JP2008302904A - Collision predictive device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a collision predictive device attaining enhancement of collision predictive accuracy. <P>SOLUTION: Target points P1-P5 of a plurality of positions of the other vehicle 104 are detected respectively, and a collision form of the one's own vehicle 102 and the other vehicle 104 is predicted based on the collision portion predicted regarding the respective target points P1-P5. Thereby, since the target points P1-P5 of the plurality of positions are detected respectively and the collision portion is predicted regarding the respective target points P1-P5, not only collision of the limited point (conventional representative point) of the other vehicle 104 but also collision of which portion of the one's own vehicle 102 can be predicted regarding a plurality of positions of the other vehicle 104. Accordingly, in conventional grouping processing, the collision predictive accuracy can be enhanced by effectively utilizing member P2, P4, P5 information comprehended in the representative points P1, P3 and not target-outputted. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to a collision prediction apparatus that predicts a vehicle collision.

Conventionally, a technique has been proposed that includes a radar device that detects an object such as another vehicle and predicts a collision between the detected object and the host vehicle (for example, Patent Document 1). In the technique described in Patent Document 1, based on the TTC (Time To Collation) of a representative point having the shortest distance from the host vehicle among a plurality of points in the target detected by the radar device, the target and the host vehicle It is known to perform a vehicle collision determination according to the result of the collision determination.
JP 2003-2322853 A

  A radar apparatus (for example, a millimeter wave radar) used for such collision determination detects an object (hereinafter also referred to as “target”), calculates an approximate curve from the target output result, and automatically Predicts the location of a collision with a car. FIG. 7 is a model diagram for explaining grouping representative points and grouping members in the grouping process at the time of collision determination using a conventional radar apparatus. In the prior art, a grouping representative point (hereinafter referred to as “representative point”) P101 having a high priority based on conditions such as TTC among a plurality of detected targets, and the same speed as the representative point P101. The moving targets are grouping members (hereinafter referred to as “members”) P102 and P103, and information on the members P102 and P103 is included in the information on the representative point P101. In this way, by performing the grouping process in which the information of the members P102 and P103 is included in the information of the representative point P101, the target number (information amount) is reduced and necessary information is efficiently transmitted with a small number of transmissions. It was. In the prior art, collision prediction using only representative points is required, and improvement in prediction accuracy is required.

  The present invention has been made to solve such a problem, and an object of the present invention is to provide a collision prediction apparatus that improves the accuracy of collision prediction.

  The collision prediction apparatus according to the present invention is based on a detection unit that detects target points at a plurality of positions in another vehicle, and a time series change in position information of each target point at a plurality of positions and a comparison between position information of the host vehicle. Target point collision site prediction means for predicting the collision site of each target point with respect to the own vehicle, and collision mode prediction means for predicting the collision mode between the host vehicle and other vehicles based on the prediction results of the collision site at a plurality of positions It is characterized by providing these.

  Such a collision prediction device detects target points at a plurality of positions of other vehicles, and predicts a collision mode between the own vehicle and the other vehicles based on the collision site predicted for each target point. As a result, target points at a plurality of positions are detected and the collision site is predicted for each target point, so that not only a limited number of other vehicles (conventional representative points) but also a plurality of other vehicles It is possible to predict which part of the vehicle will collide with the position. Therefore, in the conventional grouping process, the collision prediction accuracy can be improved by effectively utilizing the member information that is included in the representative points and has not been output as a target.

  Here, it is preferable that the collision mode prediction means predicts a collision site of the host vehicle with another vehicle as the collision mode. For example, it is preferable to predict whether the collision site is the front portion or the rear portion of the host vehicle.

  Moreover, it is preferable that a collision aspect prediction means predicts the magnitude | size of the collision surface with the other vehicle in the own vehicle as a collision aspect. For example, when the vehicle collides with the central portion of the host vehicle, the collision surface is predicted to be large, and when the vehicle collides with the side portion of the host vehicle, the collision surface is predicted to be small.

  Moreover, it is preferable that the collision mode prediction unit predicts the collision mode based on the position distribution of the collision site predicted by the target point collision site prediction unit.

  Moreover, it is preferable that a collision mode prediction means predicts a collision mode based on the coefficient according to the prediction result by the target point collision site | part prediction means.

  The coefficient is preferably different between when the collision site is located closer to the center of the host vehicle and when closer to the outer side of the host vehicle. This makes it possible to make a determination in consideration of the magnitude of damage and ease of avoidance. For example, it can be predicted that the collision surface is smaller and the damage of the own vehicle is less when the vehicle is collided with the outside of the vehicle than when the vehicle is collided with the center of the own vehicle, and the collision is avoided. It can be determined that there is a high possibility of being able to do it.

