Vehicle collision time calculation method, system, and storage medium

Abstract

The application discloses a kind of vehicle collision time calculation methods, comprising: obtaining the equivalent rectangular four vertex coordinates of ego vehicle and target vehicle at current time;Every vertex of ego vehicle and target vehicle is traversed by inner product operation, whether there is separation axis is judged, if there is separation axis, then calculate TTC, if there is no separation axis, then output current TTC and minimum distance is 0;The vertex distance of ego vehicle and target vehicle at current time is calculated, all vertices are traversed, a pair of nearest vertices is selected, and the first specified time length is taken as the first step to traverse all time, to obtain the TTC of coarse step prediction nearest point;With the nearest point TTC of coarse step prediction as center, first step interval is taken forward and backward respectively, assuming that ego vehicle is stationary in selected interval, target vehicle moves at uniform speed with relative speed, and the minimum distance and TTC are calculated by fine step prediction;The result of coarse step prediction and precision prediction is added to obtain the TTC of the nearest point of ego vehicle and target vehicle.

Description

Technical Field

[0001] The present invention relates to the automotive field, and in particular to a method for calculating vehicle collision time in a vehicle intelligent driving or advanced driver assistance system. Background Art

[0002] In intelligent driving or advanced driver assistance systems, smart cars need to judge and assess the collision risk between themselves and target vehicles on the road before making lane change decisions. This mainly includes trajectory prediction of themselves and road vehicles, TTC calculation, and collision probability calculation.

[0003] Currently, there are two main methods for calculating the TTC between the ego vehicle and road vehicles: CV and CA. One method assumes that the ego vehicle and target vehicle are moving at a constant speed, while the other assumes that the ego vehicle and target vehicle are moving at a constant acceleration and uses kinematic equations to calculate the TTC of the two vehicles. However, both methods ignore the external shape of the ego vehicle and the road target, as well as their lateral motion, and therefore cannot be used for safety functions.

[0004] Chinese patent CN202010286536.X discloses a vehicle collision prediction method and device, relating to the field of vehicle networking technology. Because the probability of a vehicle accident is determined when spatial distance and time conflict simultaneously, the collision probability determined by the method is more accurate than that of vehicle collision prediction methods in the prior art. The method comprises: first, the vehicle collision prediction device obtains first parameters of the current vehicle, wherein the first parameters include at least: the current vehicle's first speed, first acceleration, first direction angle, first position, first length, and first width. Then, the vehicle collision prediction device obtains second parameters of the first vehicle, wherein the second parameters include at least: the first vehicle's second speed, second acceleration, second direction angle, second position, second length, and second width. Finally, the vehicle collision prediction device determines the probability of an accident between the current vehicle and the first vehicle based on the obtained first and second parameters. This solution only considers acceleration and direction, and does not consider the impact of vehicle shape and lateral movement on collision. Therefore, the accuracy of TTC prediction needs to be improved.

[0005] Chinese patent CN202210515123.3 discloses a vehicle collision prediction method, device and equipment, the method comprising: obtaining first driving data of a first target vehicle and a planned driving path and second driving data of a second target vehicle around the first target vehicle, the second target vehicle traveling based on the planned driving path; determining the collision risk position range corresponding to the first target vehicle based on the first driving data and the second driving data; determining the collision risk information between the first target vehicle and the second target vehicle based on the planned driving path and the collision risk position range; the present application obtains the planned driving path of the second target vehicle, and determines the collision risk position range of the first target vehicle based on the driving data of the first target vehicle and the second target vehicle, and performs collision prediction based on the planned driving path and the collision risk position range; it can effectively improve the accuracy and efficiency of collision prediction. It determines the collision position range through the driving data of two target vehicles, and does not consider the influence of vehicle shape and lateral movement on collision. The accuracy of TTC prediction needs to be improved.

