JP5083062B2 - Collision prediction device - Google Patents

Description

  The present invention relates to a collision prediction apparatus that predicts a collision site of another vehicle with a host vehicle.

As a conventional collision prediction apparatus, for example, as described in Patent Document 1, a future course of a mobile body existing around the host vehicle and a future course of the host vehicle are predicted. From these routes, it is determined whether there is a possibility of collision between the moving body and the host vehicle, and if it is determined that there is a possibility of collision, an alarm is generated for the inside of the host vehicle. It has been.
JP 2003-81037 A

  However, in the above prior art, it cannot be determined to which part of the host vehicle the moving body collides. As a method for predicting the collision site, for example, it is conceivable to integrate the collision probability for each site of the host vehicle and determine that the vehicle collides with the site if the integrated value is larger than a preset threshold value. However, with this method, for example, when the host vehicle is approaching a curved road from a straight road, the collision probability related to a collision with an object such as a guardrail that is accumulated during the straight road running despite starting the steering operation. May cause collision impact mitigation (PCS) devices to operate.

  The objective of this invention is providing the collision prediction apparatus which can improve the prediction precision of the collision site | part with an object.

  The present invention relates to a collision prediction device for predicting a collision portion with an object in the own vehicle, and for each portion of the own vehicle, a collision probability integrating means for adding a collision probability when it is predicted that there is a possibility of collision with the object. A collision determination unit that determines a collision site with an object based on an integrated value of the collision probability for each site of the host vehicle, a steering determination unit that determines a steering state of the host vehicle, and a predetermined amount of steering by the steering determination unit. And an integrated value correcting means for correcting so as to reduce the integrated value of the collision probability when it is determined that the above has been performed.

  In the collision prediction apparatus of the present invention as described above, the collision probabilities when it is predicted that there is a possibility of collision with an object for each part of the host vehicle are sequentially integrated, and based on the integrated value of the collision probability. To determine where the object collides. At this time, the steering state of the host vehicle is determined, and when it is determined that a predetermined amount of steering has been performed, correction is performed so as to decrease the integrated value of the collision probability, and the corrected integrated value of the collision probability is based on the corrected integrated value of the collision probability. To identify the location of collision with the object. Therefore, for example, when the steering vehicle is increased when the host vehicle is approaching a curved road from a straight road, the integrated value of the collision probability obtained when traveling on the straight road is reduced. Thereby, since the contribution degree with respect to the collision site | part prediction of the result obtained at the time of straight road driving | running | working becomes low, the prediction precision of the collision site | part with an object improves.

  Further, the present invention provides a collision probability that integrates the collision probability when it is predicted that there is a possibility of colliding with an object for each part of the own vehicle in a collision prediction device that predicts a collision part with the object in the own vehicle. An integration unit, a collision determination unit that determines a collision part with an object based on an integrated value of a collision probability for each part of the host vehicle, a steering determination unit that determines a steering state of the host vehicle, and a predetermined amount by the steering determination unit When it is determined that a predetermined amount of steering has been performed, there is provided an integrated increase width setting means for setting the integrated increase width of the collision probability smaller than when it is determined that a predetermined amount of steering is not performed. It is.

  In the collision prediction apparatus of the present invention as described above, the collision probabilities when it is predicted that there is a possibility of collision with an object for each part of the host vehicle are sequentially integrated, and based on the integrated value of the collision probability. To determine where the object collides. At this time, the steering state of the host vehicle is determined, and when it is determined that the predetermined amount of steering is performed, the cumulative increase amount of the collision probability is set smaller than when it is determined that the predetermined amount of steering is not performed. To do. Therefore, for example, when the host vehicle travels on a curved road, the collision probability is difficult to be integrated, and thus, for example, a malfunction of the collision damage reduction (PCS) device can be prevented. Further, when the predetermined amount of steering is correction steering, it is possible to predict the collision site by utilizing the past collision probability accumulation. Thereby, the prediction precision of the collision site | part with an object improves.

  Preferably, the steering determination means includes means for estimating a curve radius at which the host vehicle travels, and determines a steering state of the host vehicle based on a differential amount of the curve radius.