  According to the collision prediction apparatus of the present invention, it is possible to improve the collision prediction accuracy.

  Hereinafter, a preferred embodiment of a collision prediction apparatus according to the present invention will be described with reference to the drawings. In the description of the drawings, the same or corresponding elements are denoted by the same reference numerals, and redundant description is omitted. FIG. 1 is a block diagram showing a collision prediction apparatus according to an embodiment of the present invention.

  A collision prediction apparatus 100 illustrated in FIG. 1 includes a collision determination electronic control unit (hereinafter referred to as “collision determination ECU”) 10 that predicts a collision between another vehicle 104 (see FIG. 2) and the host vehicle 102. The collision determination ECU 10 is electrically connected to the front millimeter wave radar 12, the front side millimeter wave radar 14, the steering angle sensor 16, the yaw rate sensor 18, and the vehicle speed sensor 20. The collision determination ECU 10 determines whether the own vehicle 102 and the other vehicle 104 are based on data detected by the front millimeter wave radar 12, the front side millimeter wave radar 14, the steering angle sensor 16, the yaw rate sensor 18, and the vehicle speed sensor 20. Predict the collision mode. Details thereof will be described later.

  The forward millimeter wave radar 12 is provided at the front center of the host vehicle 102, and the front side millimeter wave radar 14 is provided at the front left and right sides of the host vehicle 102. The front millimeter wave radar 12 and the front side millimeter wave radar 14 function as detection means for detecting the target points P1 to P5 at a plurality of positions in the other vehicle 104, and the positions of the target points P1 to P5. Information (information on direction and distance) and relative speed are output as information on other vehicle 104. The steering angle sensor 16 detects the steering angle of the host vehicle 102, the yaw rate sensor 18 detects the yaw rate of the host vehicle 102, and the vehicle speed sensor 20 detects the vehicle speed of the host vehicle 102.

  The collision determination ECU 10 is electrically connected to the alarm ECU 22, the seat belt ECU 24, and the brake ECU 26. The alarm ECU 22 is electrically connected to an alarm device 32 that outputs an alarm sound and a vibrator 34 for transmitting vibration to the driver, and notifies the driver of the alarm. The seat belt ECU 24 operates the seat belt actuator 36 to wind up and tension the seat belt. The brake ECU 26 performs a deceleration control of the host vehicle 102 by sending a target hydraulic pressure signal to a brake actuator 38 that adjusts the hydraulic pressure of the wheel cylinder and controlling the brake actuator 38 to adjust the hydraulic pressure of the wheel cylinder.

  The collision determination ECU 10, the alarm ECU 22, the seat belt ECU 24, and the brake ECU 26 are configured to be able to exchange data with each other by being connected by a communication circuit such as a CAN (Control Area Network). The collision determination ECU 10, the alarm ECU 22, the seat belt ECU 24, and the brake ECU 26 include a CPU that performs arithmetic processing, a ROM and a RAM that serve as a storage unit, an input communication circuit, an output communication circuit, a power supply circuit, and the like.

  Here, in the CPU of the collision determination ECU 10, a time series change calculation unit, a comparison unit, a target point collision site prediction unit, and a collision mode prediction unit are configured by executing a program stored in the storage unit.

  The time-series change calculating unit calculates the time information of the position information of the target points P1 to P5 at a plurality of positions in the other vehicle 104 based on the data detected by the front millimeter wave radar 12 and the front side millimeter wave radar 14. The series change is calculated. The comparison unit compares the time series change of the position information of each of the target points P1 to P5 calculated by the time series change calculation unit with the position information of the host vehicle 102.

  The target point collision site prediction unit predicts the collision site of each of the target points P1 to P5 with respect to the host vehicle 102 based on the comparison result. The collision mode prediction unit predicts the collision mode between the host vehicle 102 and the other vehicle 104 based on the collision site prediction results at a plurality of positions by the target point collision site prediction unit.

  FIG. 2 is a plan view showing a positional relationship between the host vehicle and another vehicle and a detection range by the millimeter wave radar. In FIG. 2, a first road 108 and a second road 110 that are orthogonal to each other at the intersection 106 are shown. The host vehicle 102 travels straight on the first road 108 toward the intersection 106, and the other vehicle 104 moves to the second road. 110 shows that the vehicle 110 is traveling straight toward the intersection 106. If the host vehicle 102 goes straight ahead, there is a possibility of colliding with another vehicle 104 at the intersection 106. A range surrounded by a virtual line indicates a detectable range S1 by the front millimeter-wave radar 12 and a detectable range S2 by the front-side millimeter wave radar 14. In the state of FIG. The front side millimeter wave radar 14 detects the other vehicle 104.

