A collision warning method, system and storage medium

Abstract

The application provides a collision warning method, system and storage medium, comprising: constructing a target detection area in an electronic map based on target vehicle action information of a target vehicle; defining at least one auxiliary vehicle in the target detection area, and determining an auxiliary vehicle blind area for each auxiliary vehicle; collecting roadside sensing data of the auxiliary vehicle blind area through a roadside sensing device; when the roadside sensing data of the blind area contains early warning target action information, calculating a target vehicle action trajectory of the target vehicle and a warning target action trajectory of the early warning target; and generating collision warning information corresponding to the early warning target according to the target vehicle action trajectory and the warning target action trajectory. Through roadside sensing devices on both sides of the road, global visual field information of the blind area can be obtained, and whether the target vehicle and the early warning target will collide can be effectively predicted. Collision warning information is sent to remind the target vehicle before the collision is predicted to occur, and the target vehicle is helped to avoid the occurrence of the collision event in advance.

Description

Technical Field

[0001] This invention relates to the field of intelligent transportation technology, and in particular to a collision warning method, system and storage medium. Background Technology

[0002] During vehicle operation, blind spots are unavoidable due to the vehicle's own structure. Furthermore, blind spots can also be created by terrain, buildings, and other vehicles obstructing the view.

[0003] Pedestrians, non-motorized vehicles, and other vulnerable road users face significant safety hazards during travel, especially the so-called "ghost peek" incident, which occurs when a non-motorized vehicle or pedestrian suddenly darts out from a blind spot when there are vehicles or other obstacles blocking the view ahead.

[0004] To avoid traffic safety issues caused by "ghost peek" incidents, the relevant technology adopts the following solution: real-time detection of obstacles ahead that pose a "ghost peek" risk, identifying them as either vehicle-type or non-vehicle-type obstacles, acquiring blind spot images directly and indirectly, determining whether there are pedestrians in the blind spot based on the blind spot images, and when a pedestrian is present, defining a hypothetical collision zone by combining the vehicle's current position, current speed, pedestrian's position in the blind spot, and pedestrian's speed, and finally making a speed limit decision based on the hypothetical collision zone.

[0005] In the aforementioned solutions, most vehicles on the market lack complete perception capabilities, and even high-level intelligent vehicles cannot establish data links between enterprises or vehicles due to data privacy and security concerns, as well as technological barriers. Secondly, the surrounding environment may not have other vehicles or vehicles capable of acquiring images, or the acquired images may not cover the entire blind spot. Therefore, obtaining accurate and effective blind spot data poses a significant challenge to resolving the "ghost peek" incident. Summary of the Invention

[0006] In view of this, the purpose of the present invention is to provide a collision warning method, system and storage medium, which can acquire roadside perception data of the blind spot of the auxiliary vehicle in real time through roadside perception devices on both sides of the road, and the target vehicle can acquire global field of vision information of the blind spot of the auxiliary vehicle, thus solving the problem of difficulty in acquiring blind spot data.

[0007] Firstly, this application provides a collision warning method, comprising the following steps:

[0008] Based on the target vehicle's movement information, a target detection area corresponding to the target vehicle is constructed in the electronic map;

[0009] At least one auxiliary vehicle is defined within the target detection area, and an auxiliary vehicle blind spot is determined for each auxiliary vehicle relative to the target vehicle; the auxiliary vehicle blind spot is the field of view blocked by the auxiliary vehicle relative to the target vehicle.

[0010] Roadside sensing data of the blind spot of the auxiliary vehicle is collected by roadside sensing devices corresponding to the blind spot of the auxiliary vehicle; when the roadside sensing data of the blind spot contains the warning target action information of the preset type of warning target, the target vehicle action trajectory of the target vehicle is calculated based on the target vehicle action information, and the warning target action trajectory of the warning target is calculated based on the warning target action information.

[0011] Based on the trajectory of the target vehicle and the trajectory of the warning target, a collision warning message corresponding to the warning target is generated.

[0012] In one embodiment, constructing a target detection area corresponding to the target vehicle in an electronic map based on the target vehicle's movement information specifically includes:

[0013] The current position coordinates of the target vehicle are determined based on the target vehicle's movement information, and the current position coordinates of the target vehicle are defined as the first position coordinates. The current lane corresponding to the target vehicle in the electronic map is determined based on the first position coordinates.

[0014] Determine a second position coordinate located in the current lane and at a distance of a preset first length from the first position coordinate; and determine the area covered by the current lane located between the first position coordinate and the second position coordinate as the reference detection area.

[0015] In the electronic map, the areas on both sides of the reference detection area and whose distance from the edge of the reference detection area does not exceed a preset first width are defined as auxiliary detection areas;

[0016] The target vehicle detection area is generated based on the benchmark detection area and the auxiliary detection area.

[0017] In one embodiment, determining the auxiliary vehicle blind spot relative to the target vehicle for each of the auxiliary vehicles specifically includes:

[0018] The blind spot direction of the auxiliary vehicle's blind spot is determined based on the target vehicle's target vehicle movement information and the auxiliary vehicle's auxiliary vehicle movement information;

[0019] Based on the auxiliary vehicle's movement information, the current position coordinates of the auxiliary vehicle are determined, and the current position coordinates of the auxiliary vehicle are defined as the third position coordinates. A fourth position coordinate is determined in the blind spot direction and at a distance of a preset second length from the third position coordinates.

[0020] Along the blind spot direction, a rectangular region with the second length and a preset second width as its side lengths is established between the third position coordinate and the fourth position coordinate, and the rectangular region is defined as the blind spot of the auxiliary vehicle.

[0021] In one embodiment, before calculating the target vehicle trajectory based on the target vehicle trajectory based on the target vehicle trajectory based on the target vehicle trajectory based on the target vehicle trajectory based on the target vehicle trajectory based on the target vehicle trajectory information, when the roadside perception data of the blind spot contains target action information of a preset type, the method further includes:

[0022] The current location coordinates of the warning target are determined based on the warning target's action information, and the current location coordinates of the warning target are defined as the fifth location coordinates;

[0023] Based on the fifth position coordinates, it is determined whether the warning target is located in the target detection area; if so, the auxiliary vehicle blind spot where the warning target is currently located is determined according to the fifth position coordinates, and the auxiliary vehicle blind spot where the warning target is currently located is defined as the target blind spot.