  In this case, it is possible to easily and reliably determine the steering state (addition / returning, etc.) of the host vehicle based on the differential amount of the curve radius.

  At this time, it is preferable to further include a differential amount correction means for correcting the curve radius to be smaller when the curve radius is larger than when the curve radius is small.

  A road with a large curve radius and a road with a small curve radius differ greatly in the differential amount of the curve radius even if the steering amount is the same. Specifically, the differential amount of the curve radius is larger on a road having a large curve radius than on a road having a small curve radius. Therefore, the steering state of the host vehicle can be determined more accurately by correcting the differential amount of the curve radius.

  ADVANTAGE OF THE INVENTION According to this invention, the prediction precision of the collision site | part of the own vehicle and an object can be improved. This makes it possible to reduce the annoyance felt by the driver, for example, by operating the collision damage reduction (PCS) device at an appropriate timing.

  DESCRIPTION OF EMBODIMENTS Hereinafter, a preferred embodiment of a collision prediction apparatus according to the present invention will be described in detail with reference to the drawings.

  FIG. 1 is a schematic configuration diagram showing an embodiment of a collision prediction apparatus according to the present invention. In the figure, a collision prediction device 1 according to the present embodiment is a device that predicts a collision portion of an own vehicle with an object such as another vehicle, a bicycle, or a guardrail. The collision prediction apparatus 1 includes an obstacle capturing sensor 2, a steering angle sensor 3, a yaw rate sensor 4, a wheel pulse sensor 5, and a system ECU (Electronic Control Unit) 6.

  The obstacle capturing sensor 2 is a sensor that detects an object (obstacle) that exists in the vicinity of the host vehicle. For example, a millimeter wave radar sensor or an image sensor is used. The steering angle sensor 3 is a sensor that detects the steering angle of the steering. The yaw rate sensor 4 is a sensor that detects the yaw rate of the host vehicle. The wheel pulse sensor 5 is a sensor that detects the speed of the wheel, and outputs a pulse every rotation of the wheel.

  The system ECU 6 includes a microcomputer including a CPU, a ROM, a RAM, and the like as main components. An alarm device 7, a collision avoidance support device 8, a seat belt control unit 9, a seat control unit 10, a brake control unit 11, and an airbag control unit 12 are connected to the system ECU 6.

  The warning device 7 gives a warning to the driver of the host vehicle by a buzzer (warning sound), sound, display, and the like according to a control signal from the system ECU 6. The collision avoidance support device 8 performs steering control for avoiding a collision or alleviating an impact caused by the collision in accordance with a control signal from the system ECU 6, and uses, for example, a gear ratio variable steering system (VGRS). Thus, the cutting angle of the front wheel is optimally controlled.

  The seat belt control unit 9 controls the tightening force of the seat belt according to a control signal from the system ECU 6. The seat control unit 10 controls the position / posture of the seat so as to reduce the impact on the occupant due to the collision in response to a control signal from the system ECU 6. The brake control unit 11 automatically controls the brake according to a control signal from the system ECU 6 to apply a braking force to the host vehicle. The airbag control unit 12 controls the airbag deployment operation in accordance with a control signal from the system ECU 6.

  The system ECU 6 receives output signals from the obstacle capturing sensor 2, the steering angle sensor 3, the yaw rate sensor 4, and the wheel pulse sensor 5, performs predetermined processing, estimates a collision portion with an object in the own vehicle, and performs the collision. When there is a high possibility that an object will collide with a part, an alarm device 7, a collision avoidance support device 8, a seat belt control unit 9, a seat control unit 10, and a brake control unit according to a time margin (Time To Collision: TTC) 11 and the airbag control unit 12 are operated.

  Here, FIG. 2 shows a basic concept of estimating a collision site with an object in the own vehicle. In the figure, first, the own vehicle A is divided into a plurality of sections. Although FIG. 2 shows the front and left sides of the host vehicle, the same applies to the right side of the host vehicle.

