A vehicle, a target threat estimation method thereof, and a computer storage medium

Abstract

The application discloses a vehicle and a target threat estimation method and a computer storage medium thereof. The method comprises the following steps: acquiring a brake delay time; acquiring main motion delay information of a host vehicle after the brake delay time and target motion delay information of a target vehicle after the brake delay time; predicting main motion prediction information of the host vehicle based on the main motion delay information; predicting target motion prediction information of the target vehicle based on the target motion delay information; and acquiring lateral threat estimation information and / or longitudinal threat estimation information of the host vehicle and the target vehicle according to the main motion prediction information and the target motion prediction information. The vehicle of the application extracts motion information of the target vehicle and motion information of the host vehicle, does not depend on other unstable external environment information, introduces the idea of prediction, respectively calculates threat evaluation indexes of the target vehicle and the host vehicle in the lateral direction and the longitudinal direction, and estimates the dangerous degree of the target vehicle to the host vehicle through the above dynamic process parameters.

Description

Technical Field

[0001] This application relates to the field of road safety technology, and in particular to a method for estimating the threat of vehicles and their targets, and a computer storage medium. Background Technology

[0002] As traffic hubs, intersections are a crucial component of urban road traffic systems, enriching traffic content and improving the efficiency of traffic participants. Due to the frequent interactions between vehicles and pedestrians, and between vehicles themselves (including motorized and non-motorized vehicles), within the limited area of ​​intersections, accident situations are highly complex, and traffic accidents occur frequently. This is especially true at unsignalized intersections, where there is no traffic light control, and vehicle speeds are relatively higher than at signalized intersections. If drivers cannot react quickly enough to pedestrians suddenly crossing the intersection, the risk of an accident is extremely high. Statistics show that in recent years, a significant portion of all traffic accidents worldwide have occurred at intersections. This is not only a major current problem but is also predicted to account for an even larger share of road traffic accidents in the future.

[0003] However, current vehicle threat estimation algorithms rely heavily on lane markings to make hazard assessments, making them unsuitable for scenarios like intersections or where lane markings are missing. Furthermore, some algorithms utilize time-to-collision (TTC) and time-of-headway (THW). However, simple hazard estimation metrics cannot dynamically characterize the relative motion between the vehicle and the target vehicle; dynamic process parameters must be extracted to estimate the degree of hazard. Summary of the Invention

[0004] This application provides a method for estimating the threat of a vehicle and its target, and a computer storage medium.

[0005] One technical solution adopted in this application is to provide a target threat estimation method based on a crossroads, the target threat estimation method comprising:

[0006] Obtain the braking delay time;

[0007] Acquire the main motion delay information of the main vehicle after the braking delay time, and the target motion delay information of the target vehicle after the braking delay time;

[0008] Based on the main motion delay information, predict the main motion prediction information of the main vehicle;

[0009] Based on the target motion delay information, predict the target motion prediction information of the target vehicle;

[0010] Based on the main motion prediction information and the target motion prediction information, obtain the lateral threat estimation information and / or longitudinal threat estimation information of the main vehicle and the target vehicle.

[0011] The lateral threat estimation information includes the rate of change of the curvature of the main vehicle;

[0012] The step of obtaining lateral threat estimation information and / or longitudinal threat estimation information of the master vehicle and the target vehicle according to the master motion prediction information and the target motion prediction information includes:

[0013] Based on the main motion prediction information and the target motion prediction information, obtain the two nearest sides and the diagonal sides between the target vehicle and the main vehicle;

[0014] Using the two nearest edges and the diagonal edge, determine the corner coordinates of the target vehicle;

[0015] Based on the corner coordinate data, the approximate collision point of the main vehicle is obtained;

[0016] Based on the approximate collision point, the rate of change of curvature of the main vehicle used to steer and avoid collision is determined.

[0017] The step of obtaining the two closest edges between the target vehicle and the host vehicle includes:

[0018] Obtain the midpoints of each edge of the target vehicle;

[0019] Obtain the distance between the midpoint of each edge and the center of gravity of the main vehicle;

[0020] The two closest edges are selected as the two closest edges between the target vehicle and the host vehicle.

[0021] The step of obtaining the midpoints of each edge of the target vehicle includes:

[0022] Add a preset compensation amount to the length and width of the target vehicle, and obtain the centroid coordinates and length and width projection values ​​of the target vehicle after adding the preset compensation amount;

[0023] Based on the centroid coordinates and the length and width projections, determine the corner points of each side of the target vehicle;

[0024] Obtain the midpoints of each edge of the target vehicle based on the corner points of each edge.

[0025] The target threat estimation method further includes:

[0026] Obtain the maximum rate of change of lateral acceleration;

[0027] Based on the curvature change rate of the main vehicle, the current lateral acceleration change rate is obtained;

[0028] When the current rate of change of lateral acceleration is greater than the maximum rate of change of lateral acceleration, automatic emergency steering or vehicle takeover is activated.

[0029] The longitudinal threat estimation information includes acceleration values;

[0030] The step of obtaining lateral threat estimation information and / or longitudinal threat estimation information of the master vehicle and the target vehicle according to the master motion prediction information and the target motion prediction information includes:

[0031] Obtain the nearest longitudinal distance between the master vehicle and the target vehicle at each moment within the predicted time period;

[0032] Based on the nearest longitudinal distance at each time moment, the acceleration value at each time moment is obtained;

[0033] Compare the acceleration value at each moment with the acceleration values ​​at historical moments;

[0034] If the difference between the acceleration value at each moment and the acceleration value at a historical moment is greater than 0, the acceleration value at the current moment is retained.

[0035] When the difference between the acceleration value at each moment and the acceleration value at a historical moment is less than 0, the acceleration value at the current moment is replaced by the acceleration value at the historical moment.

[0036] Based on the acceleration value at each moment, longitudinal threat estimation information of the host vehicle and the target vehicle is obtained.

[0037] The target threat estimation method further includes:

[0038] When the acceleration value is greater than the acceleration value or less than the braking value, automatic emergency steering or vehicle takeover is activated.

[0039] Wherein, the step of performing automatic emergency steering or activating takeover vehicle when the acceleration value is greater than the acceleration value or less than the braking value includes:

[0040] The rate of change of acceleration is obtained based on the acceleration value;

[0041] When the acceleration value is greater than the acceleration acceleration value, or when the rate of change of acceleration is greater than the maximum rate of change of longitudinal acceleration acceleration, automatic emergency steering or vehicle takeover is activated.

[0042] Alternatively, if the acceleration value is less than the braking acceleration value, or the rate of change of acceleration is less than the maximum rate of change of longitudinal braking acceleration, automatic emergency steering or vehicle takeover may be performed.

[0043] Another technical solution adopted in this application is to provide a vehicle, the vehicle including a memory and a processor coupled to the memory;

[0044] The memory is used to store program data, and the processor is used to execute the program data to implement the target threat estimation method described above.

[0045] Another technical solution adopted in this application is to provide a computer storage medium for storing program data, which, when executed by a computer, is used to implement the target threat estimation method described above.