[0024] The geometric range of the target blind spot is obtained based on the roadside perception data of the target blind spot;

[0025] The geometric range of the target blind spot is written into the collision warning information.

[0026] In one embodiment, after writing the geometric range of the target blind spot into the collision warning information, the method further includes:

[0027] Based on the target vehicle's movement information, determine whether the target vehicle is stationary; if so, send the collision warning information to the target vehicle; if not, determine whether the warning target is stationary based on the warning target's movement information; when the warning target is stationary, send the collision warning information to the target vehicle; when the warning target is not stationary, calculate the target vehicle's movement trajectory and the warning target's movement trajectory.

[0028] In one embodiment, before generating collision warning information corresponding to the warning target based on the target vehicle's trajectory and the warning target's trajectory, the method further includes:

[0029] Determine whether there is an intersection between the trajectory of the target vehicle and the trajectory of the warning target; if so, define the intersection as a collision point, calculate the collision distance between the collision point and the target vehicle and the minimum safe distance between the target vehicle and the warning target, and write the collision distance and the minimum safe distance into the target vehicle control information; if not, send a collision warning to the target vehicle.

[0030] In one embodiment, after calculating the collision distance between the collision point and the target vehicle, and the minimum safe distance between the target vehicle and the warning target, and generating target vehicle control information based on the collision distance and the minimum safe distance, the method further includes:

[0031] Generate safety distance coordinates based on the minimum safety distance;

[0032] Calculate the expected deceleration speed and expected deceleration distance of the target vehicle to reach the safe distance coordinates, and write the expected deceleration speed and the expected deceleration distance into the target vehicle control information;

[0033] Determine whether the collision distance is greater than the minimum safe distance; if so, calculate the action time of the warning target to reach the collision point and write the action time into the collision warning information; if not, send the target vehicle control information to the target vehicle, generate a target vehicle control command based on the target vehicle control information, and drive the target vehicle to execute the target vehicle control command.

[0034] In one embodiment, after sending the target vehicle control information to the target vehicle, generating a target vehicle control command based on the target vehicle control information, and driving the target vehicle to execute the target vehicle control command, the method further includes:

[0035] Identify at least one surrounding vehicle located within a preset distance range of the target vehicle, and calculate the positional relationship data between the surrounding vehicle and the target vehicle;

[0036] Based on the location relationship data and the target vehicle control information, surrounding vehicle control information is generated, the surrounding vehicle control information is sent to the surrounding vehicles, surrounding vehicle control commands are generated based on the surrounding vehicle control information, and the surrounding vehicles are driven to execute the surrounding vehicle control commands.

[0037] In a second aspect, this application provides a collision warning system, comprising: a processor and a memory; wherein the memory stores a computer program for being loaded by the processor and executed as described in any one of the first aspects.

[0038] Thirdly, this application provides a computer-readable storage medium storing instructions for loading by a processor and executing the collision warning method as described in any one of the first aspects.

[0039] In the collision warning method of this embodiment, roadside perception devices on both sides of the road can acquire roadside perception data of the blind spot of the auxiliary vehicle in real time. The target vehicle can acquire global field of vision information of the blind spot of the auxiliary vehicle within the target detection area, which solves the problem of difficulty in acquiring blind spot data. Based on the action information of the target vehicle and the warning target, it can effectively predict whether the target vehicle and the warning target will collide. Before the collision is predicted, a collision warning message is sent to remind the target vehicle, so that the driver of the target vehicle can take timely actions such as stopping or slowing down, and help the target vehicle avoid the occurrence of collision events in advance. Attached Figure Description

[0040] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation on the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0041] Figure 1 This is a system architecture diagram of the collision warning method of this application.

[0042] Figure 2 This is a flowchart illustrating the collision warning method in the embodiments of this application.

[0043] Figure 3 This is a schematic diagram of the construction of the detection area in an embodiment of this application.

[0044] Figure 4 This is a schematic diagram illustrating the creation of a blind zone in an embodiment of this application.

[0045] Figure 5 This is a schematic diagram of the collision warning system based on roadside perception in an embodiment of this application. Detailed Implementation

[0046] The specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are merely some, not all, of the embodiments of the present invention. Based on the description of the present invention, all other embodiments obtained by those skilled in the art without inventive effort are within the scope of protection of the present invention.

[0047] In the description of this invention, unless otherwise explicitly specified and limited, the terms "set," "install," "connect," etc., should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium. Those skilled in the art can understand the specific meaning of the above terms according to the specific circumstances.

[0048] The terms “upper,” “lower,” “left,” “right,” “front,” “back,” “top,” “bottom,” “inner,” and “outer,” etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings, or the orientation or positional relationship commonly used when the product of the invention is in use. They are only for the convenience of description and simplification, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on the present invention.

[0049] The terms “first,” “second,” “third,” etc., are used merely to distinguish elements with similar attributes, not to indicate or imply relative importance or a specific order.

[0050] The terms “include,” “comprising,” or any other variation thereof are intended to cover non-exclusive inclusion, which includes not only the elements listed but also other elements not expressly listed.

[0051] "Collision risk of vulnerable road users" (vulnerable road users generally refer to pedestrians or non-motorized vehicles, referred to as warning targets in this embodiment) has always been a significant safety hazard in transportation, especially the incident commonly known as "ghost peek" or "sudden pedestrian appearance." To effectively address the "ghost peek" problem, the industry has proposed different approaches from multiple perspectives, including management and technology, mainly including:

[0052] a. From a management perspective, this can be achieved by upgrading facilities and equipment at high-risk intersections, such as adjusting the position of the stop line for motor vehicles near the pedestrian crossing to allow for visual space; or by strengthening the management of pedestrians and non-motorized vehicles crossing the intersection / section.

