JP7551242B2 - Driving Support Devices - Google Patents

Description

The present invention relates to a driving assistance device that assists in driving a vehicle.

Conventionally, there are devices that assist in driving a vehicle by avoiding collisions with objects located around the vehicle and reducing damage in the event of a collision. For example, the driving assistance device described in Patent Document 1 is a collision determination device that determines whether or not a collision will occur with an object located around the vehicle, calculates an estimated path of the vehicle and an estimated path of the surrounding object, and if there is an intersection between the vehicle and the surrounding object, determines that a collision will occur with the object.

JP 2020-8288 A (see paragraphs 0019 to 0059, etc.)

The collision determination device of Patent Document 1 calculates an estimated path of the host vehicle based on the rate of change in steering amount of the host vehicle and the host vehicle speed, and calculates the area of the host vehicle at each position on the estimated path. The area of the host vehicle is calculated based on vehicle specifications that indicate the size of the host vehicle. In addition, an estimated path of the object is calculated based on the position of the object detected by the object detection device and the relative speed of the object with respect to the host vehicle, and calculates the area of the object at each position on the estimated path. The area of the object is calculated based on the size of the object calculated by the object detection device. Then, the determination unit determines that a collision has occurred between the host vehicle and the object when there is an overlap between the area of the host vehicle and the area of the object at each position on the estimated path of the host vehicle.

As described above, the collision determination device described in Patent Document 1 requires various calculation and determination processes, but since there are usually multiple objects for which collision determination must be performed, there is a problem that the processing load on the ECU and other calculation units increases.

The present invention has been made in consideration of the above problems, and aims to provide a driving assistance device that can reduce the processing load on the calculation unit by performing detailed collision prediction only for targets with a possibility of collision.

In order to achieve the above object, a driving assistance device of the present invention is a driving assistance device that assists driving of a host vehicle by a driving assistance function, and includes vehicle position prediction means for predicting a future position of the host vehicle after a predetermined time, area setting means for setting a vehicle area surrounding the future position, target position prediction means for predicting a future position of a target after the predetermined time, judgment means for judging whether the future position of the target predicted by the target position prediction means is included in the vehicle area set by the area setting means, function activation means for activating the driving assistance function when it is judged by the judgment means that the future position of the target is included in the vehicle area, and object detection means for deriving a traveling direction of the target. and a target travel direction derivation means, wherein the judgment means performs a first simple judgment in which the target is subject to the judgment if the distance between the vehicle and the target is within a predetermined distance set in advance, but is not subject to the judgment if the distance is not within the predetermined distance , and when it is judged that the target is subject to the judgment in the first simple judgment, performs a second simple judgment in which the target is not subject to the judgment if there is no point where a straight line that is parallel to the travel direction of the target derived by the target travel direction derivation means and passes through the position of the target passes through a possible existence area, which is an area including the predicted path of the vehicle, and is subject to the judgment if there is a point where it passes through the possible existence area (Claim 1).

According to the invention of claim 1, the judgment means does not make a judgment on all targets, but rather , a first simple judgment is used to judge whether or not the target is to be judged based on whether the distance between the vehicle and the target is within a predetermined distance, thereby narrowing down the targets to be judged to those within the predetermined distance, and for the targets narrowed down as targets to be judged , a second simple judgment is further used to judge not to judge the target if there is no point where a straight line that is parallel to the target's traveling direction and passes through the target's position passes through a possible presence area, which is an area that includes the vehicle's predicted course, but to judge the target if there is a point where it passes through the possible presence area, thereby further reducing the burden of the judgment processing on the judgment means and enabling the driving assistance function to be activated based on a more accurate judgment.

1 is a block diagram showing a configuration of a driving assistance system according to an embodiment of the present invention; FIG. 11 is a diagram showing the flow of a function activation process. 13 is a flowchart showing a flow of a future position estimation process. FIG. 2 is a diagram showing a vehicle and an object (another vehicle) present in front of the vehicle. FIG. 2 is a diagram for explaining a vehicle area. 13 is a flowchart showing the flow of an operational function determination process. 10A and 10B are diagrams for explaining an example of a simple determination of the presence or absence of a collision possibility.

The following describes in detail an embodiment of the present invention with reference to the attached drawings.

<Driving assistance system>
FIG. 1 is a block diagram showing a configuration of a driving assistance system 1 according to an embodiment of the present invention.