  FIG. 3 is a plan view showing an example of the positional relationship among grouping ranges, representative points, and members of the millimeter wave radar. The grouping conditions for the front millimeter wave radar 12 and the front side millimeter wave radar 14 are within a predetermined range in the Z direction (travel direction of the host vehicle 102), and in the X direction (perpendicular to the travel direction of the host vehicle 102). In the (direction), those having a predetermined range and those having a relative speed difference within a predetermined range are grouped.

  The forward millimeter wave radar 12 detects the position of the front center of the other vehicle 104 as a representative point (target point) P1, and detects the position of the left side of the front as a member (target point) P2 of the representative point P1. Information about the representative point P1 and the member P2 is output as a target. The front side millimeter wave radar 14 detects the position of the front side of the other vehicle 104 as a representative point (target point) P3, and detects the position of the right mirror as a member (target point) P4 of the representative point P3. The position of the rear side of the other vehicle 104 is detected as a member (target point) P5 of the representative point P3, and the position information and relative speed of each target point are output as targets.

  FIG. 4 is a diagram illustrating an example of an approximate curve for estimating a collision trajectory. In FIG. 4, a coordinate system having the origin at the center of the rear surface of the host vehicle 102 is set, the horizontal distance of the host vehicle 102 is shown on the horizontal axis, and the TTC (until the collision with the other vehicle 4 is shown on the vertical axis. Time). The collision determination ECU 10 functions as a time series change calculation unit, a comparison unit, and a target point collision site prediction unit, and shows time series changes in position information of the target points P1 to P5 as shown in FIG. Approximate curves (straight lines) L1 to L5 are calculated.

  Next, the operation procedure of the collision prediction process by the collision determination ECU 10 will be described. FIG. 5 is a flowchart showing the operation procedure of the collision prediction process by the collision determination ECU. When the other vehicle 104 enters the detectable range, the front millimeter wave radar 12 and the front side millimeter wave radar 14 output the position information and relative speed information of the representative points P1, P3 and members P2, P4, P5 as targets. To do. Then, the collision determination ECU 10 inputs the target outputs from the front millimeter wave radar 12 and the front side millimeter wave radar 14 (S1). At this time, the position information and relative speed information of the members P2, P4, and P5 are newly added with information indicating which representative point is the same target.

  Next, the collision determination ECU 10 approximates the time series change of the position information of each of the target points P1 to P5 based on the information output by the front millimeter wave radar 12 and the front side millimeter wave radar 14. A curve is calculated, and a collision locus between the host vehicle 102 and the other vehicle 104 is estimated (S2).

  Subsequently, the collision determination ECU 10 calculates the own vehicle cabin front collision determination value Fn = α × γ + β × γ (1) (S3) and the own vehicle cabin rear collision determination value Rn = β × γ + α × γ. (2) is calculated (S4). In addition, the own vehicle cabin front collision determination value of the target point P1 is denoted as F1. FIG. 6 is a diagram illustrating a setting example of the coefficient α region and the coefficient β region. In FIG. 6, a coordinate system having the origin at the center of the rear surface of the host vehicle 102 is set, and the horizontal axis is the X axis and the vertical axis is the Z axis.

  The front side of the host vehicle 102 is a coefficient α region, and the coefficient α region is divided into a plurality (six in this embodiment) at regular intervals in the X-axis direction. Specifically, two regions in the center of the host vehicle are set as α3, a region α2 is set outside the region α3, and a region α1 is set outside the region α2. The coefficient α satisfies the relationship of α3> α2> α1, and is set in consideration of the fact that the degree of damage increases when the vehicle collides with the center of the vehicle. For example, when the target point is predicted to collide with the region α3 based on the estimated collision locus, the coefficient α3 (α2> α1) is employed.

  Further, the rear side of the host vehicle 102 is set as a coefficient β region, and the coefficient β region is divided into a plurality (six in this embodiment) at equal intervals in the X-axis direction from the center of the host vehicle, like the coefficient α region. To the outside, α3, α2, and α1 are set. The coefficient β satisfies the relationship β3> β2> β1, and is set in consideration of the fact that the degree of damage increases when the vehicle collides with the center of the vehicle. For example, when the target point is predicted to collide with the region β3 based on the estimated collision locus, the coefficient β3 (β2> β1) is employed.

