Vehicle anti-collision warning method and system based on active perception control

Abstract

The application discloses a vehicle anti-collision early warning method and system based on active sensing control, and comprises the following steps: S1, scanning a target vehicle to obtain the relative position between the target vehicle and the ego vehicle; S2, estimating the motion state of the target vehicle based on the relative position as the initial condition; S3, delimiting a target area, and searching and tracking the target vehicle in the target area based on the motion state of the target vehicle; S4, obtaining the position and speed of the target vehicle in real time based on the tracking of the target vehicle; and S5, determining the collision risk between the target vehicle and the ego vehicle based on the position and speed of the target vehicle, and performing anti-collision early warning according to the collision risk. The application can timely early warn the collision risk of the vehicle behind and the vehicle at the side and rear of the ego vehicle, achieves the effect of preventing vehicle collision, and guarantees the driving safety of the vehicle.

Description

Technical Field

[0001] This invention relates to the field of vehicle driving safety, specifically to a vehicle collision avoidance warning method and system based on active perception control. Background Technology

[0002] With the development of autonomous driving technology, users have increasingly higher demands for efficient and stable driver assistance systems. Collision avoidance warning technology is an important technology that provides drivers with risk information during vehicle operation and is a crucial component of driver assistance technology. It assesses the risks posed by road vehicles and obstacles, issues warnings, and helps avoid collisions.

[0003] Currently, existing vehicle collision avoidance warning methods or systems suffer from significant deviations in monitoring or tracking target vehicles near the current vehicle, and there are also delays in obtaining real-time motion information of the target vehicles. Consequently, the collision avoidance warnings are not timely enough, and the effect is unsatisfactory. Therefore, to solve these problems, a vehicle collision avoidance warning method and system based on active perception and control is needed. Summary of the Invention

[0004] In view of this, the purpose of the present invention is to overcome the defects in the prior art and provide a vehicle collision avoidance warning method and system based on active perception control, which can promptly warn of the collision risk of vehicles behind and to the sides of the vehicle, thereby achieving the effect of preventing vehicle collisions and ensuring vehicle driving safety.

[0005] The vehicle collision avoidance warning method based on active perception control of the present invention includes the following steps:

[0006] S1. Scan the target vehicle to obtain the relative position between the target vehicle and your own vehicle;

[0007] S2. Using the relative position as the initial condition, estimate the motion state of the target vehicle;

[0008] S3. Define the target area, and within the target area, search for and track the target vehicle based on its motion state;

[0009] S4. Based on tracking the target vehicle, obtain the target vehicle's position and speed in real time;

[0010] S5. Based on the position and speed of the target vehicle, determine the collision risk between the target vehicle and your own vehicle, and issue a collision avoidance warning based on the collision risk.

[0011] Furthermore, the target vehicles include vehicles directly behind the vehicle and vehicles to the side and rear of the vehicle.

[0012] Furthermore, using the relative position as an initial condition, the motion state of the target vehicle is estimated, specifically including:

[0013] Determine the motion model of the target vehicle; the motion model includes a linear motion model and a turning motion model.

[0014] An interactive vehicle motion estimation model is constructed. The initial conditions are input into the motion estimation model for interactive iteration, and the future motion speed and direction of the target vehicle are output.

[0015] Furthermore, the target area is defined, specifically including:

[0016] Using the middle position of the rear edge of the vehicle as the starting point, the starting point as the origin of the coordinate system, the opposite direction of the vehicle's driving direction as the positive direction of the X-axis, and the Y-axis perpendicular to the X-axis through the origin of the coordinate system, a two-dimensional coordinate system is formed.

[0017] Using the side along the positive Y-axis as the right side, the target region is defined on the right side, specifically including:

[0018] Starting from the starting point, a forward scan is performed directly behind the vehicle. The point with the furthest effective distance in the forward scan is taken as the endpoint. The starting point and the endpoint are connected to form a line l1. In the width direction of the lane, the point with a longitudinal distance of width W between it and the endpoint is taken as the second endpoint. From the second endpoint, an extension line l2 parallel to the line l1 is drawn, and the length of the extension line l2 is equal to that of the line l1. The end point of the extension line l2 is taken as the second starting point. Starting from the starting point, the endpoint, the second endpoint, the second starting point, and the starting point are connected in sequence to form a rectangle D1, and rectangle D1 is taken as the first target area. Here, 1 / 4 of the lane width is taken as the width W.