Driving assistance system 1 is a system that is mounted on vehicle A (see FIG. 4) and assists in driving vehicle A. As driving assistance functions for avoiding a collision between vehicle A and a target T (see FIG. 4) such as a vehicle ahead or reducing damage due to the collision, driving assistance system 1 has a collision warning function that warns of the possibility of a collision before the automatic brake is activated, a primary braking function (gentle braking function) that decelerates vehicle A at a primary target deceleration by the automatic brake to encourage the driver to take action to avoid the collision, and a secondary braking function (hard braking function) that decelerates vehicle A at a secondary target deceleration that is greater than the primary target deceleration by the automatic brake to avoid the collision and reduce damage due to the collision.

The driving assistance system 1 includes an ECU (Electronic Control Unit) 11. The ECU 11 is equipped with a microcomputer (microcontroller) 12. The microcomputer 12 has a built-in CPU 13 and memory 14. In addition to the ECU 11, vehicle A is equipped with multiple ECUs for controlling each part, and the ECU 11 is connected to the other ECUs so that bidirectional communication is possible using the CAN (Controller Area Network) communication protocol.

Vehicle A is also equipped with a camera 21. Camera 21 is, for example, a stereo camera capable of continuously capturing still images at a predetermined frame rate, and is installed, for example, in front of the rearview mirror in the center of the front part of the vehicle interior so as to capture a wide-angle image of the area in front of vehicle A. Camera 21 extracts target pixels corresponding to the same object in each image captured by the image sensors from a pair of image data input from the image sensors of both the left and right eyes, detects the amount of deviation in the position of the target pixels between the pair of images, and calculates the distance to the same object using the principle of triangulation. The output signal of camera 21 is input to ECU 11.

Vehicle A is further provided with a vehicle speed sensor 22, a steering angle sensor 23, and a yaw rate sensor 24. Vehicle speed sensor 22 outputs a detection signal in the form of a pulse signal synchronized with the rotation of a rotating body (e.g., a drive shaft) that rotates as vehicle A travels. Steering angle sensor 23 outputs a detection signal corresponding to the steering angle (absolute steering angle) relative to the steering angle midpoint of the steering mechanism (e.g., a steering wheel). The steering angle takes a positive value when the steering mechanism is turned to the right from the steering angle midpoint (the steering wheel is turned to the right), and takes a negative value when the steering mechanism is turned to the left (the steering wheel is turned to the left). Yaw rate sensor 24 outputs a detection signal corresponding to the yaw rate, which is the angular velocity of rotation around a vertical axis passing through the center of gravity of vehicle A. The detection signals of vehicle speed sensor 22, steering angle sensor 23, and yaw rate sensor 24 are input to ECU 11.

Vehicle A is equipped with a hydraulic brake system. The brake system includes a brake pedal, a brake booster, a master cylinder, a brake actuator 25, and brakes provided on each wheel. The brake pedal is located in a position that is convenient for the driver in the driver's seat to step on it with his right foot. When the brake pedal is stepped on, the force input to the brake pedal is transmitted to the brake booster. The brake booster uses the negative pressure generated in the engine's intake system, and the brake pedal force is amplified by the pressure difference between the negative pressure and atmospheric pressure. The force amplified by the brake booster is transmitted from the brake booster to the master cylinder, and hydraulic pressure corresponding to the force is generated from the master cylinder. The hydraulic pressure of the master cylinder is transmitted to the brake actuator 25, and hydraulic pressure is supplied from the brake actuator 25 to the wheel cylinders of the brakes provided on each wheel, and the hydraulic pressure applies a braking force to the wheels from each brake. The brake actuator 25 also has an electric pump built in, and when the automatic brake is activated, the electric pump is driven by power from the battery, and the hydraulic pressure generated by the electric pump is supplied to each wheel cylinder.

Vehicle A is also equipped with an alarm 26. The alarm 26 outputs various types of alarms, which may be output by light, sound, or voice.

<Function Activation Processing>
FIG. 2 is a diagram showing the flow of the function activation process.

In vehicle A (see FIG. 4), the ECU 11 executes a function activation process to selectively activate the collision warning function, the primary braking function, and the secondary braking function, regardless of whether the vehicle is moving or stopped.