[0006] Chinese patent CN201911398952.2 discloses a method and device for predicting collisions of autonomous vehicles. This method treats both obstacles and the vehicle as polygons with raised two-dimensional surfaces. Based on the direction of movement of the obstacle and the direction of movement of the vehicle, it determines some points on the edge of the polygon where collisions are likely to occur, and performs corresponding obstacle prediction path planning and vehicle prediction path planning. When it is determined that each obstacle prediction path intersects with the predicted path of each vehicle, it determines whether the vehicle and the obstacle will collide based on the distance traveled by the vehicle, the set safety distance, and the minimum distance. It can then calculate the time of collision and accurately determine the specific location of the collision of the vehicle body, making the prediction of whether the vehicle and the obstacle will collide more accurately. Moreover, this method is applicable to structured roads such as straight roads, curved roads, and intersections, as well as unstructured roads, without the need for differentiated processing. This solution only considers speed and direction, and does not consider the impact of vehicle shape and lateral movement on collisions. The accuracy of TTC prediction needs to be improved. Summary of the Invention

[0007] The Summary of the Invention introduces a series of simplified concepts, which are simplifications of existing technologies in the field and are further described in detail in the Detailed Description of the Invention. The Summary of the Invention is not intended to define the key features and essential features of the claimed technical solution, nor is it intended to determine the scope of protection of the claimed technical solution.

[0008] The technical problem to be solved by the present invention is to provide a vehicle collision time calculation method, system and storage medium that can improve the TTC calculation accuracy of the closest point without increasing the cost.

[0009] To address the above technical issues, the present invention provides a vehicle collision time calculation method that takes into account the shapes of the ego vehicle and the target when calculating the TTC between the ego vehicle and the target within the prediction time. The projection method is used to calculate the closest point distance between the ego vehicle vertex and the target vertex (using the length and width of the ego vehicle and the target input by the perception system, with the vehicle shape replaced by a rectangle) in the current prediction step (the entire prediction time is 4 seconds). To improve computational efficiency, a method combining coarse-step prediction and fine-step prediction is used, including the following steps:

[0010] S1, obtain the coordinates of the four vertices of the equivalent rectangle of the self-vehicle and the target vehicle at the current moment;

[0011] S2, traverse each vertex of the ego vehicle and the target vehicle and perform inner product calculation to determine whether there is a separating axis. If there is a separating axis, calculate the TTC. If there is no separating axis, output the current TTC and the minimum distance is 0;

[0012] S3: Calculate the vertex distance between the ego vehicle and the target vehicle at the current moment, traverse all vertices, select a pair of closest vertices, and then traverse all moments with the first specified time length as the first step length to obtain the coarse-step prediction of the TTC of the closest point;

[0013] S4, with the nearest point TTC predicted by the coarse step as the center, take the first step length interval forward and backward respectively. In the selected interval, assume that the ego vehicle is stationary and the target vehicle moves at a constant speed relative to the target vehicle. The fine step prediction calculates the minimum distance and TTC;

[0014] S5, add the results of the coarse-step prediction and the precision prediction to obtain the TTC of the closest point between the ego vehicle and the target vehicle.

[0015] Optionally, the vehicle collision time calculation method is further improved, with the first step length ranging from 0.1S to 0.4S.

[0016] Optionally, the vehicle collision time calculation method can be further improved. When implementing step S3, the coordinates of the four vertices of the self-vehicle rectangle and the target vehicle rectangle are calculated respectively, and each vertex of one rectangle is projected onto the edge of the other rectangle so that point P (p x , p y ) onto another line g is expressed as:

[0017] Pg=(q x ,q y )+λ(m x , m y );

[0018] (m x , m y ) is the direction vector, (q x ,q y ) is the projection point on the line Traverse the entire prediction time and get the TTC of the nearest point of the coarse step prediction. is the direction vector of point P on line g.

[0019] Optionally, the vehicle collision time calculation method is further improved, with a prediction step range of 0.01s to 0.09s, preferably a prediction step of 0.05s, and the minimum distance equation is as follows:

[0020]

[0021] Where (x i ,y i ) is the target vertex, (m x , m y ) is the direction of movement, (x j ,y j ) is a vertex on the ego-vehicle matrix, then the time at which the minimum value of the distance equation is:

[0022]

[0023] To solve the above technical problems, the present invention provides a computer-readable storage medium, which stores a computer program therein. When the computer program is executed, it is used to implement the steps in any one of the above-mentioned vehicle collision time calculation methods.