[0046] The beneficial effects of this application are: the vehicle acquires braking delay time; acquires the main motion delay information of the host vehicle after the braking delay time, and the target motion delay information of the target vehicle after the braking delay time; based on the main motion delay information, predicts the main motion prediction information of the host vehicle; based on the target motion delay information, predicts the target motion prediction information of the target vehicle; and according to the main motion prediction information and the target motion prediction information, acquires lateral threat estimation information and / or longitudinal threat estimation information of the host vehicle and the target vehicle. This application's vehicle, by extracting the motion information of the target vehicle and the host vehicle, without relying on other unstable external environmental information, introduces the concept of prediction, calculates the threat evaluation indicators of the target vehicle and the host vehicle in the lateral and longitudinal directions respectively, and estimates the degree of danger of the target vehicle to the host vehicle through the above dynamic process parameters. Attached Figure Description

[0047] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0048] Figure 1 This is a schematic diagram of acceleration, steering, and braking as potential vehicle collision avoidance actions provided in this application;

[0049] Figure 2 This is a flowchart illustrating the first embodiment of the target threat estimation method provided in this application;

[0050] Figure 3 This is a flowchart of the overall algorithm for the target threat estimation method provided in this application;

[0051] Figure 4 The target threat estimation method provided in this application is for the target vehicle at T d A schematic diagram of linear motion at time t;

[0052] Figure 5 The target threat estimation method provided in this application is for the target vehicle at T d A schematic diagram of an embodiment of circular motion at a given moment;

[0053] Figure 6 The target threat estimation method provided in this application is for the target vehicle at T d A schematic diagram of another embodiment of the circular motion at a given moment;

[0054] Figure 7 This is a schematic diagram of the predicted motion trajectory of the main vehicle using the Bezier curve provided in this application;

[0055] Figure 8 This is a flowchart illustrating a second embodiment of the target threat estimation method provided in this application;

[0056] Figure 9 This is a schematic diagram of the nearest edge at time TTR in the target threat estimation method provided in this application;

[0057] Figure 10 This is a schematic diagram of the nearest edge calculation process at the TTR time in the target threat estimation method provided in this application;

[0058] Figure 11 X is the target threat estimation method provided in this application. fl / Y fl The calculation process;

[0059] Figure 12 This is a schematic diagram of the steering and collision avoidance process in the target threat estimation method provided in this application;

[0060] Figure 13 The target threat estimation method provided in this application is for the driver at t i A diagram showing how to turn at all times to avoid the edges (x1, y1) and (x2, y2);

[0061] Figure 14 This is a schematic diagram of the turning behavior parameters in the target threat estimation method provided in this application;

[0062] Figure 15 This is a flowchart illustrating the third embodiment of the target threat estimation method provided in this application;

[0063] Figure 16 This is a schematic diagram of the target coordinate points required for the longitudinal InPath in the target threat estimation method provided in this application;

[0064] Figure 17 This is a schematic diagram of an embodiment of the lateral relative positional relationship between the target vehicle and the host vehicle in the target threat estimation method provided in this application;

[0065] Figure 18 This is a schematic diagram of another embodiment of the lateral relative positional relationship between the target vehicle and the host vehicle in the target threat estimation method provided in this application;

[0066] Figure 19 This is a schematic diagram of the longitudinal InPath judgment process in the target threat estimation method provided in this application;

[0067] Figure 20 This is a flowchart illustrating the longitudinal (braking, acceleration) collision avoidance parameterization calculation provided in this application;

[0068] Figure 21 This is a schematic diagram of the turning behavior parameters in the target threat estimation method provided in this application;

[0069] Figure 22 This is a parameterized schematic diagram of braking and acceleration in the target threat estimation method provided in this application;

[0070] Figure 23 This is a schematic diagram of the structure of an embodiment of the vehicle provided in this application;

[0071] Figure 24 This is a schematic diagram of the structure of an embodiment of the computer storage medium provided in this application. Detailed Implementation

[0072] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of the embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this application.

[0073] To address the aforementioned problems, the purpose of this application is twofold: firstly, to propose a hazard estimation algorithm that covers the use scenario of intersections, extracting as much target and vehicle motion information as possible without relying on other unstable external environmental information; and secondly, to introduce the concept of prediction, calculating the threat evaluation indicators of the target and vehicle in the lateral and longitudinal directions respectively.

[0074] Please see Figure 1 , Figure 1 This is a schematic diagram of acceleration, steering, and braking as potential vehicle collision avoidance actions provided in this application.

[0075] Most general emergency collision avoidance threat estimation algorithms focus on obstacles traveling in the same direction. For intersection scenarios, threat estimation for obstacles crossing the road is not compatible. In intersection scenarios, drivers may have the following concerns: Figure 1The proposed algorithm evaluates how a vehicle can avoid collisions with other objects within a limited prediction range at an intersection by steering, braking, or accelerating. First, the outer contour of the main vehicle is represented by a rectangle, and the outer contour of the estimated obstacle target is described by a rectangle representing its size, position, and orientation. The motion of the main vehicle is described using a linear bicycle model, and the obstacle target's motion is predicted using either a constant acceleration model or a constant rate of curvature change model based on the magnitude of its curvature. When evaluating how the driver can maneuver to avoid a collision between the main vehicle and the target vehicle, vehicle dynamics and the driver's preferred ability to take over the vehicle are taken into account. The potential driver takeover ability is parameterized, and an analytical expression is approximately derived to estimate the set of analytical solutions for the driver to avoid collisions through steering, braking, and acceleration.

[0076] By solving the collision avoidance analysis set of steering, braking, and acceleration behaviors, the potential threat level of a vehicle colliding with an obstacle at the current moment can be estimated. Then, based on the steering demand performance index, braking demand performance index, and acceleration demand performance index, it can be determined whether the intervention of the auxiliary system is needed to avoid a traffic accident.

[0077] Please see details. Figure 2 and Figure 3 , Figure 2 This is a flowchart illustrating the first embodiment of the target threat estimation method provided in this application. Figure 3 This is a flowchart of the overall algorithm for the target threat estimation method provided in this application.

[0078] like Figure 2 As shown, the target threat estimation method in this application embodiment may specifically include the following steps:

[0079] Step S11: Obtain the braking delay time.

[0080] In this embodiment of the application, before the vehicle estimates the threat, the execution time of the actuator is considered, and the actuator delay time, i.e. the vehicle braking delay time, is obtained by linear interpolation based on the current acceleration value of the vehicle.

[0081] Specifically, the actuator delay time includes: [0.01 0.01 0.01 0.03 0.05 0.1 0.15 0.15 0.07 0.05], and the vehicle acceleration interpolation table is: [-5 -4 -3 -2 -1 -0.5 -0.3 0.3 1 3].

[0082] Step S12: Obtain the main motion delay information of the main vehicle after the braking delay time, and the target motion delay information of the target vehicle after the braking delay time.

[0083] In this embodiment, the vehicle predicts the target vehicle, i.e., the motion delay information of the target vehicle after the braking delay time. The motion information is based on the target vehicle's motion, such as linear motion or circular motion, specifically including but not limited to linear motion information and circular motion information.