In the function activation process, first, a target T (see FIG. 4), such as a vehicle moving in front of vehicle A, is recognized from the output signal of camera 21, and information such as the moving direction and moving speed of the target T is obtained (step S1). If multiple targets T are recognized by camera 21, information such as the moving direction and moving speed of each target T is obtained. In addition, information such as the vehicle speed, steering angle, and yaw rate of vehicle A, which is the subject vehicle, is obtained from the detection signals of vehicle speed sensor 22, steering angle sensor 23, and yaw rate sensor 24 (step S1). The steering angle and yaw rate information are an example of turning information, which is information on the turning state of vehicle A.

Next, the future positions of the target object T and the vehicle A, which are the positions t seconds later (t>0), are estimated (step S2). The future positions are estimated for each target object T.

Next, a simple determination is made as to whether or not there is a possibility of collision for each target T (step S3). For targets T for which it is determined that there is no possibility of collision, collision prediction is not performed for each function (collision warning function, primary brake function, secondary brake function) described below.

The simplified determination of the possibility of collision is performed for each target T when the distance R1 between the future position of vehicle A (estimated position in t seconds) and the future position of each target T (estimated position in t seconds) is greater than a predetermined threshold Rx (R≧Rx). On the other hand, when the distance R1 is less than the predetermined threshold Rx (R<Rx), it is determined that there is a possibility of collision between the target T and vehicle A. This determination will be described later. When multiple targets T are recognized by the camera 21, the simplified determination of the possibility of collision is performed for each target T.

In addition to the simplified judgment based on the estimated distance between vehicle A and each target T after t seconds, if the direction of travel of target T is a direction away from the predicted path of vehicle A, it may be determined that there is no possibility of collision even if the estimated distance R1 between vehicle A and target T after t seconds is smaller than a predetermined threshold value Rx. This judgment will also be described later.

Then, based on the future positions of the target T and the vehicle A, a collision between the target T and the vehicle A is predicted for each of the collision warning function, the primary brake function, and the secondary brake function (step S4). Note that this collision prediction is performed for the target T that is determined to have a collision possibility in the simplified determination described above, but is not performed for the target T that is determined to have no collision possibility.

Then, based on the prediction result, it is determined whether to activate the collision warning function, the primary brake function, or the secondary brake function (step S5).

When it is determined that the collision warning function, the primary brake function, or the secondary brake function is to be activated, a command requesting activation of that function is sent to the brake actuator 25 or the alarm 26, along with a control instruction value for controlling the operation of the brake actuator 25 or the alarm 26 (step S6).

<Future position estimation>
Fig. 3 is a flowchart showing the flow of the future position estimation process. Fig. 4 is a diagram showing a vehicle A and a target T (another vehicle) located in front of the vehicle A. The estimation of the future positions of the vehicle A and the target T is performed, for example, at a period of 50 ms, and in each period, the future positions of the vehicle A and the target T are estimated at short time intervals from the current time until a certain time (for example, 5 seconds) ahead. In addition, when the camera 21 recognizes multiple targets T, the future positions of all the targets T are estimated.

The future positions of target T and vehicle A are estimated specifically as follows:

For example, in an orthogonal coordinate system with the origin at the current center C of vehicle A, the vehicle width direction (left and right direction) of vehicle A as the X-axis direction, and the traveling direction (front and back direction) as the Y-axis direction, the coordinates representing the current position of target T are (X'0, Y'0). Also, the X and Y components of the movement vector (velocity) of target T are VX and VY, respectively. In this case, the X coordinate X'(t) and Y coordinate Y'(t) of the future position of target T (position t seconds later) are:

X'(t)=X'0+VX*t...(1)
Y'(t)=Y'0+VY*t...(2)

The future position (X'(t), Y'(t)) of the target T is estimated using equations (1) and (2) (step S21).

The X coordinate X(t) and Y coordinate Y(t) of the future position of vehicle A are estimated using equations (3) and (4), respectively.

X(t)=R(1-cos(Δθ))+u(t)...(3)
Y(t)=R*sin(Δθ)+v(t)...(4)

Here, R is the turning radius of vehicle A, which is calculated by dividing the vehicle speed V of vehicle A by the yaw rate Yr according to equation (5). The yaw rate Yr takes a positive value when the turning direction is clockwise when viewed from above the center of gravity of vehicle A, and takes a negative value when the turning direction is counterclockwise.

R=V/Yr...(5)

Δθ is the rotation angle due to the turning of vehicle A, and is calculated by multiplying the yaw rate Yr of vehicle A by time t according to equation (6).