[0053] b. From the perspective of traditional technology, when a single management method cannot effectively prevent pedestrians or non-motorized vehicles from crossing, add intelligent sensing roadside facilities, such as infrared beams or visual sensing, to monitor collision events of pedestrians or non-motorized vehicles and issue warnings (voice prompts) to minimize the probability of collisions caused by pedestrians or non-motorized vehicles suddenly appearing.

[0054] c. From a new technological perspective, focusing on the core issue that major traffic participants such as vehicle drivers, pedestrians, or non-motorized vehicles cannot effectively detect risks due to blind spots, AI-enabled methods are used to conduct "360-degree no-blind-spot" monitoring and predictive analysis of key areas through visual perception and vehicle-road cooperative technology, providing timely early warnings to risk subjects.

[0055] Methods a and b primarily achieve safety by sensing and alerting pedestrians or non-motorized vehicles, but they are difficult to effectively reach and control vehicles and drivers. Method c, based on the system concept of vehicle-road cooperation, can achieve real-time linkage with vehicle drivers, but due to differences in algorithms related to blind spot perception, risk assessment, and early warning strategies involved in "ghost peeks," the effects vary greatly.

[0056] The proposed solution primarily utilizes the vehicle-side to detect obstacles posing a "ghost peek" risk in real time, identifying them as either vehicle-related or non-vehicle-related. It acquires blind spot images directly and indirectly, then determines the presence of pedestrians within the blind spot. If a pedestrian is present, it uses the vehicle's current position and speed, along with the pedestrian's position and speed within the blind spot, to define a hypothetical collision zone. Finally, it makes a speed limit decision based on this hypothetical collision zone. This method, by connecting other vehicles to the vehicle, can extend its perception capabilities to some extent and makes different and effective decisions for different obstacles and blind spots, thus expanding its applicability.

[0057] The above scheme acquires blind spot images through direct and indirect means, utilizing images from cameras of vehicles near the blind spot before making a decision. This method has the following limitations:

[0058] (1) Blind spot data is difficult to obtain. On the one hand, most vehicles on the market do not have complete perception capabilities, and even high-level intelligent driving vehicles cannot establish data links between enterprises and between vehicles due to data privacy and security, technical barriers and other reasons; secondly, there may not be vehicles or vehicles with image acquisition capabilities in the surrounding environment, or the acquired images may not cover the entire range of the blind spot.

[0059] (2) The computing power and network bandwidth of the vehicle are difficult to meet. Theoretically, after the main vehicle detects an obstacle, it needs to send the blind spot image request behind the obstacle to the auxiliary vehicle (obstacle vehicle). The auxiliary vehicle then calls its camera to take an image of the specified area and sends it back to the main vehicle, which then performs collision detection. From a global perspective, blind spots between vehicles on the road exist at all times. If blind spot images are constantly being calculated and uploaded, it may put a lot of pressure on the vehicle's network and computing power, while the network bandwidth and computing power of the vehicle are relatively limited.

[0060] (3) Problem of delayed braking. 80% of traffic accidents are caused by human error or delayed reaction. If the driver receives collision risk information but does not observe or react in time, a collision risk may occur.

[0061] (4) Limited coverage of the warning system. Vehicles operate independently and do not share information. Only the vehicle itself can obtain blind spot information, while vehicles behind may not be able to obtain information about pedestrians or non-motorized vehicles in the blind spot. When the vehicle anticipates the risk of a collision in the blind spot and suddenly slows down, it may cause a rear-end collision, leading to a traffic accident.

[0062] In the collision warning method of this embodiment, roadside perception devices on both sides of the road can acquire roadside perception data of the blind spot of the auxiliary vehicle in real time. The target vehicle can acquire global field of vision information of the blind spot of the auxiliary vehicle within the target detection area, which solves the problem of difficulty in acquiring blind spot data. Based on the action information of the target vehicle and the warning target, it can effectively predict whether the target vehicle and the warning target will collide. Before the collision is predicted, a collision warning message is sent to remind the target vehicle, so that the driver of the target vehicle can take timely actions such as stopping or slowing down, and help the target vehicle avoid the occurrence of collision events in advance.

[0063] Figure 1 This is a system architecture diagram of a collision warning method provided in this embodiment. The warning command center establishes connections with an in-vehicle terminal cluster and roadside sensing devices. The in-vehicle terminal cluster includes multiple in-vehicle terminals, and each target vehicle 10 is equipped with at least one in-vehicle terminal. The roadside sensing devices may include image acquisition devices such as cameras 20, and signal acquisition devices such as millimeter-wave radar and lidar. The warning command center may include a server 40 with data processing and storage capabilities and a cloud platform 30. The in-vehicle terminals are monitoring terminals used for monitoring and managing the target vehicles 10. Installed inside the target vehicles 10, the in-vehicle terminals can collect real-time movement information of the target vehicles 10 and interact with the cloud platform to achieve warning and control of the target vehicles 10.

[0064] like Figure 2 As shown, the present invention provides a collision warning method, comprising the following steps:

[0065] Step S100: Based on the target vehicle's movement information, construct a target detection area corresponding to the target vehicle in the electronic map;

[0066] Step S200: Define at least one auxiliary vehicle within the target detection area, and determine the auxiliary vehicle blind spot of each auxiliary vehicle relative to the target vehicle; the auxiliary vehicle blind spot is the field of view blocked by the auxiliary vehicle relative to the target vehicle;

[0067] Step S300: Collect roadside perception data of the blind spot of the auxiliary vehicle through the roadside perception device corresponding to the blind spot of the auxiliary vehicle; when the roadside perception data of the blind spot contains the warning target action information of the preset type of warning target, calculate the target vehicle action trajectory of the target vehicle based on the target vehicle action information, and calculate the warning target action trajectory of the warning target based on the warning target action information.

[0068] Step S400: Generate collision warning information corresponding to the warning target based on the target vehicle's trajectory and the warning target's trajectory.