[0084] Linear motion process as follows Figure 4 As shown, Figure 4 The target threat estimation method provided in this application is for the target vehicle at T d A schematic diagram of linear motion at any given time. The circular motion process is as follows: Figure 5 and Figure 6 As shown, Figure 5 The target threat estimation method provided in this application is for the target vehicle at T d A schematic diagram of an embodiment of circular motion at a given moment. Figure 6 The target threat estimation method provided in this application is for the target vehicle at T d A schematic diagram of another embodiment of the circular motion at a given moment.

[0085] In the embodiments of this application, Figure 4 Linear motion can also include longitudinal motion calculation and lateral motion calculation. The formula for calculating lateral motion is as follows:

[0086]

[0087] v xpre =v x +a x t

[0088] The formula for calculating longitudinal motion is as follows:

[0089]

[0090] v ypre =v y +a y t

[0091] Where t is the prediction time, taking into account the target vehicle's stopping time based on the current moment and the braking delay time T. d Take the minimum value.

[0092] The above formula yields the target vehicle's linear motion prediction information set, including: {x} pre y pre v xpre v ypre a xpre a ypre}

[0093] Where x, y, v x v y ax a y Given the target vehicle's position, speed, and acceleration information at the current moment, x pre y pre v xpre v ypre a xpre a ypre Braking delay T d The target vehicle's center of gravity position, speed, and acceleration information.

[0094] like Figure 5 and Figure 6 As shown, the main solutions for circular arc motion include: calculating the central angle, obtaining the coordinates of the rotation center based on the curvature and direction angle, and finally obtaining the horizontal and vertical coordinates after rotation based on the current coordinates, the coordinates of the rotation center, and the central angle.

[0095] Motion analysis is divided into tangential motion and centripetal motion, as detailed below:

[0096] σ=crvt*L

[0097] In the formula, σ is the central angle, crvt is the curvature of the target motion, and L is the arc length of the approximate curvilinear motion.

[0098] θ'=θ+σ

[0099] Wherein, θ is the target's initial heading angle, and θ' is the target's heading angle at the braking delay time.

[0100] x0 = x + cos(θ + pi / 2)

[0101] y0 = y + sin(θ + pi / 2)

[0102] x pre =x0+(x-x0)*cos(σ)+(y-y0)*sin(σ)

[0103] y pre =y0+(x-x0)*sin(σ)+(y-y0)*cos(σ)

[0104] The above formula yields the target vehicle's circular motion prediction information set, including: {x} pre y pre spd pre a p re,ag pre ,crvt}.

[0105] Where x0, y0 are the target rotation centers, x, y are the target's current centroid positions, and x pre y pre The position of the centroid after the target braking delay, spdpre The velocity of the center of mass after the target is delayed; a pre Let be the acceleration of the target's center of mass after the time delay, and let crvt be the curvature of the target's center of mass after the time delay.

[0106] By using the above method, the vehicle can predict the motion information of the target vehicle and obtain the target motion delay information after the braking delay time.

[0107] Furthermore, in one embodiment of this application, the vehicle obtains the main motion delay information of the host vehicle after the braking delay time using the following formula:

[0108] v0 h =v h +a h t

[0109] c0 h =c h +c1 h t

[0110] h0 h =c0 h *d

[0111]

[0112] Among them, v h Main vehicle speed, a h The acceleration of the main vehicle, c h Main vehicle motion curvature, c1 h Let c be the rate of change of curvature of the main vehicle, and d be the displacement during the delay time. When c > 0 h When the speed exceeds a certain threshold, the vehicle is considered to be moving in a circular arc.

[0113]

[0114] Otherwise, it is linear motion:

[0115]

[0116] The braking delay T of the main vehicle is obtained from the above formula. d The subsequent information set includes {x0} h y0 h h0 h a0 h v0 h c0 h}

[0117] In the above formula, x0 h y0 h Position information after main vehicle braking delay, h0 h Let a0 be the heading angle. h v0h c0 h The acceleration, velocity, and curvature of motion after the braking delay are given.

[0118] Since the above vehicle information is calculated in its own coordinate system, this embodiment also includes a coordinate transformation process, that is, unifying the motion information of the target vehicle into the coordinate system of the master vehicle, as follows:

[0119] After calculating the motion information of the target vehicle and the master vehicle after the braking delay Td, the motion information of the target vehicle is converted into the motion information of the master vehicle (T). d After the position is determined, the linear motion conversion formula is as follows:

[0120]

[0121] In the above formula, x l y l For the target vehicle relative to T d Real-time location information of the main vehicle, WeChat l vy l ,ax l ,ay l Target vehicle relative to T d Real-time movement information of the main vehicle.

[0122] Step S13: Based on the main motion delay information, predict the main motion prediction information of the main vehicle.

[0123] Specifically, in the embodiments of this application, the vehicle uses a braking delay time (T) d Starting from the motion prediction point, the main vehicle predicts its trajectory over 3 seconds using a Bezier curve. A third-order Bezier curve is then used to predict the main vehicle's trajectory. The prediction formula is as follows:

[0124] x = P0(1-t) 3 +3P1(1-t) 2 t+3P2(1-t)t 2 +P3t 3

[0125] Expanding the above formula further, we get:

[0126]

[0127] In the above formula, x and y are the coordinates of the moving point on the trajectory, respectively, and P 0x and P 0y It is the control point at the starting point of the curve, P. 3x and P 3y P1 and P2 are the coordinates of the two intermediate control points, and t varies from 0 to 1s. Figure 7The trajectory shown in the third-order Bezier curve is based on the premise that the horizontal and vertical coordinates of P0 to P3 are known.

[0128] In actual vehicle operation, the current vehicle speed v0 is known. h and acceleration a0 h Knowing the position of P0 at t=0 (i.e., at T0), we know that at t=0:

[0129]

[0130] in, The ordinate value at time T0. The x-coordinate value is given at time T0; knowing the position of P3 at time t = 1s, P3 is the predicted position of the vehicle at 1s.

[0131]

[0132] in, The ordinate value at time T1 The x-coordinate value at time T1;

[0133] Given the current vehicle curvature and rate of change of curvature (C) r The curvature of the vehicle's motion at 1 second can be determined by calculation:

[0134]

[0135] Based on the above formula to determine the position of P3, only vehicles with changes in curvature are calculated for circular motion. For vehicles with no change in curvature, the kinematic formulas are directly applied for solution.

[0136] 1) Solving for the position at 1s

[0137] Both P0 and P3 are positions based on the vehicle's own coordinate system. However, when the vehicle is moving in a circular arc, the position cannot be solved simply using linear motion equations. Because it is curvilinear motion, in 1 second, the vehicle can be considered as moving in a circle, with the radius of curvature of the vehicle as the center, and the vehicle as a point on this circle. This transforms the vehicle's motion into a position (P3) around a certain center of rotation in the future 1 second. The solved position (P3) coordinates are relative to the center of the circle, not the vehicle itself. Therefore, the solved P3 is then geometrically solved using the vehicle as the reference frame to obtain P3.