Δθ=Yr*t...(6)

When vehicle A is not stopped, that is, when vehicle A is traveling, the actual measured value of vehicle speed V can be obtained from the detection signal of vehicle speed sensor 22. Also, the actual measured value of yaw rate Yr can be obtained from the detection signal of yaw rate sensor 24. Therefore, when vehicle A is traveling (NO in step S22), the actual measured values of vehicle speed V and yaw rate Yr are substituted into equations (5) and (6), and the turning radius R and rotation angle Δθ obtained thereby are substituted into equations (3) and (4) (step S23), and the future position (X(t), Y(t)) of vehicle A is estimated (step S24).

However, while vehicle A is stopped, the actual measured values of vehicle speed V and yaw rate Yr cannot be obtained. Therefore, while vehicle A is stopped, a virtual vehicle speed, which is a virtual value of the vehicle speed after vehicle A starts moving, is set to vehicle speed V, and this virtual vehicle speed is substituted into equation (5). Also, a virtual yaw rate, which is a virtual value of the yaw rate after vehicle A starts moving, is set to yaw rate Yr, and this virtual yaw rate is substituted into equation (6). Then, the turning radius R and rotation angle Δθ obtained from this are substituted into equations (3) and (4) (step S25), and the future position (X(t), Y(t)) of vehicle A is estimated (step S24).

The virtual vehicle speed is stored as virtual vehicle speed information in the memory 14 of the ECU 11, and may be a predetermined vehicle speed value, or a vehicle speed value learned from the acceleration of vehicle A during past starts, and may be a constant value or a value that increases over time.

The virtual yaw rate is set, for example, by obtaining the steering angle from the detection signal of the steering angle sensor 23 while vehicle A is stopped, and calculating the steering angle and the virtual vehicle speed. The steering angle takes a positive value when the steering mechanism is turned to the right from the steering angle midpoint, and a negative value when it is turned to the left, so the virtual yaw rate takes a positive value when vehicle A turns to the right, and a negative value when it turns to the left. Therefore, the positive or negative virtual yaw rate represents the turning direction of vehicle A after it starts moving.

<Collision prediction>
5 is a diagram for explaining the vehicle area AA(t). Collision prediction is performed only for the object T that is determined to have a collision possibility in the above-mentioned simplified collision possibility determination. Collision prediction is performed at a predetermined cycle (e.g., 50 ms).

To predict a collision between target T and vehicle A, ECU 11, which is an example of a region setting means, sets a rectangular vehicle region AA(t) surrounding vehicle A located at a future position (X(t), Y(t)). The distance between the center C of vehicle A and each of the four vertices (corners) P1, P2, P3, and P4 of vehicle region AA(t) is r, and the angle between a line connecting center C and vertex P1 and a line extending from center C of vehicle A to the right in the vehicle width direction is θ. The X coordinate P1x(t) and Y coordinate P1y(t) of vertex P1 of the right front corner of vehicle region AA(t) are expressed by equations (7) and (8), respectively.

P1x(t)=X(t)+rcos(θ-Δθ)...(7)
P1y(t)=Y(t)+rsin(θ−Δθ) (8)

The X coordinate P2x(t) and Y coordinate P2y(t) of the left front corner vertex P2 are expressed by equations (9) and (10), respectively.

P2x(t)=X(t)+rcos(π-θ-Δθ)...(9)
P2y(t)=Y(t)+rsin(π-θ-Δθ)...(10)

The X coordinate P3x(t) and the Y coordinate P3y(t) of the left rear corner vertex P3 are expressed by equations (11) and (12), respectively.

P3x(t)=X(t)+rcos(π+θ−Δθ) (11)
P3y(t)=Y(t)+rsin(π+θ−Δθ) (12)

The X coordinate P4x(t) and the Y coordinate P4y(t) of the vertex P4 of the right rear corner are expressed by equations (13) and (14), respectively.

P4x(t)=X(t)+rcos(2π-θ-Δθ)...(13)
P4y(t)=Y(t)+rsin(2π-θ-Δθ)...(14)

The distance r is set variably according to each of the collision warning function, the primary brake function, and the secondary brake function. Specifically, the distance r is set to a smaller value as the urgency of the control for activating each of the collision warning function, the primary brake function, and the secondary brake function increases. Since the secondary brake function is intended to avoid a collision and reduce the damage caused by the collision, a delay in its activation may increase the damage caused by the collision, and the control for activating the secondary brake function has a high urgency. Since the primary brake function is intended to encourage the driver to take action to avoid a collision, a delay in its activation does not directly lead to damage caused by the collision, and therefore the control for activating the primary brake function has a lower urgency than the control for activating the secondary brake function. Since the collision warning function is intended to notify the driver of the possibility of a collision before the automatic brake is activated, the control for activating the collision warning function has a lower urgency than the control for activating the primary brake function. Therefore, for the secondary braking function, the distance r is set to a predetermined value r1, for the primary braking function, the distance r is set to a value r2 that is greater than the value r1, and for the collision warning function, the distance r is set to a value r3 that is even greater than the value r2.

