Vehicle driving assistance system, driving assistance method, and recording medium

Abstract

To provide a vehicle driving support device capable of predicting the possibility of collision and collision position in driving assistance when a vehicle is driving while turning.SOLUTION: The vehicle driving support device is configured to: calculate a yaw rate change rate based on a yaw rate of the own vehicle obtained at multiple times; predict a sequential position of the own vehicle as an own-vehicle position assuming that the yaw rate change rate is kept constant, and predict a sequential position of another vehicle as an other-vehicle position; determine the possibility of a collision depending on whether there is a collision position where the own-vehicle position partially overlaps with the other-vehicle position; and when determining that there is no possibility of collision until the end of the prediction as the point at which the own vehicle turns at a set angle, terminate the sequential prediction of the own-vehicle position and the other-vehicle position.SELECTED DRAWING: Figure 4

Description

Technical Field

[0005]

[0001] The present invention relates to a driving support device, a driving support method, and a recording medium for determining the presence or absence of a vehicle collision.

Background Art

[0002] A technique for predicting the collision position between a host vehicle and another vehicle and performing driving support is known. In the driving support device of Patent Document 1, it is assumed that the host vehicle maintains the current vehicle speed and travels straight, and the other vehicle maintains the current speed. When it is predicted that the host vehicle and the other vehicle will collide in an intersection area where the area expected to be passed by the other vehicle and the area expected to be passed by the host vehicle overlap, the other vehicle is specified as a crossing target. Then, the braking device of the host vehicle is controlled so that the host vehicle decelerates at a deceleration from a time point before the predicted collision time point, which is the time point when it is predicted that the host vehicle and the crossing target will collide.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] The collision determination device of Patent Document 1 is effective when the host vehicle is traveling straight. However, there is a problem that it is difficult for the collision determination to function when the host vehicle is turning and traveling at a curve or an intersection.

Means for Solving the Problems

[0005] A vehicle driving assistance device in one embodiment of the present invention calculates the yaw rate change rate from the yaw rate of the own vehicle obtained at multiple points in time, predicts the position of the own vehicle sequentially as the own vehicle position, and predicts the position of other vehicles sequentially as other vehicle positions, determines whether or not there is a collision based on whether or not there is a collision point where the sequential own vehicle position partially overlaps with the other vehicle position, and terminates the prediction of the sequential own vehicle position and other vehicle position when it is determined that there is no collision up to the prediction termination point, which is the point in time when the own vehicle turns to a set angle. [Effects of the Invention]

[0006] The position of the vehicle where the yaw rate change rate is maintained is represented by a clothoid curve. Then, by using the vehicle's position as the successive vehicle position and predicting from the prediction reference point to the prediction end point, which is the point at which the vehicle turns to a set angle, it is possible to determine whether or not a collision will occur based on the presence or absence of a collision point with another vehicle. [Brief explanation of the drawing]

[0007] [Figure 1] A driver assistance system installed in the vehicle. [Figure 2] A diagram showing the predicted travel paths of your vehicle and other vehicles. [Figure 3] A flowchart of the driver assistance process. [Figure 4] Flowchart of the collision detection process. [Figure 5] A diagram illustrating the calculation of the yaw rate change rate. [Figure 6] A diagram showing a driving path where no collisions are predicted up to the end of the prediction process. [Modes for carrying out the invention]

[0008] Figure 1 shows a vehicle equipped with the driver assistance device 1 according to an embodiment of the present invention. To distinguish it from other vehicles VS that are the target of collision determination, the vehicle is shown as the own vehicle VM. The driver assistance device 1 of the own vehicle VM has an ECU 11, a camera 12, a radar sensor 13, and wiring 14 connecting the ECU 11 to the camera 12 and radar sensor 13. Two cameras 12 are installed side by side in the left-right direction of the vehicle, and one radar sensor 13 is installed on each side at the front of the vehicle. The ECU 11 is also connected to the braking device 2 of the own vehicle VM via communication wiring 3.

[0009] The video footage captured by the two cameras 12 and the data detected by the two radar sensors 13 are sent to the ECU 11 via wiring 14 and analyzed by the CPU 111 using software stored in memory 112. The CPU 111 and memory 112 form a computer. When the ECU 11 determines that collision response is necessary, it takes action such as sending a braking command to the braking system 2 via communication wiring 3.

