WO2016204213A1 - Vehicle control device and vehicle control method - Google Patents

Abstract

A vehicle control device (10, 10a), which creates a fusion target by fusing a radar target of an object that is in front of a host vehicle and acquired as a reflected wave of a carrier wave, and an image target of an object acquired by image processing of an image obtained by imaging the area in front of the host vehicle, and which controls the host vehicle in regard to the object detected as the fusion target, assesses whether or not the object has transitioned from being detected using the fusion target to being detected using only the radar target, and, when the object is determined to have transitioned to being detected using only the radar target, assesses whether or not the distance to the object is a prescribed near distance. When the distance to the object is determined to be a prescribed near distance, vehicle control in relation to the object is carried out.

Description

Vehicle control apparatus and vehicle control method Cross-reference of related applications

This application is based on Japanese Application No. 2015-121397 filed on June 16, 2015 and Japanese Application No. 2016-1112096 filed on June 3, 2016. Incorporate.

The present invention relates to a vehicle control device and a vehicle control method for performing vehicle control on an object ahead of the host vehicle.

When the radar target acquired by the radar sensor and the image target acquired by the image sensor are collated and it is determined that the radar target and the image target are from the same object, the radar target A technique for generating a new target (fusion target) by fusing the image target with the image target is known. By generating this fusion target, the recognition accuracy of an object such as a preceding vehicle ahead of the host vehicle can be improved. And the vehicle control of the own vehicle with respect to an object can be appropriately performed by using the positional information on the object specified using a fusion target (refer patent document 1).

JP 2005-145396 A

However, when the host vehicle approaches the object, a part (lower end) of the object is removed from the shooting angle of view of the image sensor, so that a fusion target may not be generated. In this case, there arises a problem that vehicle control cannot be performed on the object specified by the fusion target.

The present invention has been made in view of the above, and has as its main object to provide a vehicle control device and a vehicle control method capable of appropriately performing vehicle control on an object ahead of the host vehicle.

The vehicle control apparatus according to an aspect of the present invention provides a first target information of an object in front of the host vehicle acquired as a reflected wave of a carrier wave, and a second of the object acquired by image processing of a captured image in front of the host vehicle. A vehicle control device that fuses target information to generate a fusion target and performs vehicle control of the host vehicle with respect to the object detected as the fusion target, wherein the object is detected by the fusion target. A state determination unit that determines whether or not the object has transitioned to a state in which the object is detected only by the first target information, and the state detection unit detects the object by only the first target information. A distance determining unit that determines whether or not the distance to the object when it is determined that the state has transitioned to a predetermined state is a predetermined short distance, and the distance determining unit determines that the distance to the object is a predetermined short distance The distance When it is, and a vehicle control unit for implementing a vehicle control for the object.

According to the present invention, when the second target information cannot be acquired from the state in which the object is detected as a fusion target, the object is transitioned to a state in which only the first target information is detected. The vehicle control for the object is performed on condition that the distance to the object is a predetermined short distance. Therefore, even after the object is no longer detected as a fusion target, it is possible to perform vehicle control on the object with high reliability.

The schematic block diagram of the vehicle control apparatus in 1st and 2nd embodiment. The functional block diagram of ECU in 1st Embodiment. Explanatory drawing of the relationship between a vehicle distance and an image lost. The figure which shows the relationship between relative speed and collision margin time. The flowchart of vehicle control. The block diagram of the logic circuit for determining permission and prohibition of PB. The functional block diagram of ECU in 2nd Embodiment. The top view at the time of the own vehicle approaching an object. The timing chart which shows the aspect which permits the action | operation of PB. The top view at the time of the own vehicle approaching an object. The timing chart which shows the aspect which permits the action | operation of PB. The top view at the time of the own vehicle approaching an object. The timing chart which shows the aspect which prohibits the action | operation of PB. The block diagram of the logic circuit for determining permission and prohibition of PB in another example. The figure which shows the relationship between the operating width W and the lateral speed of an object.

Hereinafter, embodiments of the present invention will be described in more detail with reference to the accompanying drawings. However, the present invention may be implemented in many different forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the invention to those skilled in the art. Note that similar reference numbers indicate similar components throughout the drawings.

(First embodiment)
A vehicle control system 100 according to the present embodiment is mounted on a vehicle, detects an object existing in front of the vehicle, and performs various controls to avoid or reduce a collision with the object. system). In the following description, a vehicle equipped with the vehicle control system 100 is referred to as a host vehicle.