  The coefficient (weight multiplier based on the number of extrapolations) γ is a coefficient for weighting according to the TTC and the number of extrapolations. FIG. 7 is a graph showing an example of setting the weight multiplier γ based on the number of extrapolations. In FIG. 7, the horizontal axis indicates the number of extrapolations, and the vertical axis indicates the coefficient γ. “Extrapolation” means that, when the target is detected by the millimeter wave radars 12 and 14, the reflection level from the target object that is indispensable falls below a predetermined detection level, and the target cannot be detected so far. Based on the detection result, it is assumed that the place that will be detected this time is estimated. “Extrapolation frequency” refers to the number of times of extrapolation continuously. In FIG. 7, G1 is a setting example when TTC is short, G2 is a setting example when TTC is long, and both G1 and G2 are γ = 1 when the number of extrapolation is 0, and extrapolation As the number of times increases, the coefficient γ decreases. G2 is set lower than G1. In other words, even if the millimeter wave radar loses sight of the target immediately before the collision, the coefficient γ is not lowered below 0.8.

  As described above, in step 3 and step 4, the own vehicle cabin front determination value Fn = α × γ + β × γ and the own vehicle cabin rear determination value Rn = β × γ−α × γ for each of the target points P1 to P5. Is calculated. At this time, it is preferable to weight the coefficient α and the coefficient β depending on whether each of the target points P1 to P5 is a representative point or a member. Weighting is performed so as to increase.

  Further, in step 3, when ΣFn is calculated and ΣFn> own vehicle cabin front collision determination range threshold value THFC is satisfied, it is predicted that another vehicle will collide with the front of the own vehicle, and the own vehicle cabin front collision Determination flag = ON. Note that the own vehicle cabin front collision determination range threshold value THFC is set by experiment or the like.

  Further, in step 4, when ΣRn is calculated and ΣRn> own vehicle cabin rear collision determination range threshold value THRC is established, it is predicted that another vehicle will collide with the rear of the own vehicle, and the own vehicle cabin rear collision determination flag = Set to ON. Note that the own vehicle cabin rear collision determination range threshold value THRC is set by experiment or the like.

  Subsequently, the collision determination ECU 10 determines whether or not only the front collision determination flag is ON (S5). If it is determined that only the front collision determination flag is ON, the process proceeds to step 9, and if it is not determined that only the front collision determination flag is ON, the process proceeds to step 6.

  In step 6, the collision determination ECU 10 determines whether only the rear collision determination flag is ON. If it is determined that only the rear collision determination flag is ON, the process proceeds to step 10, and if it is not determined that only the rear collision determination flag is ON, the process proceeds to step 7.

  In step 7, the collision determination ECU 10 determines whether both the front collision determination flag and the rear collision determination flag are ON. If it is determined that both the front collision determination flag and the rear collision determination flag are ON, the process proceeds to step 8, and it is not determined that both the front collision determination flag and the rear collision determination flag are ON. If so, the process ends.

  In Step 8, the collision determination ECU 10 determines whether or not (ΣFn−THFC)> (ΣRn−THRC). When it is determined that (ΣFn−THFC)> (ΣRn−THRC), the process proceeds to step 9, and when it is not determined that (ΣFn−THFC)> (ΣRn−THRC), the process proceeds to step 10.

  In step 9, the collision determination ECU 10 outputs a signal to the alarm ECU 22 and issues an alarm to the driver using the alarm device 32 and the vibrator 34. The collision determination ECU 10 outputs a signal to the seat belt ECU 24, controls the seat belt, operates the seat belt retractor, winds the seat belt, and tensions it. In addition, the collision determination ECU 10 outputs a signal to the brake ECU 26 and controls the brake actuator 36 to adjust the hydraulic pressure of the wheel cylinder, so that the collision of the host vehicle 102 is reduced in order to avoid a collision to the center of the host vehicle cabin. Take control. That is, when it is predicted that another vehicle will collide with the front portion of the own vehicle cabin, the collision determination ECU 10 performs alarm output, seat belt control, and brake control to reduce damage in the center of the own vehicle cabin.

  In step 10, the collision determination ECU 10 outputs a signal to the alarm ECU 22 and issues an alarm to the driver using the alarm device 32 and the vibrator 34. Further, the collision determination ECU 10 outputs a signal to the seat belt ECU 24, controls the seat belt, operates the seat belt retractor, winds the seat belt, and tensions it. In step 10, brake control is not performed. That is, when it is predicted that another vehicle will collide with the rear part of the own vehicle cabin, the collision determination ECU 10 performs alarm output and seat belt control, does not perform brake control, and reduces damage in the center of the own vehicle cabin.