[0024] To solve the above technical problems, the present invention provides a vehicle collision time calculation system, which can be implemented based on existing vehicle hardware and computer programming technology, including:

[0025] An interface module, which obtains the coordinates of the four vertices of the equivalent rectangle of the ego vehicle and the target vehicle at the current moment from the ego vehicle's onboard system;

[0026] The judgment module traverses each vertex of the ego vehicle and the target vehicle and performs inner product calculation to determine whether there is a separating axis. If there is a separating axis, the TTC is calculated. If there is no separating axis, the current TTC is output and the minimum distance is 0.

[0027] The coarse-step prediction module calculates the vertex distance between the ego vehicle and the target vehicle at the current moment, traverses all vertices, selects a pair of closest vertices, and then traverses all moments with the first specified time length as the first step length to obtain the coarse-step prediction TTC of the closest point;

[0028] The fine-step prediction module takes the first step length interval forward and backward, centered on the nearest point TTC predicted by the coarse-step prediction. Within the selected interval, it assumes that the ego vehicle is stationary and the target vehicle is moving at a constant relative speed. The fine-step prediction calculates the minimum distance and TTC.

[0029] The output module adds the results of the coarse-step prediction and the fine-step prediction to obtain the TTC of the closest point between the ego vehicle and the target vehicle.

[0030] Optionally, the vehicle collision time calculation system may be further improved so that the first step length ranges from 0.1S to 0.4S.

[0031] Optionally, the vehicle collision time calculation system is further improved, and the coarse step prediction module obtains the TTC of the coarse step prediction closest point in the following manner;

[0032] Calculate the coordinates of the four vertices of the self-vehicle rectangle and the target vehicle rectangle respectively, and project each vertex of one rectangle onto the edge of the other rectangle so that point P(p x , p y ) onto another line g is expressed as:

[0033] Pg=(q x ,q y )+λ(m x , m y );

[0034] (m x , m y ) is the direction vector, (q x ,q y ) is the projection point on the line,

[0035] Traverse the entire prediction time and get the TTC of the nearest point of the coarse step prediction. is the direction vector of point P on line g.

[0036] Optionally, the vehicle collision time calculation system is further improved, and the precision prediction calculation module obtains the minimum distance and TTC in the following manner;

[0037] The prediction step range is 0.01s to 0.09s, and the preferred prediction step is 0.05s. The minimum distance equation is as follows:

[0038]

[0039] Where (x i ,y i ) is the target vertex, (m x , m y ) is the direction of movement, (x j ,yj ) is a vertex on the ego-vehicle matrix, then the time at which the minimum value of the distance equation is:

[0040]

[0041] The present invention takes into account the vehicle shape and the lateral and longitudinal motions when calculating TTC, and applies the separating axis theorem. By combining coarse-step prediction with fine-step prediction, computing costs are saved, thereby improving the accuracy of TTC calculation in a low-load controller. Through testing and verification with a large amount of road data and comparison with traditional calculation results, the present invention can significantly improve the calculation accuracy of TTC. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] The drawings herein are intended to illustrate the general characteristics of methods, structures, and / or materials used in certain exemplary embodiments of the present invention, supplementing the descriptions in the specification. However, the drawings herein are schematic diagrams not drawn to scale and may not accurately reflect the precise structure or performance characteristics of any given embodiment. The drawings herein should not be interpreted as defining or limiting the range of values ​​or properties encompassed by the exemplary embodiments of the present invention. The present invention is further described in detail below in conjunction with the drawings and specific embodiments:

[0043] Figure 1 It is a schematic flow chart of the present invention.

[0044] Figure 2 This is a schematic diagram of collision prediction between the ego vehicle and the target vehicle. In the figure, α represents the angle between the vector corresponding to an edge of the object polygon and the normal vector of an edge of the host polygon.

[0045] Figure 3 It is the logic flow chart of the separating axis theorem.

[0046] Figure 4 It is a schematic diagram of the coarse-step prediction and fine-step prediction of the nearest point.

[0047] Figure 5 It is a schematic diagram of the minimum distance between the vehicle and the target vehicle at a constant speed.