[0019] On the lane adjacent to the lane where the vehicle is located, select a point whose longitudinal distance from the second endpoint is half the width of the lane as the first target point. Along the positive Y-axis, select a point whose longitudinal distance from the first target point is width W as the second target point. Parallel to the lane length direction, draw extension lines l3 and l4 starting from the first target point and the second target point respectively, so that the endpoint of the extension line l4 exceeds the maximum tilt angle range of the lateral scan. Select the endpoint of the extension line l4 as the third target point. Select a point on the extension line l3 whose longitudinal distance from the third target point is width W as the fourth target point. Starting from the first target point, connect the second target point, the third target point, the fourth target point and the first target point in sequence to form a rectangle D2. Select rectangle D2 as the second target area.

[0020] Using the side along the negative Y-axis as the left side, the target region is delineated on the left side, specifically including:

[0021] Following the same principle as defining the target region for the right-side region, the target region is defined for the left-side region.

[0022] Furthermore, the first target region is designated as the first sub-target region θ1;

[0023] Connect the starting point and the second target point to form a line L2. Draw a perpendicular line l1 from the intersection of line L2 and its extension l3 to the extension l3. The perpendicular line l1 divides rectangle D2 into rectangles D along the negative X-axis. 21 and rectangle D 22 , reshape D 21 As the second sub-target region θ2;

[0024] Take the intersection of perpendicular line l1 and extension line l4 as point O1, and connect the starting point and point O1 to form line L3. Draw a perpendicular line l2 to the extension line l3 through the intersection of line L3 and extension line l3. Perpendicular line l2 will divide rectangle D. 22 Divide the area into rectangles D along the negative X-axis. 221 and rectangle D 222 , reshape D 221 As the third sub-target region θ3;

[0025] Following the same method used to obtain the third sub-target region θ3, we can obtain the fourth sub-target region θ4, ..., the nth sub-target region θ3, and so on. n Until the intersection point O of the subsequent perpendicular line and its extension l4. n-2 The line Ln connecting the starting point to the extended line l3 will continue to intersect with the extended line l3 until the line Ln no longer intersects with the extended line l3.

[0026] Furthermore, the collision risk between the target vehicle and the vehicle itself is determined, and a collision avoidance warning is issued based on the collision risk, specifically including:

[0027] Construct a collision risk model between the target vehicle and your own vehicle:

[0028]

[0029] Where TTC is the estimated time before a collision occurs between the target vehicle and the vehicle itself; d rel The distance between the target vehicle and the vehicle itself; v obj v represents the speed of the target vehicle. ego The speed of the vehicle itself; a obj The acceleration of the target vehicle quantity; a ego The acceleration of its own vehicle;

[0030] Set collision time threshold (TTC) t If the collision time TTC is less than the collision time threshold TTC tThen, a collision warning will be issued to the vehicle.

[0031] A vehicle collision avoidance warning system based on active perception control includes a scanning unit, a processing unit, and an alarm unit;

[0032] The scanning unit is used to scan the target vehicle;

[0033] The processing unit is used to obtain the relative position between the target vehicle and its own vehicle based on the scanning data acquired by the scanning unit; estimate the motion state of the target vehicle using the relative position as an initial condition; delineate a target area, and search for and track the target vehicle within the target area based on its motion state; acquire the position and speed of the target vehicle in real time based on the tracking of the target vehicle; determine the collision risk between the target vehicle and its own vehicle based on the position and speed of the target vehicle, and provide a collision avoidance warning based on the collision risk.

[0034] The alarm unit is used to issue a collision warning reminder based on the collision avoidance warning results analyzed by the processing unit.

[0035] Furthermore, the scanning unit includes a rotating platform mounted on the vehicle itself and a lidar mounted on the rotating platform.