The ECU 11 sets the distance r to r1, and calculates the coordinates (P1x(t), P1y(t)) of vertex P1, the coordinates (P2x(t), P2y(t)) of vertex P2, the coordinates (P3x(t), P3y(t)) of vertex P3, and the coordinates (P4x(t), P4y(t)) of vertex P4, each of which is a distance r1 away from the center C of vehicle A at its future position (X(t), Y(t)) t seconds later. It then determines whether the future position (X'(t), Y'(t)) of target T is included within the vehicle area AA(t) defined by the vertices P1 to P4, and if the future position (X'(t), Y'(t)) of target T is included within the vehicle area AA(t), it is determined that the secondary brake function is "predicted to collide (possible collision)" since it predicts the possibility of a collision between vehicle A and target T.

The ECU 11 sets the distance r to r2, and calculates the coordinates (P1x(t), P1y(t)) of vertex P1, the coordinates (P2x(t), P2y(t)) of vertex P2, the coordinates (P3x(t), P3y(t)) of vertex P3, and the coordinates (P4x(t), P4y(t)) of vertex P4, each of which is a distance r2 away from the center C of vehicle A at the future position (X(t), Y(t)). It then determines whether the future position (X'(t), Y'(t)) of target T is included within the vehicle area AA(t) defined by the vertices P1 to P4, and if the future position (X'(t), Y'(t)) of target T is included within the vehicle area AA(t), it is determined that the primary brake function is "predicted collision (possible collision)" since it predicts the possibility of a collision between vehicle A and target T.

The ECU 11 sets the distance r to r3, and determines the coordinates (P1x(t), P1y(t)) of vertex P1, the coordinates (P2x(t), P2y(t)) of vertex P2, the coordinates (P3x(t), P3y(t)) of vertex P3, and the coordinates (P4x(t), P4y(t)) of vertex P4, each of which is a distance r3 away from the center C of vehicle A at the future position (X(t), Y(t)). It then determines whether the future position (X'(t), Y'(t)) of target T is included within the vehicle area AA(t) defined by those vertices P1 to P4, and if the future position (X'(t), Y'(t)) of target T is included within the vehicle area AA(t), it is determined that the collision warning function predicts the possibility of a collision between vehicle A and target T, and determines that "collision predicted (possible collision)."

<Simple Determination of Possibility of Collision 1>
Next, the simplified determination of the possibility of collision will be described. As described above, for the object T recognized by the camera 21, a simplified determination of whether or not there is a possibility of collision with the object is performed before a collision prediction is performed to determine whether or not each function is to be activated.

In this embodiment, the future position (position t seconds later) of vehicle A and the future position (position t seconds later) of each target T recognized by the camera 21 are estimated. In the simplified judgment, the distance R1 between the estimated future position of vehicle A and the future position of the target T is calculated based on the future position of the target T. The calculation of the distance R1 is performed for all targets T recognized by the camera 21.

The future position of vehicle A (position t seconds later) is the future position of the center C of vehicle A (see Figure 5), and the future position of target T (position t seconds later) is the future position of the center of target T.

The X coordinate X(t) of the future position of vehicle A (the position of center C of vehicle A t seconds later) is calculated using equation (3), and the Y coordinate Y(t) is calculated using equation (4).

The X coordinate X'(t) of the future position of the target T (the position of the center of the target T t seconds later) is calculated from equation (1), and the Y coordinate Y'(t) is calculated from equation (2). Here, the coordinates (X'0, Y'0) representing the current position of the target T can be estimated based on the type of the target T and the model shape stored according to that type. Specifically, for the target T, the type of the target T, such as a person, a vehicle, or a two-wheeled vehicle, is determined from the image of the camera 21. In addition, the model shape according to the type of the target is stored in the memory 14 of the microcomputer 12, and the coordinates (X'0, Y'0) indicating the current position of the center of the target T can be estimated based on the type of the target T estimated from the image of the camera 21 and the model shape stored according to that type.