[0138] Considering the rate of change of curvature, and based on the empirical formula, we take a weight of 0.7 at T0 and a weight of g.3 at T1. Therefore, the final position of P3 is:

[0139]

[0140] In the above formula, C rt0 Let C be the rate of change of curvature at the current moment. rt1 The curvature change rate at 1 second is given by a weight of 0.7, while the weight at 1 second in the future is g.3, indicating greater confidence in the curvature change rate at the current moment, and a more conservative prediction of the vehicle's trajectory.

[0141] 2) Solving for the vehicle's velocity and acceleration in 1 second.

[0142] To determine the vehicle velocity and acceleration at position P3, we first need to solve for P1 and P2, and then apply the formula...

[0143]

[0144] Taking the derivative, i.e., differentiating with respect to x, yields the vehicle speed in physical terms. Similarly, differentiating with respect to y:

[0145] x′=v=3P1-3P0+(6P0-12P1+6P2)t+(9P1-3P0-9P2+3P3)t 2

[0146] Taking t = g, we get

[0147] Taking the derivative with respect to the vehicle speed, we get the acceleration:

[0148]

[0149] Taking t=0, we get

[0150] Having determined the coefficients P1, P2, and P3 (i.e., position information), the vehicle velocity and acceleration at 1 second are calculated using the same method as for the coefficients P, except that the value of t is changed from g to 1, as follows: Taking the derivative of x, we obtain the vehicle velocity at 1 second: x′=v=3P1-3P0+(6P0-12P1+6P2)+(9P1-3P0-9P2+3P3)=3P3-3P2 Taking the second derivative of x, we obtain the vehicle acceleration at 1 second: x″=A=6P1-12P2+6P3.

[0151] The velocity and acceleration calculated above are based on a curvilinear coordinate system, and P0 is set to O in the initial position T0. Therefore, the calculated velocity and acceleration are relative to time T0. In order to calculate the velocity and acceleration in the actual vehicle coordinate system, it is necessary to calculate based on the vehicle acceleration and velocity at time T0. The actual vehicle speed and acceleration at time T1 require knowing the heading angle at time T0. At this time, A and V obtained by Bezier play the role of calculating the heading, ensuring the continuity of the curve.

[0152] Based on the above methods, the longitudinal and lateral velocities (V) of the vehicle at time T1, calculated by Bezier, are obtained respectively.Lgt and V Lat Derivation of vehicle heading at time T1:

[0153] Heading T1 =atan2(V Lgt V Lat )

[0154] Calculate the vector velocity at time T1 based on the vehicle's velocity and acceleration kinematic formulas at time T0:

[0155] Speed T1 =Speed T0 +A T0

[0156] A T1 =A T0

[0157] In this embodiment, the rate of change of curvature within 0-1s is obtained by converting the steering wheel angular velocity through a two-degree-of-freedom model. The rate of change of curvature used within 1-2s is inverted with the rate of change of curvature within 0-1s. The rate of change of curvature within 2-3s is forcibly written as 0 and is considered and calculated as a constant curvature.

[0158] The above calculation is a recursive operation, that is, P3 after each solution is P0 for the next scheduling cycle. The motion control points are calculated recursively over 3 seconds, and finally 12 Bessel control points are obtained.

[0159] Step S14: Based on the target motion delay information, predict the target motion prediction information of the target vehicle.

[0160] In this embodiment of the application, the vehicle predicts the trajectory of the target vehicle within 3 seconds, with a time interval of 0.05 seconds. The linear motion is predicted and calculated using a constant acceleration model, and the circular motion is predicted and calculated using a constant curvature model. A 60-dimensional array is used to represent the motion information of each motion point, and the delay information includes the lateral and longitudinal positions and the heading angle information.

[0161] Step S15: Based on the main motion prediction information and the target motion prediction information, obtain the lateral threat estimation information and / or longitudinal threat estimation information of the main vehicle and the target vehicle.

[0162] In this application, the vehicle obtains lateral threat estimation information between the main vehicle and the target vehicle according to the main motion prediction information and the target motion prediction information. The lateral threat estimation information includes the curvature change rate of the main vehicle. The vehicle obtains the two nearest sides and the diagonal sides between the target vehicle and the main vehicle based on the main motion prediction information and the target motion prediction information. Using the two nearest sides and the diagonal sides, the vehicle determines the corner coordinate data of the target vehicle. Based on the corner coordinate data, the vehicle obtains the approximate collision point of the main vehicle. Based on the approximate collision point, the vehicle determines the curvature change rate of the main vehicle when steering to avoid collision.

[0163] Specifically, the vehicle calculates an analytical solution to avoid a collision by steering within a finite prediction time to assess future lateral threats to the target. The main idea is to find a rate of curvature change of the main vehicle at any time t∈3 such that the outer contour of the main vehicle does not overlap with the contour of the target.

[0164] This application is explained in three parts: First, before the collision between the vehicle and the target, the target has at most two edges close to the vehicle, so the closest edges between the vehicle and the target at a certain moment are first calculated. Second, the vehicle predicts how to avoid collisions with the closest edge, the second closest edge, and the diagonal edge at a certain moment. Third, the vehicle estimates how to avoid collisions with the three edges by steering throughout the entire prediction time vector.

[0165] In one embodiment of this application, the lateral threat estimation information includes the rate of change of the curvature of the main vehicle. This application further proposes an embodiment for obtaining lateral threat estimation information; please refer to the details below. Figure 8 , Figure 9 , Figure 10 and Figure 11 , Figure 8 This is a flowchart illustrating a second embodiment of the target threat estimation method provided in this application; Figure 9 This is a schematic diagram of the nearest edge at time TTR in the target threat estimation method provided in this application; Figure 10 This is a schematic diagram of the nearest edge calculation process at the TTR time in the target threat estimation method provided in this application; Figure 11 X is the target threat estimation method provided in this application. fl / Y fl The calculation process.

[0166] like Figure 8 As shown, the specific steps are as follows:

[0167] Step S21: Based on the main motion prediction information and the target motion prediction information, obtain the two closest edges and diagonal edges between the target vehicle and the main vehicle.

[0168] In one embodiment of this application, the vehicle obtains the midpoints of each side of the target vehicle; further, a preset compensation amount is added to the length and width of the target vehicle, and the centroid coordinates and length and width projection values ​​of the target vehicle after adding the preset compensation amount are obtained; based on the centroid coordinates and the length and width projection values, the corner points of each side of the target vehicle are determined; the midpoints of each side of the target vehicle are obtained according to the corner points of each side of the target vehicle, and the distance between the midpoints of each side and the centroid of the main vehicle is obtained; the two closest sides are obtained as the two closest sides between the target vehicle and the main vehicle.

[0169] Specifically, such as Figure 9 As shown, the vehicle finds the two closest sides and the diagonal side of the main vehicle and the target vehicle at the TTR (time to Rollover) time. It also calculates the coordinates of the two closest sides and the diagonal side of the target vehicle at the first 6 sampling points and the last 3 sampling points at the TTR time, forming a time vector with a total duration of 0.5s.