[0048] The following describes the embodiments of the present invention through specific embodiments. Those skilled in the art can fully understand the other advantages and technical effects of the present invention from the contents disclosed in this specification. The present invention can also be implemented or applied through different specific embodiments, and the details in this specification can also be applied based on different viewpoints, and various modifications or changes can be made without departing from the overall design concept of the invention. It should be noted that, in the absence of conflict, the following embodiments and the features therein can be combined with each other. The following exemplary embodiments of the present invention can be implemented in a variety of different forms and should not be interpreted as being limited to the specific embodiments described herein. It should be understood that these embodiments are provided to make the disclosure of the present invention thorough and complete, and to fully convey the technical solutions of these exemplary embodiments to those skilled in the art. It should be understood that when an element is referred to as being "connected" or "coupled" to another element, the element can be directly connected or coupled to the other element, or there can be an intermediate element. In contrast, when an element is referred to as being "directly connected" or "directly coupled" to another element, there is no intermediate element. In all figures, the same figure numbers always represent the same elements.

[0049] First embodiment;

[0050] refer to Figure 1 As shown, the present invention provides a method for calculating vehicle collision time, comprising the following steps:

[0051] S1, obtaining the coordinates of the four vertices of the equivalent rectangle of the ego vehicle and the target vehicle at the current moment; illustratively, using the position information of the ego vehicle and target vehicle output by the onboard trajectory prediction module and the dimensional information of the ego vehicle and target vehicle input by the onboard perception system, assuming that the ego vehicle and target vehicle are both rectangular, the coordinates of the four vertices of the ego vehicle and target vehicle at the current moment are calculated respectively;

[0052] S2, traverse each vertex of the ego vehicle and the target vehicle and perform inner product calculation to determine whether there is a separating axis. If there is a separating axis, calculate the TTC. If there is no separating axis, output the current TTC and the minimum distance is 0;

[0053] The separating axis theorem is a collision detection technique for convex polygons. To conserve computational resources, before calculating the time-to-travel (TTC) for each target vehicle and the ego vehicle, this method is run to detect collisions between the ego vehicle and the target. If a collision is detected, the target is declared a dangerous object and its TTC is set to 0, eliminating the need for precise TTC calculations. If no collision is detected, the TTC for the ego vehicle and the target is further precisely calculated.

[0054] refer to Figure 2 、 Figure 3As shown, the separating axis theorem states that if there is any directional axis onto which two convex polygons (the ego vehicle and the target vehicle) are projected, and there is a gap between the projections of the two convex polygons, then the two convex polygons are considered non-colliding. However, in a program, it is unrealistic to traverse all angles. For two 2D objects, the potential candidate separating axes are the normals of each edge of the two objects. Therefore, each edge of the two objects is traversed separately, and then both objects are projected onto the normal corresponding to this edge. If there is a gap, the two objects do not collide; otherwise, the traversal continues. If no separating axis is found after the traversal is completed, the two objects have collided.

[0055] S3: Calculate the vertex distance between the ego vehicle and the target vehicle at the current moment, traverse all vertices, select a pair of closest vertices, and then traverse all moments with the first specified time length as the first step length to obtain the coarse-step prediction of the TTC of the closest point;

[0056] Exemplary, reference Figure 4 As shown in the figure, the step length of the coarse step prediction is 0.2s. In each prediction step, the coordinates of the four vertices of the ego-car rectangle and the target rectangle are calculated respectively, and each vertex of a rectangle is projected onto the edge of the other rectangle. A point P(p x , p y ) onto another line g:

[0057] Pg=(q x ,q y )+λ(m x , m y );

[0058] (m x , m y ) is the direction vector, (q x ,q y ) is the projection point on the line, is the direction vector of point P on line g;

[0059] Therefore, a pair of closest points of the current prediction step can be obtained. After traversing the entire prediction time, the minimum distance point of the coarse step prediction is obtained, which is assumed to be at the prediction time of 1.6s.