The center can also be estimated based on the type of target T estimated from the image of camera 21 and the partial position of the target T that can be recognized by camera 21. For example, if target T is estimated to be a four-wheeled vehicle and camera 21 can recognize three points, the front left end, rear right end, and rear left end, but cannot recognize the front right end, the coordinates of the remaining point (front right end) are estimated based on these three points, and the center coordinates of the four-wheeled vehicle are estimated based on the three points recognized by camera 21 and the remaining estimated point. The method of estimating the center position of target T can be changed as appropriate.

The distance R1 between the future position of vehicle A (position t seconds later) and the future position of target T (position t seconds later) is expressed by equation (15).

R=√((X(t)-X'(t)) ² +(Y(t)-Y'(t)) ² )...(15)

Then, when the calculated value of R is smaller than a predetermined threshold value Rx (R<Rx), the ECU 11 determines that there is a possibility of a collision between vehicle A and target T. In this embodiment, the threshold value Rx can be set to the distance required from the timing of automatic braking to the timing of stopping vehicle A when a normal automatic brake (e.g., secondary brake) is activated while vehicle A is traveling at 80 km/h, for example, but the distance can also be made variable depending on the current vehicle speed of vehicle A.

In this embodiment, it is configured to determine that there is a possibility of a collision between vehicle A and target T when the distance R1 between the future position of vehicle A (position in t seconds) and the future position of target T (position in t seconds) is smaller than a predetermined threshold value Rx, but it may also be configured to determine that there is no possibility of a collision between vehicle A and target T when the distance R1 between the current position of vehicle A and the current position of target T is larger than a predetermined threshold value Rx.

The future positions of vehicle A and target T are calculated at multiple times, and at each time, the calculated distance R1 is compared with a predetermined threshold Rx to determine whether or not there is a possibility of collision. For example, even for the same target T, if the distance R1 between the future positions of vehicle A and target T (position at t1) is smaller than the threshold Rx at time t1, it is determined that there is a possibility of collision, and then a collision prediction for each function is performed, but if the distance R1 between the future positions of vehicle A and target T (position at t2) is larger than the threshold Rx at time t2, it is determined that there is no possibility of collision, and then a collision prediction for each function is not performed.

<Simple collision possibility determination 2>
Even if it is determined that there is a possibility of a collision in the above-mentioned simplified collision possibility judgment 1 (distance R1 between the future position of vehicle A and the future position of target T < threshold value Rx → possibility of collision), if it is determined that the direction of travel of target T is a direction away from the predicted path of vehicle A, collision prediction may not be performed for each function (collision warning function, primary brake function, secondary brake function).

The ECU 11 calculates a predicted path of the vehicle A from the vehicle speed, steering angle, yaw rate, etc. of the vehicle A. The ECU 11 also sets a possible presence region M (see FIG. 7) of the vehicle A that is determined according to the known vehicle width, current steering angle, yaw rate, etc. of the vehicle A. For example, the possible presence region M is set based on a map stored in the memory 14. The possible presence region M is a region that includes the predicted path of the vehicle A.

The ECU 11 also derives the traveling direction of the target T, and determines whether there is a point where a straight line that is parallel to the traveling direction and passes through the position of the target T passes through the set possibility of existence region M.
Here, if there is a point passing through the presence possibility region M, the ECU 11 determines that there is a possibility of collision between the vehicle A and the target T, and if there is no point passing through the presence possibility region M, it presumes that the target T is traveling in a direction away from the vehicle A, and determines that there is no possibility of collision between the vehicle A and the target T. If the camera 21 recognizes multiple targets T, this determination is made for all targets T. A case in which it is determined that there is no possibility of collision between the vehicle A and the target T is when, for example, the target T is traveling in a leftward direction as viewed from the vehicle A while the vehicle A is turning right.

The traveling direction of the target T can be derived based on the most recent position of the target T calculated at a predetermined period (e.g., every 50 ms) and the position of the target T calculated at the previous period, but the traveling direction of the target T may also be derived based on the position of the target T going back even further.

Even if ECU 11 determines that there is a possibility of a collision based on the distance R1 between the future position of vehicle A and the future position of target T, if the target T is moving away from vehicle A, it determines that there is no possibility of a collision and excludes it from the targets for collision prediction for each function (collision warning function, primary brake function, secondary brake function).

<Operation Function Determination Process>
FIG. 6 is a flowchart showing the flow of the operational function determination process.