[0036] The beneficial effects of this invention are as follows: The vehicle collision avoidance warning method and system based on active perception control disclosed in this invention determine nonlinear detection methods according to the target vehicle motion model, initially realizing the scanning and tracking of the target vehicle motion. By delineating the target area, within the target area, an active perception algorithm is determined based on the vehicle search strategy of angle and area, realizing the simultaneous search and tracking of the target area and the target vehicle. Through the uncertain vehicle tracking strategy, the lateral error of the vehicle is eliminated, realizing further prediction and tracking of the target vehicle motion. It can track and warn of vehicles in the blind spots of the rear and side rearwards during the driving process of the vehicle, achieving the effect of preventing vehicle collisions and ensuring the driving safety of the vehicle. Attached Figure Description

[0037] The present invention will be further described below with reference to the accompanying drawings and embodiments:

[0038] Figure 1 This is a schematic diagram of the early warning method of the present invention;

[0039] Figure 2 This is a schematic diagram of a two-dimensional coordinate system dominated by the vehicle itself, according to the present invention.

[0040] Figure 3 A schematic diagram illustrating the delineation of the target area in this invention;

[0041] Figure 4This is a schematic diagram of the target vehicle detection area after the target area has been defined according to the present invention;

[0042] Figure 5 This is a schematic diagram illustrating the uncertain error that the lidar of the present invention can identify;

[0043] Figure 6 This is a schematic diagram of the IIMM-EKF-PDF process of the present invention. Detailed Implementation

[0044] The present invention will be further described below with reference to the accompanying drawings, as shown in the figures:

[0045] The vehicle collision avoidance warning method based on active perception control of the present invention includes the following steps:

[0046] S1. Scan the target vehicle to obtain the relative position between the target vehicle and the vehicle itself; wherein, the target vehicle includes vehicles directly behind the vehicle and vehicles to the side and rear of the vehicle.

[0047] S2. Using the relative position as the initial condition, estimate the motion state of the target vehicle;

[0048] S3. Define the target area, and within the target area, search for and track the target vehicle based on its motion state;

[0049] S4. Based on tracking the target vehicle, obtain the target vehicle's position and speed in real time;

[0050] S5. Based on the position and speed of the target vehicle, determine the collision risk between the target vehicle and your own vehicle, and issue a collision avoidance warning based on the collision risk.

[0051] In this embodiment, in step S1, a rotating platform is set at the rear of the vehicle, and a LiDAR is installed on the rotating platform to scan the target vehicle. The LiDAR scans clockwise, identifying the front corner of the target vehicle by scanning the front and sides, and updating the data; then it scans counterclockwise to detect the target. At the last moment before the target is lost, the time and distance of the laser reflection are determined as the initial position data of the target vehicle, and the relative position is calculated based on the initial position data of the target vehicle.

[0052] In this embodiment, step S2, using the relative position as the initial condition, estimates the motion state of the target vehicle, specifically including:

[0053] The motion of the target vehicle is summarized as a non-single model, namely a linear motion model and a turning motion model. Due to the existence of model errors in actual observation, an improved interacting multiple model (IIMM) is used to estimate the vehicle's running speed. This includes model input interaction, model matching filtering, and output interaction.

[0054] Model input interaction:

[0055]

[0056]

[0057]

[0058]

[0059]

[0060] In the formula, This is the previous estimate. For the previous covariance, mix to provide input to the next filter, u i|j For the mixed probability, p ij Let be the mode transition probability, which includes the probability of transitioning from mode i to j.

[0061] Model-matched filtering:

[0062] Since the constant speed model and the approximate coordinated turning model are nonlinear models, the models need to be linearized, so the extended Kalman filter (EKF) is used.

[0063] Output interaction:

[0064]

[0065]

[0066] Secondly, based on the constructed IIMM framework, the motion of the future target vehicle is predicted and calculated. After obtaining the initial conditions, model interaction is performed, and the initial conditions for the next iteration of interaction are calculated. Finally, the motion speed and direction of the target vehicle in each mode are calculated.