[0010] Figure 2 shows the predicted trajectories of the vehicle VM and the other vehicle VS when determining a collision. The solid lines of the vehicle VM and the other vehicle VS indicate the current position at the prediction reference time Ts. The dotted lines of the vehicle VM1 and the other vehicle VS1 indicate the predicted positions of the vehicle VM and the other vehicle VS after Δt time. The dotted lines of the vehicle VMn and the other vehicle VSn indicate the predicted positions of the vehicle VM and the other vehicle VS at the predicted collision time. The curves show the predicted trajectories of various points on the vehicle VM. MR, MC, and ML are the predicted trajectories of the front right edge Mfr, front center Mfc, and front left edge Mfl of the vehicle VM, respectively. SR and SL are the predicted trajectories of the front right edge Sfr and front left edge Sfl of the other vehicle VS, respectively. In this invention, it is predicted that the vehicle VM will travel along a clothoid curve. The predicted trajectory MC inside the vehicle is a clothoid curve. Furthermore, the vehicle's right-side predicted trajectory (MR) and left-side predicted trajectory (ML) can be obtained from the vehicle's width (VM) and its center-side predicted trajectory (MC).

[0011] Next, the control process for collision response actions through driver assistance will be explained using the flowcharts in Figures 3 and 4. The steps in the flowcharts indicate the driver assistance methods. This control is executed by processing by the CPU 111 using a program stored in memory 112. This program can also be stored on a computer-readable recording medium.

[0012] The driver assistance process shown in Figure 3 is started at predetermined time intervals and executed repeatedly. In step S1, it is determined whether the target vehicle VS has been recognized. The video footage captured by the two cameras 12 and the data detected by the two radar sensors 13 are analyzed by the ECU 11. If it is determined that the target vehicle VS has been recognized, the result in step S1 is YES and the process proceeds to step S2. If it is not determined that the vehicle has been recognized, the result is NO and the process ends.

[0013] The vehicle VM calculates the yaw rate Yr from the detected vehicle VM speed, lateral acceleration, and steering wheel rotation angle. In step S2, if the absolute value of the yaw rate Yr remains above or below the turning reference value for a predetermined period (YES), the process proceeds to step S3 to perform collision detection during turning. Otherwise, the process ends (NO) and waits for the start time of the next driving assistance process.

[0014] Figure 4 illustrates the flow of the collision detection process in step S3. In step S31, the yaw rate change rate Yrc is calculated from the yaw rate Yr of the vehicle VM at multiple time points. In this embodiment, the equation of the straight line L is calculated by the least squares method from the yaw rates Yr at a total of three time points: two past time points shown as black dots in Figure 5 and the current time point t=0. The yaw rate change rate Yrc is the slope of the straight line L. The yaw rate Yr may be calculated using values ​​from four or more time points, or it may be calculated by connecting the yaw rates Yr at two time points. If the straight line L is sloped, that is, if the yaw rate change rate Yrc is not 0, the predicted trajectory will be a clothoid curve. Then, by using the equation of the straight line L, assuming that the yaw rate change rate Yrc is maintained after the current time point Ts, the position of the vehicle VM is calculated sequentially as the vehicle position, and the predicted trajectory MC, which is the predicted trajectory of the vehicle VM, can be obtained.

[0015] In the next step, S32, the point in time when the vehicle VM rotates to the set angle of 90° while the yaw rate change rate Yrc is maintained is calculated and set as the prediction end time Te. In this embodiment, prediction is only performed until the vehicle VM has rotated 90°. In this embodiment, the set angle is 90°, but other angles may also be used.

[0016] In the next step, S33, the image obtained from the camera 12 is recognized using data detected by the radar sensor 13, and the rectangle occupied by the other vehicle VS on the horizontal plane is determined as the other vehicle rectangle. The size and shape of the other vehicle rectangle differ depending on whether the image is recognized as a small vehicle or a large vehicle. Multiple other vehicle rectangles are stored in memory 112, and the other vehicle rectangle is selected based on the image recognition. In addition, the position, speed, and direction of the other vehicle VS are calculated by recognizing the image at multiple points in time. The rectangle occupied by the own vehicle VM on the horizontal plane is stored as a set value in memory 112, and the current speed is also stored in memory 112 through measurement.