[0069] In step S100, the target vehicle's onboard terminal registers an early warning function on the cloud platform of the early warning command center. The target vehicle's size, color, license plate number, and other attribute information are stored in the memory of the early warning command center's server. While the target vehicle is in motion, its onboard terminal collects real-time target vehicle movement information and sends this information to the cloud platform. This target vehicle movement information includes the target vehicle's attribute information, location information, speed information, acceleration information, and heading information.

[0070] The server in the early warning and command center can construct a target detection area corresponding to the target vehicle in a preset electronic map based on the target vehicle's movement information. This includes the following steps:

[0071] Step S101: Determine the current position coordinates of the target vehicle based on the target vehicle movement information, and define the current position coordinates of the target vehicle as the first position coordinates. Determine the current lane of the target vehicle in the electronic map based on the first position coordinates.

[0072] Step S102: Determine the second position coordinates located in the current lane and at a distance of a preset first length from the first position coordinates; determine the area covered by the current lane located between the first position coordinates and the second position coordinates as the reference detection area.

[0073] Step S103: In the electronic map, the areas on both sides of the reference detection area and whose distance from the edge of the reference detection area does not exceed a preset first width are determined as auxiliary detection areas;

[0074] Step S104: Generate the target vehicle detection area based on the benchmark detection area and the auxiliary detection area.

[0075] In step S101, as Figure 3As shown, the first location coordinates are the actual latitude and longitude coordinates of the target vehicle 51. The vehicle terminal sends the first location coordinates to the cloud platform. The server finds the corresponding coordinate point in the electronic map based on the first location coordinates received by the cloud platform, and uses this coordinate point as the location coordinates of the target vehicle 51. The electronic map in this embodiment is a high-precision map with an absolute accuracy of not less than 50cm and a relative error of no more than 20cm per 100m, which can accurately locate the current lane of the target vehicle based on the first location coordinates. Here, the current lane refers to the actual lane position currently occupied by the target vehicle.

[0076] In steps S102-S104, a second position coordinate with a distance of a first length L1 from the first position coordinate is determined on the current lane. Using the first and second position coordinates as endpoints, a reference detection area 61 is constructed on the first lane. An auxiliary detection area 62 with a first width W1 is then widened on both the left and right sides of the reference detection area 61. Combining the reference detection area 61 and the auxiliary detection areas 62 generates the target detection area. The reference detection area is the region on the trajectory line in front of the target vehicle, where a collision between the target vehicle and the warning target is most likely to occur. Pedestrians, non-motorized vehicles, and other warning targets can easily enter the reference detection area 61 through the auxiliary detection areas on the left and right sides. The target detection area constructed in this embodiment can effectively monitor warning targets located in the region in front of the vehicle and promptly send collision warning information to the target vehicle.

[0077] Optionally, the current lane can be a straight lane, and the reference detection area, the auxiliary detection area, and the target detection area obtained by combining the two are all rectangular areas adapted to the current lane. The current lane can also be a non-straight lane with curvature or corners, and the reference detection area, the auxiliary detection area, and the target detection area obtained by combining the two are all non-rectangular areas adapted to the current lane.

[0078] Optionally, the first length L1 in this embodiment can be a fixed value such as 20m, 50m, or 100m. The first length L1 can also be adaptively generated based on the speed of the target vehicle. For example, the first length L1 has three settings: when the target vehicle's speed is between 0 and 30 km / h, the first length L1 is set to 20m; when the target vehicle's speed is between 30 and 60 km / h, the first length L1 is set to 50m; and when the target vehicle's speed exceeds 60 km / h, the first length L1 is set to 100m. It is understood that the above values ​​for the first length L1 are merely examples and can be adaptively set according to the actual conditions of the target vehicle and the road.

[0079] Optionally, the first width W1 in this embodiment can be set according to the width of the current lane or according to the width of the target vehicle. Preferably, since it is difficult for the driver of the target vehicle to directly obtain the movement of the warning targets in the left and right lanes before they enter the current lane, the first width W1 in this embodiment is equal to the width of the current lane, which can accurately obtain the movement of the warning targets in the left and right lanes and send collision warning information to the target vehicle in a timely manner.

[0080] In step S200, as Figure 3 and Figure 4 As shown, the auxiliary vehicle 52 is another vehicle located in front of the target vehicle 51 that can block the target vehicle 51's line of sight, and the auxiliary vehicle blind spot 70 is the area located in which the auxiliary vehicle 52 blocks the target vehicle's line of sight.

[0081] Specifically, creating a blind zone for each auxiliary vehicle relative to the target vehicle within the target detection area includes the following steps:

[0082] Step S201: Determine the blind spot direction of the auxiliary vehicle's blind spot based on the target vehicle's target vehicle movement information and the auxiliary vehicle's auxiliary vehicle movement information;

[0083] Step S202: Determine the current position coordinates of the auxiliary vehicle based on the auxiliary vehicle movement information, define the current position coordinates of the auxiliary vehicle as the third position coordinates, and determine the fourth position coordinates in the blind spot direction and at a distance of a preset second length from the third position coordinates;

[0084] Step S203: Along the blind spot direction, establish a rectangular region with the second length and the preset second width as the side lengths between the third position coordinate and the fourth position coordinate, and define the rectangular region as the blind spot of the auxiliary vehicle.

[0085] In step S201, the target vehicle's action information includes the target vehicle's first position coordinates, and the auxiliary vehicle's action information includes the auxiliary vehicle's third position coordinates. The blind spot direction of the auxiliary vehicle's blind spot can be determined by connecting the first and third position coordinates.

[0086] In steps S202 and S203, a fourth position coordinate is determined in the blind spot direction at a distance of a second length L2 from the third position coordinate. Using the third and fourth position coordinates as endpoints, an auxiliary vehicle blind spot is constructed in front of the auxiliary vehicle.