[0067] In this embodiment, step S3, delineating the target area, specifically includes:

[0068] Using the middle of the rear edge of the vehicle as the starting point, and this starting point as the origin, with the opposite direction of the vehicle's travel direction as the positive X-axis, and a Y-axis perpendicular to the X-axis drawn through the origin, a two-dimensional coordinate system is formed; the constructed two-dimensional coordinate system is as follows: Figure 2 As shown, A is the vehicle itself, and B and C are the target vehicles.

[0069] Using the side along the positive Y-axis as the right side, the target region is defined on the right side, such as... Figure 3 , 4 As shown, it specifically includes:

[0070] Starting from the starting point, a forward scan is performed directly behind the vehicle. The point with the furthest effective distance in the forward scan is taken as the endpoint. The starting point and the endpoint are connected to form a line l1. In the width direction of the lane, the point with a longitudinal distance of width W between it and the endpoint is taken as the second endpoint. From the second endpoint, an extension line l2 parallel to the line l1 is drawn, and the length of the extension line l2 is equal to that of the line l1. The end point of the extension line l2 is taken as the second starting point. Starting from the starting point, the endpoint, the second endpoint, the second starting point, and the starting point are connected in sequence to form a rectangle D1, and rectangle D1 is taken as the first target area. Here, 1 / 4 of the lane width is taken as the width W.

[0071] On the lane adjacent to the lane where the vehicle is located, select a point whose longitudinal distance from the second endpoint is half the width of the lane as the first target point. Along the positive Y-axis, select a point whose longitudinal distance from the first target point is width W as the second target point. Parallel to the lane length direction, draw extension lines l3 and l4 starting from the first target point and the second target point respectively, so that the endpoint of the extension line l4 exceeds the maximum tilt angle range of the lateral scan. Select the endpoint of the extension line l4 as the third target point. Select a point on the extension line l3 whose longitudinal distance from the third target point is width W as the fourth target point. Starting from the first target point, connect the second target point, the third target point, the fourth target point and the first target point in sequence to form a rectangle D2. Select rectangle D2 as the second target area.

[0072] Using the side along the negative Y-axis as the left side, the target region is delineated on the left side, specifically including:

[0073] Following the same principle as defining the target region for the right-side region, the target region is defined for the left-side region.

[0074] Since vehicles occupy most of the lateral space in the lane, a portion of the scanning area is invalid. By using the target area delineation method described above, the detection or search of invalid areas can be eliminated, thereby improving the efficiency of detecting or searching for target vehicles.

[0075] Furthermore, such as Figure 3 As shown, the first target region is taken as the first sub-target region θ1;

[0076] Connect the starting point and the second target point to form a line L2. Draw a perpendicular line l1 from the intersection of line L2 and its extension l3 to the extension l3. The perpendicular line l1 divides rectangle D2 into rectangles D along the negative X-axis. 21 and rectangle D 22 , reshape D 21 As the second sub-target region θ2;

[0077] Take the intersection of perpendicular line l1 and extension line l4 as point O1, and connect the starting point and point O1 to form line L3. Draw a perpendicular line l2 to the extension line l3 through the intersection of line L3 and extension line l3. Perpendicular line l2 will divide rectangle D. 22 Divide the area into rectangles D along the negative X-axis. 221 and rectangle D 222 , reshape D 221 As the third sub-target region θ3;

[0078] Following the same method used to obtain the third sub-target region θ3, we can obtain the fourth sub-target region θ4, ..., the nth sub-target region θ3, and so on. n Until the intersection point O of the subsequent perpendicular line and its extension l4. n-2 The line Ln connecting the starting point to the extended line l3 no longer intersects with the extended line l3. In this embodiment, a total of 6 sub-target regions are divided.

[0079] The above method further delineates the target area, enabling discretization of the region of interest and further improving the search and tracking of target vehicles.

[0080] In this embodiment, in step S3, after defining the target area, a search strategy based on angle and region is formulated to control the search area of ​​the LiDAR and improve vehicle tracking efficiency. The search probability of each region depends on the search score of that region.