  Then, based on the output signal from the obstacle capturing sensor 2 or the like, it is estimated to which part section of the host vehicle the object may collide. Specifically, the position B of the object detected by the obstacle capturing sensor 2 is sequentially obtained, the object trajectory C is predicted using, for example, the least square method, and the intersection of the object expected trajectory C and the host vehicle. The part section corresponding to is assumed to be a collision predicted part section predicted to have a possibility of a collision. Then, a collision probability set in advance for the collision predicted region is added. Thereafter, the above process is repeated, and the collision probabilities are sequentially accumulated for each part section of the host vehicle. When the integrated value of the collision probability exceeds a preset threshold value, it is determined that there is a high possibility that the object will collide with the part section.

  FIG. 3 is a flowchart showing details of the collision determination processing procedure executed by the system ECU 6.

  In the figure, first, based on the output signal of the obstacle capturing sensor 2, an obstacle existing in front or side of the host vehicle is detected (step S101). Further, based on the output signals of the steering angle sensor 3 and the yaw rate sensor 4, the curve radius of the road on which the host vehicle travels is estimated (step S102). Then, a differential value (rate of change) of the estimated curve radius (estimated curve radius) is calculated (step S103).

  Subsequently, the estimated curve radius differential coefficient is calculated by correcting the differential value of the estimated curve radius with the R correction coefficient (step S104). The estimated curve radius derivative is calculated as follows.

  That is, first, an R correction coefficient corresponding to the estimated curve radius is obtained using a table as shown in FIG. The table in FIG. 7 represents the relationship between the estimated curve radius and the R correction coefficient, and is stored as data in advance. The estimated curve radius is obtained by averaging the estimated curve radii in the past fixed time or by various filter processes.

Here, the absolute value of the estimated curve radius increases as the road becomes straight, and the amount of derivative of the estimated curve radius increases only with the correction steering that corrects the wobbling during straight traveling. For this reason, the R correction coefficient is set to be smaller as the estimated curve radius is larger. Then, an estimated curve radius differential coefficient is calculated by the following equation. At this time, the R correction coefficient is set to a value that increases the estimated curve radius differential coefficient as the steering state increases from the steady state (non-steered state) to the switch back state.
Estimated curve radius derivative = derivative of estimated curve radius x R correction factor

  In this way, taking into account that the derivative amount of the estimated curve radius increases as the road becomes straight, the derivative amount of the estimated curve radius is corrected according to the estimated curve radius. Using the threshold value, it is possible to determine whether the steering is increased or decreased. As a result, the steering state of the host vehicle can be accurately determined.

  Next, the accumulated value (see FIG. 2) of past collision probabilities accumulated for each part section of the host vehicle is corrected using the estimated curve radius differential coefficient (step S105). Specifically, a collision probability correction value corresponding to the estimated curve radius differential coefficient is obtained using a table as shown in FIG. The table in the figure represents the relationship between the estimated curve radius differential coefficient and the collision probability correction value, and is stored in advance as data.

  When the estimated curve radius differential coefficient is large, the steering state is a state in which increase or reduction is being performed, and the integrated value of the collision probability accumulated for each part section of the own vehicle is canceled ( to correct. At this time, the integrated value of the collision probability may be zero, or may be decreased by a predetermined amount or a predetermined ratio. When the estimated curve radius differential coefficient is small, the steering state is a steady state or a state close thereto, and the integrated value of the collision probability accumulated for each part section of the own vehicle is maintained as it is. In the intermediate region of the estimated curve radius differential coefficient, the integrated value of the collision probability is set (corrected) so as to decrease as the steering amount increases.

  Next, the vehicle speed of the host vehicle is calculated based on the output signal of the wheel pulse sensor 5 (step S106). Moreover, as shown in FIG. 2, the locus | trajectory of an obstruction is estimated based on the output signal of the obstruction capture sensor 2 (procedure S107). Then, based on the estimated curve radius of the road on which the host vehicle is traveling, the vehicle speed of the host vehicle, and the path of the obstacle, a collision site with the obstacle in the host vehicle is predicted as shown in FIG. 2 (step S108).

  Next, using the estimated curve radius differential coefficient obtained in step S104, the cumulative increase width of the collision probability is set (step S109). Specifically, using the table as shown in FIG. 6, the range of increase in the collision probability corresponding to the estimated curve radius differential coefficient is obtained. The table shown in the figure represents the relationship between the estimated curve radius differential coefficient and the cumulative increase in the collision probability, and is stored as data in advance.