[0087] Optionally, such as Figure 4As shown, the blind spot 70 of the auxiliary vehicle is a rectangular area. The second width W2 of the blind spot 70 is the length of the hypotenuse of the target vehicle 51. The target vehicle's movement information includes the hypotenuse length. The second length L2 of the blind spot 70 is the maximum value of the minimum safe distance S of the target vehicle 51 at its current speed and the preset safe distance length. Its expression is L2 = max(S, length). Refer to "T / CSAE 246-2022 Test and Evaluation Method for Early Warning Application Function of Intelligent Connected Vehicle V2X System". The formula for calculating the minimum safe distance S is as follows:

[0088]

[0089] Among them, V s t1 is the current speed of the target vehicle (in m / s), T is the driver's reaction time (ranging from 0.3s to 2s), t1 is the braking coordination time (ranging from 0.35s to 0.6s), t2 is the deceleration increase time (t2 is usually 0.2s), and a is the deceleration increase time. s The average deceleration of the driver during braking (a) s The value range is 3.6m. 2 / s~7.9m 2 / s), d0 is the safe distance when the target vehicle is stationary (d0 defaults to 3m), and t3 is the link delay calculated by the algorithm (t3 defaults to 0.5s).

[0090] In step S300, the roadside perception device for the blind spot of the auxiliary vehicle includes at least image acquisition devices such as cameras. Based on the image information acquired by the image acquisition devices, the server can accurately identify whether there are warning targets such as non-motorized vehicles and pedestrians in the blind spot of the auxiliary vehicle through intelligent recognition technology. It then generates the target vehicle's trajectory based on the target vehicle's current speed, acceleration, and heading, and generates the warning target's trajectory based on the warning target's current speed, acceleration, and direction of movement. The target vehicle trajectory and the warning target trajectory can be generated using methods such as multi-order Bézier curves, MPC prediction models, and spline curves; this embodiment does not impose any limitations on these methods.

[0091] The image acquisition device in this embodiment is installed on both sides of the road and has visual perception algorithm capabilities. It detects image data of vehicles, pedestrians and other early warning targets through local perception and converts them into structured data. Then, it performs multi-view trajectory fusion through the cloud platform big data computing center to form target trajectory data for global perception, thereby covering all auxiliary vehicle blind spots within the target detection area and solving the problem of difficult acquisition of auxiliary vehicle blind spot data.

[0092] When the roadside perception data of the blind spot contains warning target movement information of a preset type of warning target, before calculating the target vehicle movement trajectory of the target vehicle based on the target vehicle movement information and before calculating the warning target movement trajectory of the warning target based on the warning target movement information, the method further includes:

[0093] Step S501: Determine the current position coordinates of the warning target based on the warning target action information, and define the current position coordinates of the warning target as the fifth position coordinates;

[0094] Step S502: Determine whether the warning target is located in the target detection area based on the fifth position coordinates; if so, determine the auxiliary vehicle blind spot where the warning target is currently located based on the fifth position coordinates, and define the auxiliary vehicle blind spot where the warning target is currently located as the target blind spot;

[0095] Step S503: Obtain the geometric range of the target blind spot based on the roadside perception data of the target blind spot;

[0096] Step S504: Write the geometric range of the target blind zone into the collision warning information.

[0097] In steps S501 to S504, the warning target's movement information includes the warning target's attribute information, location information, speed information, acceleration information, and direction of movement information. The location information includes the warning target's current latitude and longitude coordinates, i.e., the fifth position coordinates. The server compares these fifth position coordinates with the location range of the detection area to identify whether the warning target is within the target detection area. If the warning target is not within the target detection area, it indicates that the probability of a collision between the warning target and the target vehicle is low, and the warning command center does not need to calculate the collision point between the target vehicle's trajectory and the warning target's movement trajectory.

[0098] When the warning target is within the target detection area, it is necessary to determine whether the warning target is within the blind spot of the auxiliary vehicle. If the warning target is not within the auxiliary vehicle's blind spot, meaning its trajectory is within the target vehicle's field of vision, the driver can directly obtain the target's location information without requiring a collision warning from the command center. If the warning target is within the auxiliary vehicle's blind spot, the blind spot is marked as the target blind spot, and its geometric range data is written into the collision warning information. This allows the driver to anticipate the target's movement and take timely actions such as stopping or slowing down, effectively avoiding the risk of a collision between the warning target and the target vehicle.

[0099] Optionally, when multiple warning targets exist on the road segment in which the target vehicle is traveling, it is necessary to determine the blind spot of each warning target and mark the corresponding blind spot as a "target blind spot." To distinguish between target blind spots and non-target blind spots, a "YES" mark can be placed on a target blind spot, indicating the presence of a warning target and requiring the driver to exercise caution when passing near that blind spot. A "NO" mark can be placed on a non-target blind spot, indicating the absence of a warning target and allowing the driver to safely pass near it. Through the "YES" and "NO" markings, the driver of the target vehicle can intuitively identify which blind spot contains a warning target, enabling timely vehicle control actions such as stopping or slowing down, thus helping the target vehicle avoid collisions in advance.

[0100] After writing the geometric range of the target blind spot into the collision warning information, the method further includes step S505: determining whether the target vehicle is stationary based on the target vehicle movement information; if yes, sending the collision warning information to the target vehicle; if no, determining whether the warning target is stationary based on the warning target movement information; when the warning target is stationary, sending the collision warning information to the target vehicle; when the warning target is not stationary, calculating the target vehicle movement trajectory and the warning target movement trajectory.

[0101] In step S505, there exists a scenario where an auxiliary vehicle located in front of the target vehicle has stopped, and the target vehicle also stops following the auxiliary vehicle. For example, at a traffic light intersection, the traffic lane is red, and the pedestrian lane is green. At this time, the target vehicle and the auxiliary vehicle are stopped in the traffic lane waiting for the red light, while pedestrians and non-motorized vehicles are crossing the pedestrian lane. Even if the target vehicle has a blind spot, since the target vehicle has stopped, a collision between the target vehicle and the pedestrian lane is almost impossible. In this scenario, there is no need to additionally predict the collision between the target vehicle and the pedestrian lane and generate collision warning information.

[0102] In addition, there are scenarios where the warning target has already stopped, waiting for the target vehicle and auxiliary vehicles ahead of it to pass through the road segment first. For example, at a traffic light intersection, the traffic lane is green and the pedestrian lane is red. At this time, the warning target is stopped waiting for the red light and has no tendency to cross the pedestrian lane. In this case, there is no need to additionally predict the collision between the target vehicle and the warning target and generate collision warning information.