[0170] The vehicle numbers the four sides of the target at the TTR moment clockwise: 0 for front, 1 for right, 2 for rear, and 3 for left. The x-coordinate, y-coordinate, and heading angle of the centroid at the TTR moment are obtained by interpolation from the 60-dimensional (30-second) target trajectory. This information is based on the main vehicle after braking delay. The target's length and width are compensated by a certain amount, and the projection values ​​of half the vehicle width / centroid to front length / centroid to rear length on the X and Y axes are calculated. Based on the X and Y axis projection values ​​and the target centroid coordinates, the coordinates of the four corner points of the target are obtained.

[0171]

[0172]

[0173] Where X and Y are the coordinates of the target centroid; L xf Let L be the distance from the center of mass to the front end projected onto the X-axis, and L be the distance from the center of mass to the rear end projected onto the X-axis. yf L is the distance from the centroid to the front end projected onto the Y-axis. yr W is the distance from the centroid to the rear end projected onto the Y-axis. x W is the projection of half the car width onto the X-axis. y For the projection of half the car width onto the Y-axis, X corner and Y corner For an array sorted clockwise, taking the calculation process of the left front corner as an example, as follows: Figure 11 As shown, the methods for other corner points are similar.

[0174] Step S22: Use the two nearest edges and the diagonal edges to determine the corner coordinates of the target vehicle.

[0175] Specifically, the vehicle calculates the sum of squares of the distances between the midpoints of each side of the target and the centroid of the main vehicle, using the numerical values ​​to represent the proximity of the positions. Finally, it outputs the indices of the two sides closest to the main vehicle, creates a time vector near the TTR time, selects 6 sampling points forward and 3 sampling points backward with the TTR time as the current point, and the total duration is 0.5s. Based on the indices of the two closest sides, it creates diagonal sides to form a closed triangle, calculates a series of coordinate points under the time vector, and uses a 6×10 array to store the corner coordinate data of all sides.

[0176] Step S23: Obtain the approximate collision point of the main vehicle based on the corner coordinate data.

[0177] Specifically, such as Figure 11 and Figure 12 As shown, Figure 12 This is a schematic diagram of the steering and collision avoidance process in the target threat estimation method provided in this application. Based on the current speed and acceleration information of the main vehicle, the stopping time is calculated. The minimum value of this time vector is taken from the target stopping time, and the truncation ends at the target stopping time. Lateral threats beyond the main vehicle's stopping time are not calculated. The lateral and longitudinal coordinates of the main vehicle's center of mass under the TTR time vector are calculated. Based on the nearest edge index obtained in the above steps, the coordinates of the target's nearest edge, second nearest edge, and diagonal edge are selected respectively. The approximate collision point on the main vehicle is then calculated. The horizontal and vertical coordinates are used to obtain the approximate collision point of the main vehicle.

[0178] Step S24: Based on the approximate collision point, determine the rate of change of curvature of the main vehicle when the main vehicle is steering to avoid collision.

[0179] Specifically, the vehicle is based on the approximate collision point, according to the approximate collision point of the main vehicle. The vertical coordinate value is used to obtain the horizontal coordinate values ​​(y1 and y2) of the corner points on the target side using linear interpolation. The calculation method for the upper and lower endpoints is the same. Based on the collision avoidance inequality when the main vehicle turns left: Similarly, for right-hand turns, the curvature change rate of the main vehicle required to avoid a collision when turning left and the curvature change rate of the main vehicle required to avoid a collision when turning right are calculated separately. The calculation is repeated to obtain the set of curvature change rates for the left and the set of curvature change rates for the right. The maximum value is taken for the left and the minimum value is taken for the right.

[0180] In this embodiment of the application, the technical principle of the vehicle avoiding collisions by steering is as follows:

[0181] like Figure 13 As shown, Figure 13 The target threat estimation method provided in this application is for the driver at t i A diagram illustrating how to constantly turn to avoid the edges (x1, y1) and (x2, y2). Figure 13The plus sign is equivalent to +, and the minus sign is equivalent to -.

[0182] This application describes an analytical solution for estimating a vehicle's collision avoidance mechanism by steering within a finite prediction time. The goal of the study is to find c1 at any time ti∈[0,T_max] such that the outer contour of the main vehicle does not overlap with the contours of other objects.

[0183] The explanation is divided into two parts: First, it estimates how the vehicle can avoid colliding with one side of the triangle at time ti; second, it evaluates how to avoid colliding with the triangle by steering throughout the prediction range.

[0184] Collision Avoidance by Steering and Taking an Edge of the Target Vehicle: The goal is to find c1 such that at time ti, steering the vehicle to edge m avoids a collision, taking the left side as an example. c1 represents the avoidance at time ti by the edge m consisting of points [(x1,y1), (x2,y2)]. The driver at time t... i At any time, turn to avoid the edge [(x1, y1), (x2, y2)], if at t i If the entire vehicle is located on the left side of the edge, then the vehicle will pass through the left side of the edge. The y-position on the right side of the main vehicle is at x ± ≈xc(ti)+d ± When the position is determined, the y-axis on the right side of the main vehicle can be approximated as:

[0185]

[0186] In the above formula, the superscripts + and - represent the upper and lower endpoints, respectively, and the y-values ​​of the corresponding edges are given by the linear equation:

[0187]

[0188] The position of the main vehicle's corresponding collision point relative to time t i The position of edge m can be approximated as:

[0189]

[0190] By steering, the driver will be satisfied. and The edges under the given conditions collide, where:

[0191]

[0192]

[0193] d = L0 - L c

[0194] The behavior of avoiding a collision by turning the steering wheel can be given by solving equations. It is possible to avoid a collision at time t. i Collision with edge m:

[0195]

[0196] In the formula:

[0197] At time t i The solution that avoids collisions by using the left side of edge m is as follows:

[0198]

[0199] Avoid collisions with rectangles:

[0200] c1≥max{c 1,i,m}

[0201] if Therefore, it is assumed that the vehicle cannot avoid a collision by turning to the left. The parameters selected from the above formula ensure that when... The final curvature is c0+c1t θ Not exceeding the maximum curvature c max .

[0202] In one embodiment of this application, the vehicle undergoes a lateral evaluation to perform automatic emergency steering or activate a takeover vehicle. Specifically:

[0203] The vehicle obtains the maximum lateral acceleration change rate; based on the main vehicle curvature change rate, the current lateral acceleration change rate is obtained; when the current lateral acceleration change rate is greater than the maximum lateral acceleration change rate, automatic emergency steering is performed or the takeover vehicle is activated.

[0204] Specifically, the vehicle parameterizes the driver's takeover of steering, with each steering action described by only one parameter (including left and right steering). The larger the parameter value, the more pronounced the driver's corresponding takeover ability. Furthermore, the vehicle is set to steer at a constant speed.