[0060] S4, with the nearest point TTC predicted by the coarse step as the center, take the first step length interval forward and backward respectively. In the selected interval, assume that the ego vehicle is stationary and the target vehicle moves at a constant speed relative to the target vehicle. The fine step prediction calculates the minimum distance and TTC;

[0061] According to the calculation result of the coarse step prediction, 1.4s-1.8s is selected as the area of the fine step prediction, the prediction step range is 0.01s-0.09s, preferably the prediction step is 0.05s, and it is assumed that the ego vehicle is in a stationary state and the target moves at a constant speed. Figure 5 As shown in the following formula:

[0062]

[0063] Where (x i , y i ) is the target vertex, (m x , m y ) is the motion direction, (x j , y j ) is a vertex on the ego vehicle matrix, and the time at which the minimum value of the distance equation is located is:

[0064]

[0065] S5, the results of the coarse step prediction and the fine step prediction are added to obtain the TTC of the closest point between the ego vehicle and the target vehicle.

[0066] Optionally, the first step range is 0.1s-0.4s.

[0067] Second embodiment

[0068] The application provides a computer readable storage medium, which internally stores a computer program, and the computer program is executed to implement the steps in the vehicle collision time calculation method of the above-mentioned embodiments.

[0069] The computer readable medium includes permanent and non-permanent, removable and non-removable media, and can be realized by any method or technology to store information. The information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, compact disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical storage, magnetic cassette tape, magnetic tape disk storage or other magnetic storage devices, or any other non-transmission medium that can be used to store information that can be accessed by a computing device. According to the definition herein, the computer readable medium does not include non-transitory computer readable media (transitory media) such as modulated data signals and carriers.

[0070] Third embodiment

[0071] The present invention provides a vehicle collision time calculation system, comprising:

[0072] The interface module calculates the coordinates of the four vertices of the ego vehicle and target vehicle at the current moment, assuming that both the ego vehicle and target vehicle are rectangular, based on the position information of the ego vehicle and target vehicle output by the trajectory prediction module and the dimensional information of the ego vehicle and target vehicle input by the perception system.

[0073] The judgment module traverses each vertex of the ego vehicle and the target vehicle and performs inner product calculation to determine whether there is a separating axis. If there is a separating axis, the TTC is calculated. If there is no separating axis, the current TTC is output and the minimum distance is 0.

[0074] The coarse-step prediction module calculates the vertex distance between the ego vehicle and the target vehicle at the current moment, traverses all vertices, selects a pair of closest vertices, and then traverses all moments with the first specified time length as the first step length to obtain the coarse-step prediction TTC of the closest point, including:

[0075] Calculate the coordinates of the four vertices of the self-vehicle rectangle and the target vehicle rectangle respectively, and project each vertex of one rectangle onto the edge of the other rectangle so that point P(p x , p y ) onto another line g is expressed as:

[0076] Pg=(q x ,q y )+λ(m x , m y )

[0077] (m x , m y ) is the direction vector, (q x ,q y ) is the projection point on the line,

[0078] Traverse the entire prediction time and get the TTC of the nearest point of the coarse step prediction. is the direction vector of point P on line g;

[0079] The fine-step prediction module takes the first step forward and backward, taking the nearest point TTC predicted by the coarse-step prediction as the center. Within the selected interval, it is assumed that the ego vehicle is stationary and the target vehicle is moving at a constant relative speed. The fine-step prediction calculates the minimum distance and TTC, including: the prediction step range is 0.01s to 0.09s, and the preferred prediction step is 0.05s. The minimum distance equation is as follows:

[0080]

[0081] Where (xi ,y i ) is the target vertex, (m x , m y ) is the direction of movement, (x j ,y j ) is a vertex on the ego-vehicle matrix, then the time at which the minimum value of the distance equation is:

[0082]

[0083] The output module adds the results of the coarse-step prediction and the fine-step prediction to obtain the TTC of the closest point between the ego vehicle and the target vehicle.

[0084] Optional, the first step length range is 0.1S ~ 0.4S.

[0085] Unless otherwise defined, all terms used herein (including technical and scientific terms) have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. It will also be understood that, unless expressly defined herein, terms such as those defined in commonly used dictionaries should be interpreted as having a meaning consistent with their meaning in the context of the relevant art, rather than being interpreted in an idealized or overly formal sense.