Based on the result of the collision prediction, the ECU 11 executes an activation function determination process to determine whether to activate the collision warning function, the primary brake function, or the secondary brake function. If the camera 21 recognizes multiple targets T, the determination process is executed only for targets T that are determined to have a collision possibility in the simplified collision possibility determination described above.

In the activation function decision process, if it is determined that a collision is predicted when the distance r = r1 (YES in step S41), it is decided to activate the secondary brake function (step S42).

If it is not determined that a collision is predicted when the distance r = r1 (NO in step S41), and if it is determined that a collision is predicted when the distance r = r2 (YES in step S43), it is decided to activate the primary brake function (step S44).

If it is not determined that a collision is predicted when the distance r is either r1 or r2 (NO in step S43), but a collision is predicted when the distance r is r3 (YES in step S45), it is determined that the collision warning function is activated (step S46).

If it is not determined that a collision is predicted when the distance r is r1, r2, or r3 (NO in step S45), it is determined that none of the collision warning function, primary brake function, and secondary brake function are to be activated (step S47).

<Action and effect>
As described above, according to the embodiment, collision prediction is not performed for all targets T recognized by the camera 21, but only for targets T that are determined to have a collision possibility through a simple collision possibility determination, thereby reducing the processing load on the ECU 11 for collision prediction.

In addition, by limiting the object T that is the target of the collision prediction, it is possible to perform a more accurate collision prediction calculation. Here, a more accurate collision prediction calculation can be achieved by shortening the calculation cycle for the future positions of the vehicle A and the object T (e.g., calculation cycle = 50 ms). In addition, by limiting the object T that is the target of the collision prediction, it is possible to reduce the memory capacity required to store variables used in the collision prediction, for example.

Furthermore, in the simplified determination of the possibility of a collision, even if the distance R1 (distance of the future position) between vehicle A and target T is smaller than a predetermined threshold value Rx, if a part of a straight line that is parallel to the traveling direction of the target T and passes through the center of the target does not pass through the possible presence area M of vehicle A, it is estimated that the target T is traveling in a direction away from the predicted path of vehicle A, and it is determined that there is no possibility of a collision (see simplified determination 2 of the possibility of a collision). In this case, even if the traveling direction of the target T is away from the predicted path of vehicle A, the target T is excluded from the targets for collision prediction, so that the processing load of the ECU 11 for collision prediction can be further reduced.

In addition, in order to selectively activate the collision warning function, the primary brake function, and the secondary brake function, when vehicle A is stopped, the future position (X(t), Y(t)) of vehicle A after starting from a stopped state is predicted based on the virtual vehicle speed and the virtual yaw rate. By using the virtual vehicle speed and the virtual yaw rate to predict the future position (X(t), Y(t)), it is possible to accurately predict the future position (X(t), Y(t)). The time series of the future position represents the path of vehicle A after it starts. Therefore, even if the position of vehicle A does not change while it is stopped due to waiting to turn right at an intersection or stopping temporarily, it is possible to accurately predict the path of vehicle A when vehicle A starts. Therefore, it is possible to predict a collision between vehicle A and target T on that path.

To predict a collision between vehicle A and target T, specifically, the future position (X(t), Y(t)) of vehicle A after a predetermined time is predicted, and a vehicle area AA(t) surrounding the predicted future position (X(t), Y(t)) is set. The vehicle area AA(t) is set to a size corresponding to each of the collision warning function, the primary brake function, and the secondary brake function. Meanwhile, the future position (X'(t), Y'(t)) of target T after a predetermined time is predicted. Then, it is determined whether the future position (X'(t), Y'(t)) of target T is included within the vehicle area AA(t), and if the future position (X'(t), Y'(t)) of target T is included within the vehicle area AA(t), the function corresponding to the size of the vehicle area AA(t) is activated from among the collision warning function, the primary brake function, and the secondary brake function. This allows the collision warning function, primary braking function, and secondary braking function to be activated at the appropriate time.

The collision warning function, the primary brake function, and the secondary brake function are classified according to the urgency of their activation, and the size of the vehicle area AA(t) is set smaller as the urgency increases. This allows the collision warning function, the primary brake function, and the secondary brake function to be activated at appropriate timing according to the urgency of their activation. That is, for functions with a relatively low urgency of activation, i.e., the collision warning, the size of the vehicle area AA(t) is set large, so that they can be activated at an early stage when the vehicle A and the target T approach each other. On the other hand, for functions with a relatively high urgency of activation, i.e., the primary brake function and the secondary brake function, the size of the vehicle area AA(t) is set small, so that they can be activated at a stage when the possibility of a collision between the vehicle A and the target T has increased.

