a gentle deceleration is performed as the decision result; when S lim When it is greater than the distance from the vehicle's current position to the traffic sign, it indicates that it is not possible to pass with gentle deceleration, and the driver is prompted to take over as a warning result.

8. The method for early warning control of traffic information in a lane based on perception fusion technology as claimed in claim 7, characterized in that: The step S5.4 specifically includes the following steps: In the obtained dangerous area situation assessment results, when there is a construction area, the location of the construction area is obtained as the safety line location, and the distance traveled by the vehicle when it is gently decelerated to a stop at the preset gentle acceleration a1 is calculated as Compare the distance S0 that the vehicle travels when gently decelerating to a stop with the distance from the vehicle's current position to the safety line: when S0 is less than the distance from the vehicle's current position to the safety line, gentle deceleration is performed as a decision result; when S0 is greater than the distance from the vehicle's current position to the safety line, the driver is prompted to take over as a warning result.

Images