  When the estimated curve radius differential coefficient is large, the steering state is a state in which increase or reduction is being performed, and the cumulative value of the collision probability accumulated for each part section of the host vehicle is canceled as described above. Since this is necessary, the cumulative increase in the collision probability is set to zero. That is, even if it is determined that there is a possibility of collision with an obstacle, the collision probability is not added. When the estimated curve radius differential coefficient is small, the steering state is a steady state or a state close thereto, and the accumulated increase width of the collision probability accumulated for each part section of the host vehicle is set to the maximum value. Further, in the intermediate region of the estimated curve radius differential coefficient, the cumulative increase range of the collision probability is set to decrease relatively rapidly from the maximum value to zero as the steering amount increases.

  Next, the current collision probability is added by an amount corresponding to the accumulated increase width to the integrated value of the collision probability at the collision site predicted in step S108 (step S110).

  Next, based on the integrated value of the collision probability for each collision site, it is determined whether or not an obstacle collides with a certain section of the host vehicle (step S111). At this time, as described above, when the integrated value of the collision probability of a certain part section exceeds a threshold value, it is determined that there is a high possibility that an obstacle will collide with the part section. When it is determined that the obstacle does not collide with the host vehicle, the process returns to step S101.

  On the other hand, when it is determined that an obstacle collides with a certain section of the host vehicle, the alarm device 7, the collision avoidance support device 8, the seat based on the collision margin time (TTC), the collision portion and the integrated value of the collision probability Control is performed to activate any one of the belt controller 9, the seat controller 10, the brake controller 11, and the airbag controller 12 (step S112).

  In the above, the above steps S108 and S110 of the system ECU 6 constitute a collision probability integrating unit that integrates the collision probability when it is predicted that there is a possibility of collision with an object for each part of the host vehicle. The procedure S111 constitutes a collision determination unit that determines a collision part with an object based on an integrated value of the collision probability for each part of the host vehicle. The above steps S102 and S103 of the steering angle sensor 3, the yaw rate sensor 4 and the system ECU 6 constitute a steering determination means for determining the steering state of the host vehicle. The procedure S105 constitutes an integrated value correcting means for correcting so that the integrated value of the collision probability is decreased when it is determined that a predetermined amount of steering is performed by the steering determining means. In step S109, when the steering determination unit determines that the predetermined amount of steering has been performed, the integration for setting the cumulative increase amount of the collision probability smaller than that when it is determined that the predetermined amount of steering has not been performed. The raising width setting means is configured. The procedure S104 constitutes differential amount correction means for correcting the curve radius to be smaller when the curve radius is larger than when the curve radius is small.

  By the way, simply judging whether or not the obstacle collides with the own vehicle based on the integrated value of the collision probability at the collision site predicted from the past movement of the own vehicle and the obstacle, as shown in FIG. The following problems occur when traveling along a curve entrance that runs from a road to a curved road.

  That is, when traveling on a straight road (see P in the figure), an object (a guard rail or the like) B at the entrance of the curve can be seen in front, so that the collision probability is integrated with respect to the collision site corresponding to the front center of the host vehicle. It becomes. After that, when entering a curved road (see Q in the figure), the TTC becomes small, but since the rudder is turned off, it does not collide with the object B.

  However, although the steering is started, it is determined that the object B collides with the host vehicle due to the influence of the cumulative value of the collision probability accumulated during traveling on the straight road. For this reason, the collision damage reduction (PCS) devices such as the alarm device 7, the collision avoidance support device 8, the seat belt control unit 9, the seat control unit 10, the brake control unit 11, and the airbag control unit 12 are reduced by reducing the TTC. Will cause trouble to the driver.

  In addition, the same problem as described above also occurs at the time of traveling at a curve exit that runs from a curved road to a straight road.

  On the other hand, in the present embodiment, it is determined whether or not there has been a steering addition / return based on the differential amount of the estimated curve radius of the traveling road. The integrated value of the collision probability for each collision site is lowered. For this reason, when the host vehicle approaches from a straight road to a curved road, the degree of influence of the integrated value of the collision probability accumulated during traveling on the straight road on the collision site prediction is reduced. In addition, when there is an increase or decrease in steering, in order to sufficiently reduce the cumulative increase in the collision probability at that moment, the degree of influence of the past collision probability integrated value on the collision site prediction gradually decreases. . As described above, it is prevented that it is determined that the object collides with the host vehicle when the integrated value of the collision probability exceeds the threshold value.