[0086] The present invention has been described in detail above by way of specific embodiments and examples, but these do not constitute limitations of the present invention. Without departing from the principles of the present invention, those skilled in the art may make many variations and improvements, which should also be considered within the scope of protection of the present invention.

Claims

1. A method for calculating vehicle collision time, characterized in that: The following steps are involved: S1, obtain the coordinates of the four vertices of the equivalent rectangle of the self-vehicle and the target vehicle at the current moment; S2, traverse each vertex of the ego vehicle and the target vehicle and perform inner product calculation to determine whether there is a separating axis. If there is a separating axis, calculate the TTC. If there is no separating axis, output the current TTC and the minimum distance is 0; S3 calculates the vertex distance between the ego vehicle and the target vehicle at the current moment, traverses all vertices, selects a pair of closest vertices, and then traverses all moments with the first specified time length as the first step length to obtain the coarse step prediction TTC of the closest point, including: Calculate the coordinates of the four vertices of the self-vehicle rectangle and the target vehicle rectangle respectively, and project each vertex of one rectangle onto the edge of the other rectangle so that the points On another line The orthogonal projection on is expressed as: ; is the direction vector, is the projection point on the line, ; Traverse the entire prediction time and get the TTC of the nearest point of the coarse step prediction. is the direction vector of point P on line g; S4, with the nearest point TTC predicted by the coarse step as the center, takes the first step length interval forward and backward respectively. In the selected interval, it is assumed that the ego vehicle is stationary and the target vehicle is moving at a constant relative speed. The fine step prediction calculates the minimum distance and TTC, including: the prediction step range is 0.01s~0.09s, and the minimum distance equation is as follows: ;in is the target vertex, is the direction of movement, is a vertex on the ego-vehicle matrix, then the time at which the minimum value of the distance equation is: ; S5, add the results of the coarse-step prediction and the precision prediction to obtain the TTC of the closest point between the ego vehicle and the target vehicle.

2. The vehicle collision time calculation method according to claim 1, wherein: The first step length ranges from 0.1S to 0.4S.

3. A computer-readable storage medium, characterized in that: A computer program is stored therein, and when the computer program is executed, it is used to implement the steps in the vehicle collision time calculation method according to any one of claims 1 or 2.

4. A vehicle collision time calculation system, characterized in that: include: An interface module, which obtains the coordinates of the four vertices of the equivalent rectangle of the ego vehicle and the target vehicle at the current moment from the ego vehicle's onboard system; The judgment module traverses each vertex of the ego vehicle and the target vehicle and performs inner product calculation to determine whether there is a separating axis. If there is a separating axis, the TTC is calculated. If there is no separating axis, the current TTC is output and the minimum distance is 0. The coarse-step prediction module calculates the vertex distance between the ego vehicle and the target vehicle at the current moment, traverses all vertices, selects a pair of closest vertices, and then traverses all moments with the first specified time length as the first step length to obtain the coarse-step prediction TTC of the closest point, including: Calculate the coordinates of the four vertices of the self-vehicle rectangle and the target vehicle rectangle respectively, and project each vertex of one rectangle onto the edge of the other rectangle so that the points On another line The orthogonal projection on is expressed as: ; is the direction vector, is the projection point on the line, ; After the entire prediction time, the TTC of the nearest point of the coarse step prediction is obtained. is the direction vector of point P on line g; The fine-step prediction module takes the first step forward and backward from the coarse-step predicted nearest point (TTC) as the center. Within the selected interval, it assumes that the ego vehicle is stationary and the target vehicle is moving at a constant relative speed. The fine-step prediction module calculates the minimum distance and TTC, including: The prediction step range is 0.01S~0.09s, and the minimum distance equation is as follows: ;in is the target vertex, is the direction of movement, is a vertex on the ego-vehicle matrix, then the time at which the minimum value of the distance equation is: ; The output module adds the results of the coarse-step prediction and the fine-step prediction to obtain the TTC of the closest point between the ego vehicle and the target vehicle.

5. The vehicle collision time calculation system according to claim 4, wherein: The first step length ranges from 0.1S to 0.4S.

Images