[0103] In this embodiment, the early warning and command center is equipped with a cloud platform with big data computing capabilities, which can significantly improve the computing power of the early warning and command center. The big data cloud platform has the characteristics of high computing power and high bandwidth. Using the big data platform to assist in the collision risk calculation of the vehicle's blind spot can not only solve the problem of low computing power and low bandwidth on the vehicle side, but also obtain data from a larger and farther field of view. This has the advantage of helping to detect risks in advance and help the target vehicle avoid collision events in advance.

[0104] Before generating collision warning information corresponding to the warning target based on the target vehicle's trajectory and the warning target's trajectory, the method further includes step S600: determining whether there is an intersection between the target vehicle's trajectory and the warning target's trajectory; if so, defining the intersection as a collision point, calculating the collision distance between the collision point and the target vehicle, as well as the minimum safe distance between the target vehicle and the warning target, and writing the collision distance and the minimum safe distance into the target vehicle control information; if not, sending collision warning information to the target vehicle.

[0105] In step S600, the server generates a target vehicle trajectory curve based on the target vehicle's movement trajectory and a warning target trajectory curve based on the warning target's movement trajectory. The collision point is the intersection of the target vehicle trajectory curve and the warning target trajectory curve. The collision point carries the predicted collision coordinates and collision time. Based on the collision coordinates, the collision distance between the collision point and the target vehicle, as well as the minimum safe distance between the target vehicle and the warning target, can be calculated.

[0106] After calculating the collision distance between the collision point and the target vehicle, and the minimum safe distance between the target vehicle and the warning target, and generating target vehicle control information based on the collision distance and the minimum safe distance, the method further includes:

[0107] Step S601: Generate safety distance coordinates based on the minimum safety distance;

[0108] Step S602: Calculate the expected deceleration speed and expected deceleration distance of the target vehicle to reach the safe distance coordinates, and write the expected deceleration speed and expected deceleration distance into the target vehicle control information;

[0109] Step S603: Determine whether the collision distance is greater than the minimum safe distance; if yes, calculate the action time of the warning target to reach the collision point and write the action time into the collision warning information; if no, send the target vehicle control information to the target vehicle, generate a target vehicle control command based on the target vehicle control information, and drive the target vehicle to execute the target vehicle control command.

[0110] In steps S601 to S603, the formula for calculating the minimum safe distance S is as follows:

[0111]

[0112] Among them, V s t1 is the current speed of the target vehicle (in m / s), T is the driver's reaction time (ranging from 0.3s to 2s), t1 is the braking coordination time (ranging from 0.35s to 0.6s), t2 is the deceleration increase time (t2 is usually 0.2s), and a is the deceleration increase time. s The average deceleration of the driver during braking (a) s The value range is 3.6m. 2 / s~7.9m 2 / s), d0 is the safe distance when the target vehicle is stationary (d0 defaults to 3m), and t3 is the link delay calculated by the algorithm (t3 defaults to 0.5s).

[0113] When the target vehicle is traveling in its current lane, the distance between the safe distance coordinate and the point of collision is the minimum safe distance, and the distance between the target vehicle's coordinates and the safe distance coordinates is the expected deceleration distance S. target According to the expected deceleration distance S target The target vehicle's speed V s and the driver's average braking deceleration a s Through formula To achieve the desired deceleration speed V target V s The target vehicle's current speed, a s The average deceleration of the driver during braking (a) s The value range is 3.6m. 2 / s~7.9m 2 / s). The time it takes for the warning target to reach the collision point can be generated based on data such as the current position coordinates of the warning target, the coordinates of the collision point, the current speed of the warning target, and the average speed of the warning target.

[0114] Optionally, the collision warning information includes at least the location of the warning target, the relative direction between the warning target and the target vehicle, the distance between the warning target and the target vehicle, the blind spot of which auxiliary vehicle the warning target is located, and the geometric range of the corresponding auxiliary vehicle blind spot.

[0115] Optionally, the first vehicle control information includes at least the location of the warning target, the relative direction between the warning target and the target vehicle, the distance between the warning target and the target vehicle, the blind spot of which auxiliary vehicle the warning target is located and the geometric range of the corresponding auxiliary vehicle blind spot, the time to reach the collision point (TTC) of the warning target, and the expected deceleration speed (V). targetand expected deceleration distance S target .

[0116] In this embodiment, when the collision risk is high, the cloud platform automatically sends control information to the target vehicle and instructs the target vehicle to execute vehicle control commands such as deceleration and stopping, so as to avoid collision events caused by the driver's untimely reaction or braking when the target vehicle receives a collision risk.

[0117] After sending the target vehicle control information to the target vehicle, generating a target vehicle control command based on the target vehicle control information, and driving the target vehicle to execute the target vehicle control command, the method further includes:

[0118] Step S701: Determine at least one surrounding vehicle located within a preset distance range of the target vehicle, and calculate the positional relationship data between the surrounding vehicle and the target vehicle;

[0119] Step S702: Generate surrounding vehicle control information based on the location relationship data and the target vehicle control information, send the surrounding vehicle control information to the surrounding vehicles, generate surrounding vehicle control commands based on the surrounding vehicle control information, and drive the surrounding vehicles to execute the surrounding vehicle control commands.

[0120] In steps S701 and S702, the preset distance range of the target vehicle is usually the vehicles behind and to the left and right of the target vehicle. The preset distance range can be customized according to the road conditions. When the road is congested, the preset distance range can be smaller, and when the road is relatively open, the preset distance range can be larger.