[0017] In the next step S34, the locations of the host vehicle VM and the other vehicle VS at the predicted time point t after Δt time are calculated. The host vehicle position of the host vehicle VM and the other vehicle position of the other vehicle VS are represented as rectangles as shown in FIG. 2. The coordinates at this time are based on the position and direction of the host vehicle VM at the current time as shown in FIG. 2. Then, from the predicted yaw rate Yr assuming that the speed and the yaw rate change rate Yrc of the host vehicle VM are constant, the host vehicle rectangle of the host vehicle VM at the predicted time point t after Δt time is calculated. For the host vehicle rectangle, and for the other vehicle VS, the other vehicle rectangle after Δt time is calculated assuming a constant linear motion using the calculated position, speed, and direction. In this way, the rectangles of the host vehicle VM1 and the other vehicle VS1 at the predicted time point t after Δt time from the current time, whose frames are shown by dotted lines in FIG. 2, are predicted as the host vehicle position and the other vehicle position.

[0018] In the next step S35, it is determined whether there is an overlapping part between the host vehicle rectangle, which is a rectangle showing the host vehicle position of the host vehicle VM, and the other vehicle rectangle, which is a rectangle showing the other vehicle position of the other vehicle VS. Based on this determination, the presence or absence of a collision is determined according to the presence or absence of a collision position where the host vehicle position contacts the other vehicle position. The host vehicle rectangle is represented by four linear equations with limited ranges. The same applies to the other vehicle rectangle. Therefore, for example, it is possible to determine whether there is an overlapping part by determining whether there is a solution to the simultaneous equations of the linear equations of the host vehicle rectangle and the linear equations of the other vehicle rectangle. Since there is no overlapping part between the host vehicle VM1 and the other vehicle VS1 in FIG. 2, it is NO, and the process proceeds to step S36.

[0019] In step S36, it is determined whether the predicted time point t has reached the prediction end time point Te. If it has not reached the prediction end time point Te (NO), the process returns to step S34. In this way, by repeating steps S34 to S36, the determination of whether the rectangles of the host vehicle position of the host vehicle VM and the other vehicle position of the other vehicle VS overlap at each successive predicted time point t after Δt time is repeated. The repetition is performed until the rectangles of the host vehicle position and the other vehicle position overlap or the prediction end time point Te elapses. <00,00092> Until the prediction end time Te has elapsed, if there is no overlap between the host vehicle rectangle and the other vehicle rectangle, YES is determined in step S36, and step S3 of the collision determination process ends. Fig. 6 shows a prediction diagram of a driving route with no rectangle overlap until the prediction end time Te. The host vehicle VMe and the other vehicle VSe show the rectangles of the host vehicle position and the other vehicle position at the prediction end time Te. The host vehicle VM and the other vehicle VS shown by the solid line indicate the current positions. The host vehicle VM1 and the other vehicle VS1 shown by the dotted line are the predicted positions of the host vehicle VM and the other vehicle VS after Δt time. The curve shows the predicted trajectory of each part of the host vehicle VM. ML and MR are the predicted trajectories of the left front end and the right front end of the host vehicle VM, respectively. Also, Sfl and Sfr are the predicted trajectories of the left front end and the right front end of the other vehicle VS, respectively.

[0021] On the other hand, when the rectangles of the host vehicle position and the other vehicle position overlap like the host vehicle VMn and the other vehicle VSn in Fig. 2, YES is determined in step S35 in Fig. 4, and the process proceeds to step S37. In step S37, it is memorized that there is a collision, and step S3 of the collision determination process ends. <​​​​​​Once step S3 is completed, the process proceeds to step S4 in Figure 3. In step S4, it is determined whether there is a memory of a collision and whether the urgency of the response is high. For example, if the collision point where the rectangles overlap occurs more than a predetermined time after the current time, which is the predicted reference time Ts, then there is sufficient time until the collision, and the automatic collision response operation is not performed. Also, if there is no memory of a collision, the automatic collision response operation is not performed. If the urgency in step S4 is determined to be NO, the process ends and waits until the start of the next driving assistance process. If it is determined to be YES, the collision response operation is performed. In this embodiment, as a collision response operation, a braking command is sent to the braking device 2 to activate the brakes.

[0024] In this embodiment, the set angle was set to 90°, and the prediction end time was defined as the point at which the vehicle turns to 90°, but other angles may also be used. Preferably, the set angle is set to 45° or more and 180° or less, and the prediction end time is defined as the point at which the vehicle turns to the set angle.