<Modification>
Although one embodiment of the present invention has been described above, the present invention can be embodied in other forms.

For example, in the above embodiment, a configuration was taken up in which the distance r between the center C of vehicle A and each of the four vertices (corners) P1, P2, P3, and P4 of the vehicle area AA(t) is set to a constant value r1, r2, and r3 for the secondary brake function, the primary brake function, and the collision warning function, respectively. However, the distance r may be set variably depending on the behavior of vehicle A (turning conditions, acceleration/deceleration conditions, etc.).

As an example, when vehicle A turns right, it is assumed that vehicle A will move to the right, so the distance r between center C and each of vertices P1 and P4 may be set to a relatively large value, and the distance r between center C and each of vertices P2 and P3 may be set to a relatively small value. When vehicle A turns left, it is assumed that vehicle A will move to the left, so the distance r between center C and each of vertices P2 and P3 may be set to a relatively large value, and the distance r between center C and each of vertices P1 and P4 may be set to a relatively small value.

When vehicle A accelerates, the distance r between center C and each of vertices P1 and P2 may be set to a relatively large value, and the distance r between center C and each of vertices P3 and P4 may be set to a relatively small value. When vehicle A decelerates, the distance r between center C and each of vertices P3 and P4 may be set to a relatively large value, and the distance r between center C and each of vertices P1 and P2 may be set to a relatively small value.

In the above embodiment, the future position (X(t), Y(t)) of vehicle A is estimated using the actual measured values of vehicle speed V and yaw rate Yr while vehicle A is traveling, and the future position (X(t), Y(t)) of vehicle A is estimated using the virtual vehicle speed and virtual yaw rate while vehicle A is stopped. However, even without using the yaw rate Yr and the virtual yaw rate, if the turning direction of vehicle A is actually measured or predicted and the turning direction and the vehicle speed of vehicle A are used, the accuracy of the estimation decreases, but for example, the future position (X(t), Y(t)) of vehicle A can be estimated by setting the turning radius R and the rotation angle Δθ to predetermined fixed values. The future position (X(t), Y(t)) of vehicle A can also be estimated using the actual measured value or virtual value of the steering angle of vehicle A and the vehicle speed of vehicle A.

The driving assistance function has been exemplified by the collision warning function, the primary braking function, and the secondary braking function. Other examples include a lane keeping assist (LKA) function that keeps vehicle A in the lane it is traveling in, an adaptive cruise control (ACC) function that makes vehicle A follow the vehicle ahead, a parking assistance function that assists with steering when parking vehicle A in a parking position, and a blind spot monitoring (BSM) function that monitors the driver's blind spots.

Furthermore, the above-mentioned embodiments and variations can be applied not only to gasoline vehicles, but also to electric vehicles, hybrid vehicles, and self-driving vehicles.

In addition, various design modifications may be made to the above-mentioned configuration within the scope of the claims.

1: Driving assistance system 11: ECU (driving assistance device, vehicle position prediction means, area setting means, target position prediction means, judgment means, function activation means, target travel direction deriving means)
A: Vehicle (own vehicle)
T: Target

Claims (1)

A driving assistance device that assists driving of a vehicle by a driving assistance function,
a vehicle position prediction means for predicting a future position of the host vehicle after a predetermined time;
an area setting means for setting a vehicle area surrounding the future position;
a target position prediction means for predicting a future position of a target after the predetermined time;
a determination means for determining whether or not the future position of the target predicted by the target position prediction means is included within the vehicle area set by the area setting means;
a function activation means for activating the driving assistance function when the determination means determines that the future position of the target is included within the vehicle area;
a target travel direction deriving means for deriving a travel direction of the target ,
The determination means is
performing a first simplified judgment in which the target is subject to the judgment when the distance between the vehicle and the target is within a preset predetermined distance, but is not subject to the judgment when the target is not within the preset distance;
a driving assistance device which performs a second simplified judgment in which, when it is determined in the first simplified judgment that the target should be subject to the judgment, if there is no point where a straight line which is parallel to the direction of travel of the target derived by the target travel direction derivation means and passes through the position of the target passes through a possible presence area which is an area including a predicted path of the vehicle, the target is not subject to the judgment, but if there is a point where it passes through the possible presence area, the target is subject to the judgment .

Images