[0121] The positional relationship data between the surrounding vehicles and the target vehicle typically includes the distance between the surrounding vehicles and the target vehicle, and the orientation of the target vehicle relative to the surrounding vehicles. There are usually multiple surrounding vehicles within the preset range, therefore the surrounding vehicle control information received by each surrounding vehicle is not entirely the same. For example, when the target vehicle is in the rightmost lane of a three-lane road, surrounding vehicle A is in the leftmost lane, surrounding vehicle B is in the rightmost lane, and the auxiliary vehicle is in the middle lane, warning of a pedestrian crossing to the right. Since the warning target does not show any tendency to move to the left, the surrounding vehicle control information received by surrounding vehicle A does not include a stop instruction; only a deceleration is required. However, the surrounding vehicle control information received by surrounding vehicle B includes a stop instruction.

[0122] Optionally, this implementation utilizes a 5G network to achieve low-latency vehicle-to-vehicle information sharing. When a collision risk is imminent, the 5G network transmits surrounding vehicle control information to vehicles behind and nearby, providing sufficient information about blind spots for vehicles behind. Based on the received surrounding vehicle control information, surrounding vehicles can simultaneously adopt the same control strategies as the target vehicle, thereby preventing rear-end collisions caused by delayed braking from vehicles behind.

[0123] In summary, in the collision warning method of this embodiment, firstly, roadside sensing devices on both sides of the road can acquire roadside sensing data of the blind spots of auxiliary vehicles in real time, and the target vehicle can acquire global field-of-view information of the blind spots of auxiliary vehicles within the target detection area, thus solving the problem of difficulty in acquiring blind spot data. Secondly, the warning command center is equipped with a cloud platform with big data computing capabilities, which can significantly improve the computing power of the warning command center. The big data cloud platform has the characteristics of high computing power and high bandwidth. Using the big data platform to calculate blind spot collision risks can not only solve the problem of low computing power and small bandwidth on the vehicle side, but also acquire data with a larger and farther field of view. This helps to detect risks in advance and helps vehicles avoid collision events in advance. Thirdly, through vehicle control via the cloud platform, when the collision risk is high, the cloud platform automatically sends control information to the target vehicle and instructs the vehicle to execute vehicle control commands such as deceleration and stopping, avoiding collision events caused by the driver's untimely reaction or braking when the vehicle receives a collision risk. Fourthly, this implementation achieves low-latency vehicle-to-vehicle information sharing through the 5G network. When a collision risk is about to occur, the 5G network will send the control information of the surrounding vehicles to the vehicles behind and nearby, providing the vehicles behind with sufficient blind spot information. The surrounding vehicles can simultaneously take the same control strategy as the target vehicle based on the received control information of the surrounding vehicles, thereby avoiding rear-end collisions caused by the vehicles behind not braking in time.

[0124] like Figure 5 As shown, based on the same conceptual approach as the aforementioned embodiments, the present invention also provides a collision warning device based on roadside perception, comprising:

[0125] The target detection region construction module M10 is used to construct a target detection region corresponding to the target vehicle in an electronic map based on the target vehicle's movement information.

[0126] The auxiliary vehicle blind spot creation module M20 is used to define at least one auxiliary vehicle within the target detection area and to determine the auxiliary vehicle blind spot of each auxiliary vehicle relative to the target vehicle; the auxiliary vehicle blind spot is the field of view blocked by the auxiliary vehicle relative to the target vehicle.

[0127] The trajectory calculation module M30 is used to collect roadside perception data of the blind spot of the auxiliary vehicle through the roadside perception device corresponding to the blind spot of the auxiliary vehicle; when the roadside perception data of the blind spot contains the target action information of the target vehicle of a preset type, the module calculates the target vehicle trajectory of the target vehicle based on the target vehicle action information, and calculates the target action trajectory of the warning target based on the warning target action information.

[0128] The collision warning information generation module M40 is used to generate collision warning information corresponding to the target vehicle based on the target vehicle's trajectory and the target vehicle's trajectory.

[0129] In the roadside perception-based collision warning device of this embodiment, roadside perception data of the blind spot of the auxiliary vehicle can be obtained in real time through roadside perception devices on both sides of the road. The target vehicle can obtain global field of view information of the blind spot of the auxiliary vehicle within the target detection area, which solves the problem of difficulty in obtaining blind spot data. Based on the action information of the target vehicle and the warning target, it can effectively predict whether the target vehicle and the warning target will collide. Before the collision is predicted, a collision warning message is sent to remind the target vehicle, so that the driver of the target vehicle can take timely actions such as stopping or slowing down, and help the target vehicle avoid the occurrence of collision events in advance.

[0130] This application also provides a collision warning system, including: a processor and a memory; wherein the memory stores a computer program, the computer program being loaded by the processor and executed as described above in the collision warning method.

[0131] This application also provides a computer-readable storage medium storing instructions for being loaded by a processor and executed as described above regarding the collision warning method.

[0132] The embodiments of the mobile terminal and computer-readable storage medium provided in this application include all the technical features of the embodiments of the above control method. The extended and explanatory content of the specification is basically the same as that of the embodiments of the above method, and will not be repeated here.

[0133] This application also provides a computer program product, which includes computer program code. When the computer program code is run on a computer, it causes the computer to perform the methods described in the various possible implementations above.

[0134] This application also provides a chip, including a memory and a processor. The memory is used to store a computer program, and the processor is used to call and run the computer program from the memory, so that a device with the chip installed performs the methods described in the various possible implementations above.

[0135] The sequence numbers of the embodiments in this application are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0136] In this application, the same or similar terms, concepts, technical solutions and / or application scenario descriptions are generally described in detail only when they appear for the first time. When they appear again, they are generally not repeated for the sake of brevity. When understanding the technical solutions and other contents of this application, the same or similar terms, concepts, technical solutions and / or application scenario descriptions that are not described in detail later can be referred to their previous relevant detailed descriptions.

[0137] In this application, the descriptions of the various embodiments have different focuses. For parts that are not described in detail or recorded in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0138] The technical features of the present application can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of the present application.

[0139] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in the above-mentioned storage medium and includes several instructions to cause a terminal device to execute the methods of each embodiment of this application. The above are only preferred embodiments of this application and do not limit the patent scope of this application. Any equivalent structural or procedural transformations made based on the content of this application's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of this application.

[0140] It should be noted that the various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. The same or similar parts between the various embodiments can be referred to each other.