[0081]

[0082]

[0083]

[0084] In the formula, P s,k For search probability, Here is the state transition matrix, β is the input weight, and δ is the input weight. i,k To determine whether to search region i at time k, Let Ψ be the angle covered by the i-th sub-region scanned by the lidar at time k. i R is the maximum angle covered by the i-th sub-region. i,k For the area search score, u k This is the input value for the lidar.

[0085] If the lidar detects an area, the search score for that area decreases. Conversely, if the area is not detected, the search score for that area increases over time, keeping pace with the time update. This allows for simultaneous searching and tracking of both the target area and the target vehicle.

[0086] In this embodiment, step S4 involves formulating a vehicle estimation strategy based on uncertainty. This addresses the problem that, due to the relatively narrow beam of a single laser, the relative distance to the detected vehicle alone is insufficient to provide enough spatial state information, and a single detection is insufficient to accurately estimate the vehicle's lateral and longitudinal motion attitude. Vehicle motion is estimated using IIMM-EKF (Interactive Multi-Model Algorithm-Extended Kalman Filter). The predicted vehicle motion and sensor orientation establish physical constraints for the vehicle motion. To merge the constraints of the two models, PDF (probability density function) truncation is used, providing a more favorable vehicle motion error covariance for identification. This constructs a better vehicle attitude uncertainty model, used to control the scanning direction of the lidar sensor and determine whether to perform search and tracking.

[0087] The state estimates are then calculated as the truncated mean using a Gaussian distribution truncated at the constraint edges. The mean and variance of the constraint states are:

[0088]

[0089]

[0090] After s iterations, the final constrained state estimate and variance are obtained:

[0091]

[0092]

[0093] Therefore, in lateral position estimation, the cutoff range is:

[0094]

[0095] In the formula, γ Y To reduce the weights (0, 1) of the vehicle's lateral position uncertainty, such as Figure 5 Even if the incident angle of a single laser beam is too large (or too small), a reasonable limit can still be obtained by predicting the vehicle's motion state.

[0096] Therefore, accurate tracking of vehicle status is ensured, and the truncated IIMM-EKF-PDF estimator is now complete. The process is as follows: Figure 6 The target vehicle's state variables were obtained, which consisted of its two-dimensional coordinates and corresponding speed.

[0097] In this embodiment, in step S5, based on accurate tracking and prediction of the vehicle's motion state (position and speed), the vehicle collision risk is predicted, and the collision risk is assessed from a kinematic perspective using collision time. The collision risk between the target vehicle and the vehicle itself is determined, and a collision avoidance warning is issued based on the collision risk, specifically including:

[0098] Construct a collision risk model between the target vehicle and your own vehicle:

[0099]

[0100] Where TTC is the estimated time before a collision occurs between the target vehicle and the vehicle itself; d rel The distance between the target vehicle and the vehicle itself; v obj v represents the speed of the target vehicle. ego The speed of the vehicle itself; a obj The acceleration of the target vehicle quantity; a ego The acceleration of its own vehicle;

[0101] Set collision time threshold (TTC) t If the collision time TTC is less than the collision time threshold TTC t If the system detects a collision, a collision warning will be issued to the vehicle. The collision time threshold (TTC) can be set according to actual operating conditions. t As shown in Table 1, the collision time threshold (TTC) is set under different operating conditions. t .

[0102] Table 1

[0103]

[0104] The present invention also relates to a vehicle collision avoidance warning system based on active perception control. The system corresponds to the above-mentioned vehicle collision avoidance warning method based on active perception control and can be understood as a system that implements the above-mentioned method. The system includes a scanning unit, a processing unit, and an alarm unit.

[0105] The scanning unit is used to scan the target vehicle;

[0106] The processing unit is used to obtain the relative position between the target vehicle and its own vehicle based on the scanning data acquired by the scanning unit; estimate the motion state of the target vehicle using the relative position as an initial condition; delineate a target area, and search for and track the target vehicle within the target area based on its motion state; acquire the position and speed of the target vehicle in real time based on the tracking of the target vehicle; determine the collision risk between the target vehicle and its own vehicle based on the position and speed of the target vehicle, and provide a collision avoidance warning based on the collision risk; wherein, the processing unit includes a microprocessor;

[0107] The alarm unit is used to issue a collision warning reminder based on the collision avoidance warning results analyzed by the processing unit. The alarm unit includes a buzzer, the signal control input of which is connected to the signal control output of the microprocessor.