[0205] like Figure 14 As shown, Figure 14 This is a schematic diagram of steering behavior parameters in the target threat estimation method provided in this application. During normal steering, the driver typically turns the steering wheel to the left or right at a constant angle (i.e., applies a constant angular velocity to the steering wheel). There is a strong relationship between the maximum steering wheel angle and the maximum steering wheel angular velocity during collision avoidance. The driver's natural steering involves maintaining a constant steering wheel angular velocity over a finite time interval until the steering wheel reaches its final angle θc. Therefore, the change in steering wheel angle over time during collision avoidance maneuvers can be expressed as:

[0206]

[0207] in, t is the current time, θ0 is the initial steering wheel angle, such as Figure 14 As shown, the rate of change of the steering wheel angle (angular velocity) determines the degree of steering effort.

[0208] By taking into account the driver's personal steering habits, t can be preset. θ The value of t, and then select t according to the actual situation. θ This allows even sharp turns of the steering wheel to be described. Let θ max This indicates the maximum steering wheel angle that the driver can turn normally. This indicates the maximum steering wheel angular velocity when the driver is driving normally. This indicates the maximum lateral acceleration that the driver does not wish to exceed. This represents the maximum rate of change of lateral acceleration.

[0209]

[0210] in, or c left >c max If the driver is unable to avoid a collision by turning left, then that moment is considered the trigger point for steering to avoid a collision.

[0211] when or c right >c max If the driver is unable to avoid a collision by turning right, then that moment is considered the trigger point for steering to avoid a collision.

[0212] Furthermore, in another embodiment of this application, the vehicle obtains longitudinal threat estimation information based on the main motion prediction information and the target motion prediction information. The longitudinal threat estimation information includes acceleration values. To assess the future longitudinal threat to the target, the vehicle estimates how the main vehicle will move within a finite prediction time t. i Within the range [0, 3], collisions are avoided by braking or acceleration. Possible braking or acceleration behaviors are represented by the parameter j. r (Rate of change of acceleration or deceleration) is used to represent this. The goal is to find j r In order to represent the rectangular outline of the main vehicle at any t i Within [0,3], there will be no overlap with other targets represented by the rectangle. This problem is divided into two parts: First, estimate the overlap between the vehicle and the target throughout the entire prediction time vector, represented by the InPath flag. Second, calculate j r To avoid collisions with the complete rectangular outline throughout the entire prediction timeframe.

[0213] Please see details. Figure 15 , Figure 15 This is a flowchart illustrating the third embodiment of the target threat estimation method provided in this application.

[0214] like Figure 15 As shown, the specific steps are as follows:

[0215] Step S31: Obtain the closest longitudinal distance between the master vehicle and the target vehicle at each moment within the predicted time.

[0216] Please see Figure 16 , Figure 16 This is a schematic diagram of the target coordinate points required for the longitudinal InPath in the target threat estimation method provided in this application.

[0217] Specifically, the vehicle calculates the projections of half the target's width, its center of gravity to the front end, and its center of gravity to the rear end onto the X and Y axes, respectively. The coordinates of the target's four corner points are then calculated using a formula. The sign of cos(theta) determines which end of the target is closer to the main vehicle. When the target's rear end is closer to the main vehicle, the rear left and rear right sides are taken as the nearest edges, and the front left and front right sides are taken as the farthest edges. When the target's front end is closer to the main vehicle, the front left and front right sides are taken as the nearest edges, and the rear left and rear right sides are taken as the farthest edges. posNear represents the two corner points of the nearest edge, and posRemote represents the two corner points of the farthest edge. The lateral relative position between the target and the main vehicle is considered from left to right.

[0218] like Figure 17 As shown, Figure 17 This is a schematic diagram of an embodiment of the lateral relative position relationship between the target and the host vehicle in the target threat estimation method provided in this application.

[0219] In this embodiment, the longitudinal relative positional relationship between the main vehicle and the target vehicle is divided into three cases: the front or rear end of the target is close to the main vehicle, the left end of the target is close to the main vehicle, and the right side of the target is close to the main vehicle.

[0220] like Figure 18 As shown, Figure 18 This is a schematic diagram of another embodiment of the lateral relative position relationship between the target and the main vehicle in the target threat estimation method provided in this application. When the nearest edge is the front end, the front left and front right coordinate points are output; when the nearest edge is the rear end, the rear left and rear right coordinate points are output; when the nearest edge is the right end of the target, the front right and rear right coordinate points of the target are output; when the nearest edge is the left end of the target, the front left and rear left coordinate points of the target are output; all cases at the time of TTR time vector are traversed, and the coordinates of the nearest edge are represented by the two points (x1, y1) and (x2, y2).

[0221] Step S32: Obtain the acceleration value at each moment based on the nearest longitudinal distance at each moment.

[0222] like Figure 19 As shown, Figure 19 This is a schematic diagram of the longitudinal InPath judgment process in the target threat estimation method provided in this application. The above steps consider six cases of the relative positional relationship between the main vehicle and the target, and output the coordinates of the two nearest corner points in each of the six cases. The following is the determination of the longitudinal InPath relationship based on the coordinates of the nearest points:

[0223] The vehicle determines the magnitude of the lateral acceleration of the main vehicle and classifies the main vehicle into linear motion and circular motion.

[0224] In one embodiment of this application, the judgment condition is judgment condition 1, that is, the lateral position of the two ends of the target is within the vehicle width range (there is overlap in the y projection), and the ordinate of the two ends of the target is in front of the front of the main vehicle. Because checking whether there is a collision risk between the target and the main vehicle (whether there is an overlap risk), only the collision that occurs at the front of the main vehicle is considered.

[0225] The specific formula for condition 1 is as follows:

[0226] min(y 1pos ,y 2pos )<w h

[0227] max(y 1pos y 2pos )>-w h

[0228] max(x 1pos x 2pos )>d h -L h

[0229] Among them, w h It is half the width of the main vehicle, d h L is the distance from the rear axle to the front end. h For the train commander.

[0230] If condition 2 is true, that is, the lateral positions of the two ends of the target are within the vehicle width range, or the target side is parallel to the main vehicle, or the target side is perpendicular to the main vehicle, the nearest and farthest longitudinal distances are updated using the corner coordinates.

[0231] The specific formula for judgment condition 2 is as follows:

[0232] min(y 1pos y 2pos )<w h &

[0233] max(y 1pos y 2pos )>-w h ||

[0234] |y 1pos -y 2pos |<0.05||

[0235] |x 1pos x 2pos |<0.05

[0236] Judgment condition 4: The current time of the farthest edge is smaller than the historical time, and the current time of the nearest edge is larger than the historical time.

[0237] The specific formula for judgment condition 4 is as follows:

[0238]

[0239] Judgment condition 3: Only a portion of the target edge is on the trajectory of the main vehicle, and the target vehicle is not parallel to the main vehicle.

[0240] The main vehicle's motion is calculated according to the circular arc, and left turns and right turns are handled separately, using the theory of intersection between a straight line and a circular arc for calculation. Finally, it is also determined whether the nearest and farthest edges have been updated. Based on the updates of the nearest and farthest edges (the main vehicle and the target vehicle overlap longitudinally) and the target being located in front of the main vehicle, it is determined whether it is inPath at that moment; the InPath cases at all moments under the TTR time vector are traversed.