[0141] 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 variations or substitutions that can be easily conceived by those 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 determined by the appended claims.

Claims

1. A collision warning method, characterized in that, Includes the following steps: Based on the target vehicle's movement information, a target detection area corresponding to the target vehicle is constructed in the electronic map; At least one auxiliary vehicle is defined within the target detection area, and an auxiliary vehicle blind spot is determined for each auxiliary vehicle relative to the target vehicle; the auxiliary vehicle blind spot is the field of view blocked by the auxiliary vehicle relative to the target vehicle. The step of determining the blind spot of each of the auxiliary vehicles relative to the target vehicle specifically includes: determining the blind spot direction of the auxiliary vehicle blind spot based on the target vehicle's movement information and the auxiliary vehicle's movement information; determining the current position coordinates of the auxiliary vehicle based on the auxiliary vehicle's movement information, defining the current position coordinates of the auxiliary vehicle as a third position coordinate, and determining a fourth position coordinate in the blind spot direction and at a distance of a preset second length from the third position coordinate; establishing a rectangular region with sides of the second length and a preset second width respectively between the third position coordinate and the fourth position coordinate along the blind spot direction, and defining the rectangular region as the auxiliary vehicle blind spot; Roadside sensing data of the blind spot of the auxiliary vehicle is collected by roadside sensing devices corresponding to the blind spot of the auxiliary vehicle; when the roadside sensing data of the blind spot contains the warning target action information of the preset type of warning target, the target vehicle action trajectory of the target vehicle is calculated based on the target vehicle action information, and the warning target action trajectory of the warning target is calculated based on the warning target action information. Based on the trajectory of the target vehicle and the trajectory of the warning target, a collision warning message corresponding to the warning target is generated.

2. The collision warning method according to claim 1, characterized in that, The construction of a target detection area corresponding to the target vehicle in the electronic map based on the target vehicle's movement information specifically includes: The current position coordinates of the target vehicle are determined based on the target vehicle's movement information, and the current position coordinates of the target vehicle are defined as the first position coordinates. The current lane corresponding to the target vehicle in the electronic map is determined based on the first position coordinates. Determine a second position coordinate located in the current lane and at a distance of a preset first length from the first position coordinate; and determine the area covered by the current lane located between the first position coordinate and the second position coordinate as the reference detection area. In the electronic map, the areas on both sides of the reference detection area and whose distance from the edge of the reference detection area does not exceed a preset first width are defined as auxiliary detection areas; The target vehicle detection area is generated based on the benchmark detection area and the auxiliary detection area.

3. The collision warning method according to claim 1, characterized in that, When the roadside perception data of the blind spot contains warning target movement information of a preset type of warning target, before calculating the target vehicle movement trajectory of the target vehicle based on the target vehicle movement information and before calculating the warning target movement trajectory of the warning target based on the warning target movement information, the method further includes: The current location coordinates of the warning target are determined based on the warning target's action information, and the current location coordinates of the warning target are defined as the fifth location coordinates; Based on the fifth position coordinates, it is determined whether the warning target is located in the target detection area; if so, the auxiliary vehicle blind spot where the warning target is currently located is determined according to the fifth position coordinates, and the auxiliary vehicle blind spot where the warning target is currently located is defined as the target blind spot. The geometric range of the target blind spot is obtained based on the roadside perception data of the target blind spot; The geometric range of the target blind spot is written into the collision warning information.

4. The collision warning method according to claim 3, characterized in that, After writing the geometric range of the target blind spot into the collision warning information, the following is also included: Based on the target vehicle's movement information, determine whether the target vehicle is stationary; if so, send the collision warning information to the target vehicle; if not, determine whether the warning target is stationary based on the warning target's movement information; when the warning target is stationary, send the collision warning information to the target vehicle; when the warning target is not stationary, calculate the target vehicle's movement trajectory and the warning target's movement trajectory.

5. A collision warning method according to claim 1, characterized in that, Before generating collision warning information corresponding to the warning target based on the target vehicle's trajectory and the warning target's trajectory, the method further includes: Determine whether there is an intersection between the trajectory of the target vehicle and the trajectory of the warning target; if so, define the intersection as a collision point, calculate the collision distance between the collision point and the target vehicle and the minimum safe distance between the target vehicle and the warning target, and write the collision distance and the minimum safe distance into the target vehicle control information; if not, send a collision warning to the target vehicle.

6. A collision warning method according to claim 5, characterized in that, After calculating the collision distance between the collision point and the target vehicle, and the minimum safe distance between the target vehicle and the warning target, and generating target vehicle control information based on the collision distance and the minimum safe distance, the method further includes: Generate safety distance coordinates based on the minimum safety distance; Calculate the expected deceleration speed and expected deceleration distance of the target vehicle when it reaches the safe distance coordinates, and write the expected deceleration speed and the expected deceleration distance into the target vehicle control information; Determine whether the collision distance is greater than the minimum safe distance; if so, calculate the action time of the warning target to reach the collision point and write the action time into the collision warning information; if not, send the target vehicle control information to the target vehicle, generate a target vehicle control command based on the target vehicle control information, and drive the target vehicle to execute the target vehicle control command.

7. A collision warning method according to claim 6, characterized in that, After sending the target vehicle control information to the target vehicle, generating a target vehicle control command based on the target vehicle control information, and driving the target vehicle to execute the target vehicle control command, the method further includes: Identify at least one surrounding vehicle located within a preset distance range of the target vehicle, and calculate the positional relationship data between the surrounding vehicle and the target vehicle; Based on the location relationship data and the target vehicle control information, surrounding vehicle control information is generated, the surrounding vehicle control information is sent to the surrounding vehicles, surrounding vehicle control commands are generated based on the surrounding vehicle control information, and the surrounding vehicles are driven to execute the surrounding vehicle control commands.

8. A collision warning system, comprising: A processor and a memory; wherein the memory stores a computer program for being loaded by the processor and executed as described in any one of claims 1-7.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores instructions for loading by a processor and executing the collision warning method as described in any one of claims 1-7.

Images