  Next, the collision determination ECU 10 performs alarm control, seat belt control, and brake control in order to avoid a collision to the center of the own vehicle cabin when it is determined to be within the front collision determination range (S4). When the collision determination ECU 10 is determined to be within the rear collision determination range, the collision determination ECU 10 performs alarm control and seat belt control in order to avoid a collision to the center of the own vehicle cabin (S5).

  As described above, in the collision prediction apparatus 100 according to the present embodiment, when predicting the collision site, the approximate curve determination is performed by effectively utilizing the information on the members P2, P4, and P5 that have not been output as targets. The collision part with respect to the own vehicle cabin position can be made. Thereby, the prediction precision of a collision site | part can be improved. Therefore, in the collision mitigation control (PCS: Pre-Crush Safety), alarm control, seat belt control, and brake control corresponding to the collision site of the host vehicle can be performed with high accuracy. In the conventional grouping process, since the member information has not been output, the target information is small and the collision site prediction is insufficient.

  Further, since the collision prediction apparatus 100 outputs the member information as a target, the amount of information related to other vehicles is increased as compared with the conventional case, and the collision site can be predicted with high accuracy. The side impact PCS for preventing the encounter collision is particularly effective because it takes a very short time to detect another vehicle approaching the host vehicle as compared to the front PCS.

  In addition, the collision prediction device controls the presence or absence of brake control depending on whether the vehicle collides with the front part of the own vehicle cabin or the rear part of the own vehicle cabin, thereby reducing the influence on the center of the own vehicle cabin. be able to.

  Further, in the collision prediction device, the coefficient α and coefficient β are set differently when the collision site is located closer to the center of the own vehicle than when the collision site is located closer to the outer side of the own vehicle. Compared to the case of collision, it is possible to predict that the collision surface is smaller and the damage of the own vehicle is less when the vehicle collides with the outside of the own vehicle, and it is more likely that the collision can be avoided. It can be judged. Therefore, it is possible to perform collision prediction in consideration of the magnitude of damage and ease of avoidance.

  As mentioned above, although this invention was concretely demonstrated based on the embodiment, this invention is not limited to the said embodiment. The collision determination ECU 10 may predict the collision mode based on the position distribution of the collision site using, for example, a histogram.

It is a block diagram which shows the collision prediction apparatus which concerns on embodiment of this invention. It is a top view which shows the positional relationship of the own vehicle and another vehicle, and shows the detection range by a millimeter wave radar. It is a top view which shows an example of the positional relationship of the grouping range of a millimeter wave radar, a representative point, and a member. It is a figure which shows an example of the approximated curve for estimating a collision locus. It is a flowchart which shows the operation | movement procedure of the collision prediction process by collision determination ECU. It is a figure which shows the example of a setting of a coefficient (alpha) area | region and a coefficient (beta) area | region. It is a graph which shows the example of a setting of the weight multiplier (gamma) by the frequency | count of extrapolation. It is a model figure for demonstrating the grouping representative point and grouping member in the grouping process in the case of the collision determination using the conventional radar apparatus.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 10 ... Collision judgment ECU (target point collision site | part prediction means, collision mode prediction means), 12 ... Front millimeter wave radar (detection means), 14 ... Front side millimeter wave radar (detection means), 102 ... Own vehicle, 104 ... Other vehicles, P1 to P5 ... Target points.

Claims (6)

Detecting means for detecting target points at a plurality of positions in other vehicles;
A target point collision site prediction means for predicting a collision site of each target point with respect to the host vehicle based on a comparison between time series changes in position information of the target points of the plurality of positions and the position information of the host vehicle; ,
A collision prediction apparatus comprising: a collision mode prediction unit that predicts a collision mode between the own vehicle and the other vehicle based on the collision site prediction result at the plurality of positions.   The collision prediction apparatus according to claim 1, wherein the collision mode prediction unit predicts a collision site of the host vehicle with the other vehicle as the collision mode.   The collision prediction device according to claim 1, wherein the collision mode prediction unit predicts a size of a collision surface of the host vehicle with the other vehicle as the collision mode.   The collision mode prediction unit predicts the collision mode based on a position distribution of the collision site predicted by the target point collision site prediction unit. The collision prediction apparatus described in 1.   The collision according to any one of claims 1 to 4, wherein the collision mode prediction unit predicts the collision mode based on a coefficient corresponding to a prediction result by the target point collision site prediction unit. Prediction device.   6. The collision prediction apparatus according to claim 5, wherein the coefficient is different between the case where the collision site is located closer to the center of the own vehicle and the case where the collision site is located closer to the outer side of the own vehicle.

Images