[0025] Furthermore, in this embodiment, the position of other vehicles is detected using information obtained from the camera 12 and the radar sensor 13, but the position of other vehicles may also be detected using information obtained from either the camera 12 or the radar sensor 13. The position of other vehicles may also be detected using information obtained from all or part of the camera 12, the radar sensor 13, and other devices.

[0026] In this embodiment, in step S2, the yaw rate Yr, which determines whether or not to start collision detection processing by comparing it with a turning reference value, is calculated from the detected speed and lateral acceleration of the vehicle VM and the steering wheel rotation angle. The yaw rate Yr used to calculate the yaw rate change rate Yrc in step S31 is calculated similarly. However, both or either of these yaw rates Yr may be calculated by correcting them with the yaw rate change rate Yrc calculated from past yaw rates Yr.

[0027] Furthermore, the specific configuration is not limited to the embodiments, and any design changes, etc., that do not depart from the spirit of the present invention are also included. [Explanation of Symbols]

[0028] VM (Virtual Vehicle) VS Other vehicles t Prediction time Ts forecast base date Te prediction end point Yr yaw rate Yrc (York Rate of Change) Mfr front right edge Mfc front center Mfl Front left edge MR (Mid-Rear View) Predicted Trajectory of Vehicle to the Right MC (Motorcycle Prediction Trajectory) ML (Multi-Level) Predicted Trajectory of Vehicle (Left Side) Sfr Front right edge Sfl Front left edge SR Other vehicle's predicted trajectory to the right SL Other vehicles left predicted trajectory 1. Driving assistance system 11 ECU 111 CPU 112 memory 12 cameras 13 Radar Sensors 14 Wiring 2 Braking device 3. Communication wiring

Claims

1. The yaw rate change rate is calculated from the yaw rate of the vehicle obtained at multiple points in time. Assuming that the yaw rate change rate is kept constant, the position of the own vehicle is predicted sequentially as the own vehicle's position, and the position of other vehicles is predicted sequentially as the other vehicle's position. The presence or absence of a collision is determined by whether or not there is a collision point where the sequential position of the own vehicle partially overlaps with the position of the other vehicle. A vehicle driving assistance device characterized in that, if it is determined that there will be no collision until the prediction end point, which is the point at which the vehicle turns to a set angle, the sequential prediction of the vehicle's position and the position of the other vehicle is terminated.

2. The vehicle driving assistance device according to claim 1, characterized in that the rate of change of the yaw rate is obtained from the yaw rate obtained at two points in time.

3. The vehicle driving assistance device according to claim 1, characterized in that the rate of change of the yaw rate is obtained by the least squares method from the yaw rate obtained at three or more time points.

4. A vehicle driving assistance device according to any one of claims 1 to 3, characterized in that the self-vehicle position is a rectangle on the horizontal plane of the self-vehicle, and the other vehicle position is a rectangle on the horizontal plane of the other vehicle.

5. A vehicle driving assistance device according to any one of claims 1 to 3, characterized in that it predicts the position of the other vehicle assuming that the other vehicle is moving in a straight line at a constant velocity.

6. On the computer, The steps include: calculating the yaw rate change rate from the yaw rate of the vehicle obtained at multiple points in time; Assuming that the yaw rate change rate is kept constant, the steps include predicting the position of the own vehicle sequentially as the own vehicle's position, and predicting the position of other vehicles sequentially as other vehicles' positions, The steps include determining whether a collision has occurred based on whether there is a collision point where the sequential position of the own vehicle partially overlaps with the position of the other vehicle, A vehicle driving assistance method characterized in that, if it is determined that there will be no collision up to the prediction end point, which is the point when the vehicle turns to a set angle, the step of terminating the sequential prediction of the vehicle's position and the position of the other vehicle is executed.

7. The steps include: calculating the yaw rate change rate from the yaw rate of the vehicle obtained at multiple points in time; Assuming that the yaw rate change rate is kept constant, the steps include predicting the position of the own vehicle sequentially as the own vehicle's position, and predicting the position of other vehicles sequentially as other vehicles' positions, The steps include determining whether a collision has occurred based on whether there is a collision point where the sequential position of the own vehicle partially overlaps with the position of the other vehicle, A computer-readable recording medium storing a program that causes a computer to execute the step of terminating the sequential prediction of the position of the own vehicle and the position of the other vehicle if it is determined that there will be no collision up to the prediction end point, which is the point when the own vehicle turns to a set angle.

Images