[0108] In this embodiment, the scanning unit includes a rotating platform mounted on the vehicle and a lidar mounted on the rotating platform. A stepper motor can be used to drive the rotating platform to rotate, enabling the lidar to perform rotational scanning. Alternatively, to improve scanning efficiency, two rotating platforms can be arranged, each equipped with a lidar. The rotation angle of the rotating platforms ranges from 0 to 33.3°. One lidar scans the right-hand area defined in the two-dimensional coordinate system, while the other lidar scans the left-hand area.

[0109] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A vehicle collision avoidance warning method based on active perception control, characterized in that: Includes the following steps: S1. Scan the target vehicle to obtain the relative position between the target vehicle and your own vehicle; S2. Using the relative position as the initial condition, estimate the motion state of the target vehicle; S3. Define the target area, and within the target area, search for and track the target vehicle based on its motion state; Define the target area, specifically including: Using the middle position of the rear edge of the vehicle as the starting point, the starting point as the origin of the coordinate system, the opposite direction of the vehicle's driving direction as the positive direction of the X-axis, and the Y-axis perpendicular to the X-axis through the origin of the coordinate system, a two-dimensional coordinate system is formed. Using the side along the positive Y-axis as the right side, the target region is defined on the right side, specifically including: Starting from the starting point, perform a forward scan directly behind the vehicle, taking the furthest effective distance point of the forward scan as the endpoint, and connect the starting point and the endpoint to form a line. In the width direction of the lane, a point with a longitudinal distance of width W from the endpoint is designated as the second endpoint. Starting from the second endpoint, a line parallel to the connecting line is drawn. extension line And make the extension line The length of the line is equal to the length of the line. Extend the line The last endpoint is used as the second starting point; starting from the starting point, the last endpoint, the second starting point, and the starting point are connected sequentially to form a rectangle. and the rectangle This is the first target area; where 1 / 4 of the lane width is taken as the width W; On the lane adjacent to your vehicle's lane, select a point whose longitudinal distance from the second endpoint is half the lane width as the first target point. Along the positive Y-axis, select a point whose longitudinal distance from the first target point is width W as the second target point. Extend lines parallel to the lane length direction, starting from both the first and second target points. as well as , making the extension line The endpoint exceeds the maximum tilt angle range of the lateral scan and extends the line. The endpoint will be used as the third target point, and will be on the extended line. The point with a vertical distance of width W between the first and third target points is designated as the fourth target point. Starting from the first target point, the second, third, and fourth target points are sequentially connected back to the first target point, forming a rectangle. and the rectangle As the second target area; Using the side along the negative Y-axis as the left side, the target region is delineated on the left side, specifically including: Following the same principle as defining the target region for the right-side region, the target region is defined for the left-side region. S4. Based on tracking the target vehicle, obtain the target vehicle's position and speed in real time; S5. Based on the position and speed of the target vehicle, determine the collision risk between the target vehicle and your own vehicle, and issue a collision avoidance warning based on the collision risk.

2. The vehicle collision avoidance warning method based on active perception control according to claim 1, characterized in that: The target vehicles include vehicles directly behind the vehicle and vehicles to the side and rear of the vehicle.

3. The vehicle collision avoidance warning method based on active perception control according to claim 1, characterized in that: Using the relative position as the initial condition, the motion state of the target vehicle is estimated, specifically including: Determine the motion model of the target vehicle; the motion model includes a linear motion model and a turning motion model. An interactive vehicle motion estimation model is constructed. The initial conditions are input into the motion estimation model for interactive iteration, and the future motion speed and direction of the target vehicle are output.