[0241] Step S33: Compare the acceleration value at each moment with the acceleration value at historical moments.

[0242] Specifically, the vehicle compares the acceleration value at each moment with the historical acceleration value at each moment.

[0243] Step S34: If the difference between the acceleration value at each moment and the acceleration value at a historical moment is greater than 0, retain the acceleration value at the current moment.

[0244] Specifically, if the difference between the vehicle's acceleration value at each moment and the acceleration value at a historical moment is greater than 0, the acceleration value at the current moment is retained.

[0245] Step S35: When the difference between the acceleration value at each moment and the acceleration value at a historical moment is less than 0, replace the acceleration value at the current moment with the acceleration value at the historical moment.

[0246] Please refer to details. Figure 20 , Figure 20 This is a schematic diagram illustrating the process of parametric calculation for longitudinal (braking, acceleration) collision avoidance provided in this application. For example... Figure 20As shown, the vehicle determines whether an InPath exists at the current time point. If InPath is true, it updates and calculates the longitudinal distance of the nearest edge. It then compares the braking jerk at the current time with the braking jerk at a previous time. If the current time is smaller, it saves the index value of that time and the braking jerk at the current time. Finally, it compares the acceleration jerk (uniform accelerometer) at the current time with the acceleration jerk (uniform accelerometer) at a previous time. If the current time is larger, it saves the index value of that time and the acceleration jerk (uniform accelerometer) at the current time.

[0247] By iterating through the time points in the TTR time dimension and repeatedly checking whether an InPath exists at that time point, the acceleration and deceleration requirements under the entire prediction time vector are finally obtained.

[0248] Step S36: Based on the acceleration value at each moment, obtain longitudinal threat estimation information between the host vehicle and the target vehicle.

[0249] In this embodiment of the application, the vehicle is capable of longitudinal evaluation, and when the vehicle's acceleration value is greater than the acceleration value or less than the braking value, it performs automatic emergency steering or activates takeover vehicle.

[0250] Specifically, the vehicle obtains the rate of change of acceleration based on the acceleration value. When the acceleration value is greater than the acceleration value or the rate of change of acceleration is greater than the maximum longitudinal acceleration rate of change, the vehicle performs automatic emergency steering or activates the takeover vehicle. Alternatively, when the acceleration value is less than the braking acceleration value or the rate of change of acceleration is less than the maximum longitudinal braking rate of change, the vehicle performs automatic emergency steering or activates the takeover vehicle.

[0251] Similarly, just like steering, braking and acceleration behaviors are parameterized as follows:

[0252] a=a0+j r t j

[0253] Among them, t j =min(t) j ,t), a0 is the initial longitudinal acceleration, j r The longitudinal acceleration rate of change

[0254] like Figure 21 and Figure 22 As shown, Figure 21 This is a schematic diagram of the turning behavior parameters in the target threat estimation method provided in this application; Figure 22 This is a parameterized schematic diagram of braking and acceleration in the target threat estimation method provided in this application.

[0255] Select time t jEven emergency braking and acceleration maneuvers can be described without exceeding the driver's preferences. The maximum longitudinal acceleration and the maximum rate of change of longitudinal acceleration that the driver does not wish to exceed during normal driving are denoted as a. max and j max During braking, the minimum acceleration that the driver does not want to fall below is denoted as a. min The minimum rate of change of acceleration is expressed as j min .

[0256] When j accel >j max or a accel >a max At that time, it was believed that the driver could not avoid the collision by accelerating;

[0257] When j brake <j max or a brake <a max At that time, it was believed that the driver could not avoid the collision by braking.

[0258] In this embodiment, the vehicle can also avoid a collision by braking or accelerating. This section discusses how the driver can avoid a collision within a limited prediction time t. i ∈[0,T max The system avoids collisions by braking or accelerating; possible braking or acceleration behaviors are represented by the parameter j. r Let it be that the goal is to find j r , t i ∈[0,T max The vehicle will not overlap with other targets represented by the rectangle. First, estimation is performed to avoid collisions with one side of the rectangle at a single time point. Second, estimation is performed to avoid collisions with the complete rectangular profile throughout the prediction time. It is assumed that the vehicle turns with a constant curvature when the driver brakes or accelerates. Because the vehicle turns with a constant curvature, it is convenient to switch the coordinate system instead of describing the vehicle's motion in a fixed curvilinear coordinate system (x, y) on the ground. Since the curvature is constant, the vehicle will travel along the x-axis of this coordinate system regardless of its speed. The origin is located at the initial position of the front of the vehicle.

[0259] 1) Use braking or acceleration to avoid a collision on one side:

[0260] In the embodiments of this application, j r The estimation is to avoid colliding with a certain edge m in time by braking or acceleration. In the curvature coordinate system, the edge is defined by [(x1, y1), (x2, y2)].

[0261] The front of the vehicle is positioned as shown below:

[0262]

[0263] The above formula can be expressed by solving the integral. and

[0264] The position of the rear end of the vehicle is given by the following formula:

[0265]

[0266] If the following occurs, then at time t i Avoid edge m by braking:

[0267]

[0268] Avoid by accelerating:

[0269]

[0270] The following two inequalities may allow for the avoidance of a collision through acceleration or braking.

[0271]

[0272]

[0273] Solve the above equations individually and extract the parameter j. r The desired rate of change of acceleration can be obtained:

[0274]

[0275] To avoid collision with edge m by accelerating:

[0276]

[0277] To avoid a collision through braking:

[0278]

[0279] 2) Brake or accelerate to avoid all the edges of the polygon:

[0280] The current goal is to find the necessary parameters to avoid a collision at any time, i.e., i = 1, 2, ..., N. Since the car will eventually come to a complete stop during deceleration, this must be considered in the solution. When the driver brakes to avoid a collision, the target's point of impact always remains in front of the vehicle. Furthermore, since the vehicle has a positive velocity v(t) ≥ 0 throughout the entire braking operation... It never lowered. Ensure the point of impact with the target is always in front of the vehicle. It is necessary to recursively update x i As:

[0281] x i =min(x i x i+1 i = N-1, N-2, ..., 1

[0282] Update x i Then, the rate of change of acceleration required to avoid a collision through braking is shown below:

[0283] j r ≤max(j r，i，m )

[0284] According to the assessment, if j r If the rate of change of acceleration exceeds the driver's normal driving behavior, the driver cannot avoid a collision by braking, i.e., j r ≤j min The deceleration required to avoid a collision is given by the above equation, i.e., a r =a0+j r t j The parameterization selected for potential braking operations ensures that the required deceleration does not exceed the preferred maximum deceleration a. r ≥a min j r ≥j min .

[0285] During acceleration maneuvers, the vehicle does not come to a complete stop. Therefore, this application can be used without modification, which gives the rate of change of acceleration required to avoid a collision throughout the entire prediction range:

[0286] j r ≥max{j r，i，m}

[0287] Where, j r，i，m This is provided by this application. It should be noted that if j r ≥j max The driver cannot avoid a collision by accelerating. Ensure a r ≤a min j r ≤j min .