  Therefore, since the prediction accuracy of the collision part at the time of traveling at the curve entrance and the curve exit is improved, the malfunction of the PCS device is prevented, and as a result, the troublesomeness given to the driver can be reduced.

  Also, in areas where it is difficult to distinguish between simple corrective steering and steering increment / return, the amount of decrease in the cumulative value of collision probability is kept small, so the cumulative value of collision probability accumulated by corrective steering is reset. Is prevented. Therefore, when corrective steering is performed, the collision portion is predicted by using the integrated value of the collision probability so far as it is. Thereby, the prediction precision of the collision site | part at the time of driving | running | working of a curve entrance and a curve exit can be improved further.

  Furthermore, in areas where it is difficult to distinguish between corrective steering and additional turning / returning of steering, it not only suppresses a decrease in the cumulative value of the collision probability, but also suppresses a decrease in the cumulative increase in the collision probability at that moment. It is possible to make a collision determination at an appropriate timing depending on whether or not there is subsequent increase / return.

  For example, when the steering is further increased after the correction steering is performed, the integrated value of the collision probability is further decreased, so that unnecessary collision damage reduction operation is suppressed. In addition, when the steering is not increased after the correction steering is performed, the collision probability is added to the integrated value of the collision probability so far, so that stable collision determination can be performed. .

It is a schematic block diagram which shows one Embodiment of the collision prediction apparatus concerning this invention. It is a conceptual diagram which shows the basic view which estimates the collision site | part with the object in the own vehicle by the collision prediction apparatus shown in FIG. It is a flowchart which shows the detail of the collision determination processing procedure performed by system ECU shown in FIG. It is a graph which shows an example of the table showing the relationship between an estimated curve radius and R correction coefficient. It is a graph which shows an example of the table showing the relationship between an estimated curve radius differential coefficient and a collision probability correction value. It is a graph which shows an example of the table showing the relationship between a presumed curve radius differential coefficient and the accumulation width | variety of the collision probability. It is a figure which shows the state which the own vehicle drive | works the curve entrance approaching from a straight road to a curve road.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 ... Collision prediction apparatus, 3 ... Steering angle sensor (steering judgment means), 4 ... Yaw rate sensor (steering judgment means), 6 ... System ECU (collision probability integration means, collision judgment means, steering judgment means, integrated value correction means, (Accumulated increase width setting means, differential amount correction means).

Claims (4)

In a collision prediction device that predicts a collision site with an object in the own vehicle,
Collision probability accumulation means for accumulating the collision probability when predicted to possibly collide with the object for each part of the host vehicle;
A collision determination means for determining a collision portion with the object based on an integrated value of the collision probability for each portion of the host vehicle;
Steering determination means for determining the steering state of the host vehicle;
A collision prediction apparatus comprising: an integrated value correcting unit that corrects so that the integrated value of the collision probability is decreased when it is determined that a predetermined amount of steering has been performed by the steering determining unit. In a collision prediction device that predicts a collision site with an object in the own vehicle,
Collision probability accumulation means for accumulating the collision probability when predicted to possibly collide with the object for each part of the host vehicle;
A collision determination means for determining a collision portion with the object based on an integrated value of the collision probability for each portion of the host vehicle;
Steering determination means for determining the steering state of the host vehicle;
When the steering determination means determines that a predetermined amount of steering has been performed, the cumulative increase width setting means for setting the cumulative increase width of the collision probability smaller than when it is determined that the predetermined amount of steering has not been performed. A collision prediction apparatus comprising:   The steering determination means includes means for estimating a curve radius on which the host vehicle travels, and determines a steering state of the host vehicle based on a differential amount of the curve radius. The collision prediction apparatus described. 4. The collision prediction apparatus according to claim 3, further comprising differential amount correction means for correcting the curve radius to be smaller when the curve radius is larger than when the curve radius is small. .

Images