4. The vehicle collision avoidance warning method based on active perception control according to claim 1, characterized in that: The first target region is designated as the first sub-target region. ; Connect the starting point and the second target point to form a line. , through the connection With extension line Extend the line from the intersection point between them perpendicular line ,perpendicular Rectangle Divide into rectangles sequentially along the negative X-axis direction. and rectangle , to rectangle As the second sub-target area ; perpendicular line With extension line The intersection point is taken as a point Connect the starting point and the point Form a line , through the connection With extension line Extend the line from the intersection point between them perpendicular line ,perpendicular Rectangle Divide into rectangles sequentially along the negative X-axis direction. and rectangle , to rectangle As the third sub-target area ; According to the obtained third sub-target area By analogy, the fourth sub-target region was obtained. ,…,No. Sub-target area until the subsequent perpendicular lines and their extensions intersection The line connecting the starting point No longer with extension line Continue until an intersection point is found.

5. The vehicle collision avoidance warning method based on active perception control according to claim 1, characterized in that: Determine the collision risk between the target vehicle and your own vehicle, and issue a collision avoidance warning based on the collision risk, specifically including: Construct a collision risk model between the target vehicle and your own vehicle: ； in, This is the estimated time when a collision is about to occur between the target vehicle and the vehicle itself. The distance between the target vehicle and the vehicle itself; The speed of the target vehicle; The speed of its own vehicle; The acceleration of the target vehicle quantity; The acceleration of its own vehicle; Set collision time threshold If the collision time Less than the collision time threshold Then, a collision warning will be issued to the vehicle.

6. A vehicle collision avoidance warning system based on active perception control, characterized in that: It includes a scanning unit, a processing unit, and an alarm unit; The scanning unit is used to scan the target vehicle; The processing unit is used to obtain the relative position between the target vehicle and its own vehicle based on the scanning data acquired by the scanning unit; and to estimate the motion state of the target vehicle using the relative position as an initial condition. Define a target area, and within that area, search for and track the target vehicle based on its motion state; based on the tracking of the target vehicle, obtain its position and speed in real time. Based on the position and speed of the target vehicle, determine the collision risk between the target vehicle and your own vehicle, and issue a collision avoidance warning based on the collision risk. Define the target area, specifically including: Using the middle position of the rear edge of the vehicle as the starting point, the starting point as the origin of the coordinate system, the opposite direction of the vehicle's driving direction as the positive direction of the X-axis, and the Y-axis perpendicular to the X-axis through the origin of the coordinate system, a two-dimensional coordinate system is formed. Using the side along the positive Y-axis as the right side, the target region is defined on the right side, specifically including: Starting from the starting point, perform a forward scan directly behind the vehicle, taking the furthest effective distance point of the forward scan as the endpoint, and connect the starting point and the endpoint to form a line. In the width direction of the lane, a point with a longitudinal distance of width W from the endpoint is designated as the second endpoint. Starting from the second endpoint, a line parallel to the connecting line is drawn. extension line And make the extension line The length of the line is equal to the length of the line. Extend the line The last endpoint is used as the second starting point; starting from the starting point, the last endpoint, the second starting point, and the starting point are connected sequentially to form a rectangle. and the rectangle This is the first target area; where 1 / 4 of the lane width is taken as the width W; On the lane adjacent to your vehicle's lane, select a point whose longitudinal distance from the second endpoint is half the lane width as the first target point. Along the positive Y-axis, select a point whose longitudinal distance from the first target point is width W as the second target point. Extend lines parallel to the lane length direction, starting from both the first and second target points. as well as , making the extension line The endpoint exceeds the maximum tilt angle range of the lateral scan and extends the line. The endpoint will be used as the third target point, and will be on the extended line. The point with a vertical distance of width W between the first and third target points is designated as the fourth target point. Starting from the first target point, the second, third, and fourth target points are sequentially connected back to the first target point, forming a rectangle. and the rectangle As the second target area; Using the side along the negative Y-axis as the left side, the target region is delineated on the left side, specifically including: Following the same principle as defining the target region for the right-side region, the target region is defined for the left-side region. The alarm unit is used to issue a collision warning reminder based on the collision avoidance warning results analyzed by the processing unit.

7. The vehicle collision avoidance warning system based on active perception control according to claim 6, characterized in that: The scanning unit includes a rotating platform mounted on the vehicle itself and a lidar mounted on the rotating platform.

Images