[0288] The beneficial effects of this application are: obtaining braking delay time; obtaining the main motion delay information of the host vehicle after the braking delay time, and the target motion delay information of the target vehicle after the braking delay time; predicting the main motion prediction information of the host vehicle based on the main motion delay information; predicting the target motion prediction information of the target vehicle based on the target motion delay information; and obtaining lateral threat estimation information and / or longitudinal threat estimation information of the host vehicle and the target vehicle according to the main motion prediction information and the target motion prediction information. The vehicle in this application extracts the motion information of the target vehicle and the host vehicle, without relying on other unstable external environmental information, and introduces the concept of prediction to calculate the threat evaluation indicators of the target vehicle and the host vehicle in the lateral and longitudinal directions respectively. The degree of danger posed by the target vehicle to the host vehicle is estimated through the above dynamic process parameters.

[0289] This application's target threat estimation method is applied to intersection scenarios, using a simple motion model to predict the trajectories of vehicles and targets. This approach is not limited to various target types, including motorcycles, cars, bicycles, and pedestrians, and its algorithm is simple with low computational load. Lateral and longitudinal collision avoidance requests are calculated predictively, and combined with parameterized driver takeover capabilities, to assess the urgency of the target, demonstrating high reliability.

[0290] The above embodiments are merely one common example of this application and do not constitute any limitation on the technical scope of this application. Therefore, any minor modifications, equivalent changes, or alterations made to the above content based on the substance of the solution of this application shall still fall within the scope of the technical solution of this application.

[0291] Please continue reading Figure 23 , Figure 23 This is a schematic diagram of the structure of a vehicle embodiment provided in this application. The vehicle 500 of this application embodiment includes a processor 51, a memory 52, an input / output device 53, and a bus 54.

[0292] The processor 51, memory 52, and input / output device 53 are respectively connected to the bus 54. The memory 52 stores program data, and the processor 51 is used to execute the program data to implement the target threat estimation method described in any of the above embodiments.

[0293] In this embodiment, processor 51 can also be referred to as a CPU (Central Processing Unit). Processor 51 may be an integrated circuit chip with signal processing capabilities. Processor 51 can also be a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. The general-purpose processor can be a microprocessor, or processor 51 can be any conventional processor.

[0294] This application also provides a computer storage medium; please refer to the following: Figure 24 , Figure 24 This is a schematic diagram of a computer storage medium according to an embodiment of the present application. The computer storage medium 600 stores program data 61, which, when executed by a processor, is used to implement the target threat estimation method of any of the above embodiments.

[0295] When the embodiments of this application are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) or processor to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0296] The above description is merely an embodiment of this application and does not limit the patent scope of this application. Equivalent structural or procedural transformations made using 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.

Claims

1. A target threat estimation method based on crossroads, characterized in that, The target threat estimation method includes: Obtain the braking delay time; Acquire the main motion delay information of the main vehicle after the braking delay time, and the target motion delay information of the target vehicle after the braking delay time; Based on the main motion delay information, predict the main motion prediction information of the main vehicle; Based on the target motion delay information, predict the target motion prediction information of the target vehicle; Based on the main motion prediction information and the target motion prediction information, obtain the lateral threat estimation information and / or longitudinal threat estimation information of the main vehicle and the target vehicle; The lateral threat estimation information includes the rate of change of the vehicle's curvature. The step of obtaining lateral threat estimation information and / or longitudinal threat estimation information of the master vehicle and the target vehicle according to the master motion prediction information and the target motion prediction information includes: Based on the main motion prediction information and the target motion prediction information, obtain the two nearest sides and the diagonal sides between the target vehicle and the main vehicle; Using the two nearest edges and the diagonal edge, determine the corner coordinates of the target vehicle; Based on the corner coordinate data, the approximate collision point of the main vehicle is obtained; Based on the approximate collision point, determine the rate of change of curvature of the main vehicle used for steering to avoid collision; The step of obtaining the two closest edges between the target vehicle and the host vehicle includes: Obtain the midpoints of each edge of the target vehicle; Obtain the distance between the midpoint of each edge and the center of gravity of the main vehicle; The two closest edges are identified as the two closest edges between the target vehicle and the host vehicle. The step of obtaining the midpoints of each edge of the target vehicle includes: Add a preset compensation amount to the length and width of the target vehicle, and obtain the centroid coordinates and length and width projection values ​​of the target vehicle after adding the preset compensation amount; Based on the centroid coordinates and the length and width projections, the corner points of each side of the target vehicle are determined; Obtain the midpoints of each edge of the target vehicle based on the corner points of each edge.

2. The target threat estimation method according to claim 1, characterized in that, The target threat estimation method also includes: Obtain the maximum rate of change of lateral acceleration; Based on the curvature change rate of the main vehicle, the current lateral acceleration change rate is obtained; When the current rate of change of lateral acceleration is greater than the maximum rate of change of lateral acceleration, automatic emergency steering or vehicle takeover is activated.

3. The target threat estimation method according to claim 1, characterized in that, The longitudinal threat estimation information includes acceleration values; The step of obtaining lateral threat estimation information and / or longitudinal threat estimation information of the master vehicle and the target vehicle according to the master motion prediction information and the target motion prediction information includes: Obtain the nearest longitudinal distance between the master vehicle and the target vehicle at each moment within the predicted time period; Based on the nearest longitudinal distance at each time moment, the acceleration value at each time moment is obtained; Compare the acceleration value at each moment with the acceleration values ​​at historical moments; If the difference between the acceleration value at each moment and the acceleration value at a historical moment is greater than 0, the acceleration value at the current moment is retained. When the difference between the acceleration value at each moment and the acceleration value at a historical moment is less than 0, the acceleration value at the current moment is replaced by the acceleration value at the historical moment. Based on the acceleration value at each moment, longitudinal threat estimation information of the host vehicle and the target vehicle is obtained.

4. The target threat estimation method according to claim 3, characterized in that, The target threat estimation method also includes: When the acceleration value is greater than the acceleration value or less than the braking value, automatic emergency steering or vehicle takeover is activated.

5. The target threat estimation method according to claim 4, characterized in that, The step of performing automatic emergency steering or activating vehicle takeover when the acceleration value is greater than the acceleration value or less than the braking value includes: The rate of change of acceleration is obtained based on the acceleration value; When the acceleration value is greater than the acceleration acceleration value, or when the rate of change of acceleration is greater than the maximum rate of change of longitudinal acceleration acceleration, automatic emergency steering or vehicle takeover is activated. Alternatively, if the acceleration value is less than the braking acceleration value, or the rate of change of acceleration is less than the maximum rate of change of longitudinal braking acceleration, automatic emergency steering or vehicle takeover may be performed.

6. A vehicle, characterized in that, The vehicle includes a memory and a processor coupled to the memory; The memory is used to store program data, and the processor is used to execute the program data to implement the target threat estimation method as described in any one of claims 1 to 5.

7. A computer storage medium, characterized in that, The computer storage medium is used to store program data, which, when executed by the computer, is used to implement the target threat estimation method as described in any one of claims 1 to 5.

Images