1A, the vehicle control system 100 includes an ECU 10, various sensors 20, and a controlled object 30.

As the various sensors 20, for example, an image sensor 21, a radar sensor 22, a vehicle speed sensor 23, and the like are provided.

The image sensor 21 is a CCD camera, a monocular camera, a stereo camera or the like, and is installed near the upper end of the windshield of the host vehicle. The image sensor 21 captures a captured image by capturing an area that extends in a predetermined range toward the front of the host vehicle at predetermined time intervals. Then, by subjecting the captured image to image processing, an object ahead of the host vehicle is acquired as target information (image target GT) and output to the ECU 10.

The image target GT includes information such as the lateral width of the object in addition to the distance and relative speed with the object in the traveling direction of the host vehicle, the lateral position indicating the position of the host vehicle in the vehicle width direction. Therefore, the ECU 10 recognizes the image target GT as information having a predetermined width.

The radar sensor 22 detects an object ahead of the host vehicle as target information (radar target LT) using a directional electromagnetic wave such as a millimeter wave or a laser, and the light is detected at the front of the host vehicle. The shaft is attached so that it faces the front of the vehicle. The radar sensor 22 scans a region extending in a predetermined range toward the front of the vehicle every predetermined time with a radar signal, and receives an electromagnetic wave reflected from the surface of the object outside the vehicle, thereby detecting the distance from the object and the object. Relative speed and the like are acquired as target information and output to the ECU 10.

The radar target LT includes information such as the distance to the object in the traveling direction of the host vehicle, the relative speed, and the lateral position indicating the position of the host vehicle in the vehicle width direction. The radar target LT corresponds to the first target information, and the image target GT corresponds to the second target information.

The vehicle speed sensor 23 is provided on a rotating shaft that transmits power to the wheels of the host vehicle, and obtains the host vehicle speed that is the speed of the host vehicle based on the rotational speed of the rotating shaft.

The ECU 10 is an electronic control unit that controls the entire vehicle control system 100. The ECU 10 is mainly composed of a CPU and includes a ROM, a RAM, and the like. The ECU 10 fuses the image target GT and the radar target LT to detect an object (vehicle, road obstacle or other vehicle) ahead of the host vehicle.

Specifically, the position of the fusion target in the traveling direction of the host vehicle is specified based on the distance and relative speed of the radar target LT, and the width of the fusion target in the vehicle width direction of the host vehicle is determined based on the lateral width and lateral position of the image target GT. Identify the location. As described above, when the fusion target is generated using the radar target LT and the image target GT and the position of the object is specified by the fusion target, among the information acquired by the radar sensor 22 and the image sensor 21, The position of the object is specified using information with higher accuracy, and the recognition accuracy of the position of the object can be improved.

The ECU 10 performs well-known image processing such as template matching on the captured image acquired from the image sensor 21, thereby detecting the type of object detected as the image target GT (another vehicle, a pedestrian, a road obstacle). Etc.). In the present embodiment, a plurality of dictionaries, which are image patterns indicating features for each object, are stored in the ROM as templates for specifying the type of each object. As the dictionary, both a whole body dictionary in which features of the entire object are patterned and a half-body dictionary in which partial features of the object are patterned are stored. Information on the type of object recognized by the image sensor 21 is also input to the ECU 10.

In this embodiment, the object recognition accuracy is improved by generating the fusion target on the condition that the type of the object detected as the image target GT is specified in the whole body dictionary. In other words, when the reliability of an object is low, such as when the type of the object detected as the image target GT is specified only by the half-body dictionary, the image target GT is not used for the generation of the fusion target. ing.

Then, the ECU 10 determines whether or not there is a possibility of collision between the object recognized as the fusion target and the host vehicle. Specifically, the lateral position closest to the host vehicle is selected as the lateral position to be controlled among the lateral position of the fusion target and the lateral position of the image target GT. Then, based on the approaching state between the lateral position of the selected object and the host vehicle, it is determined whether or not there is a possibility of collision between the host vehicle and the object.

If it is determined that there is a possibility of a collision, a collision margin time TTC (Time to the object) is calculated by a method such as dividing the distance in the traveling direction between the object and the host vehicle by the relative speed with the object. to Collision). The relative speed is obtained by subtracting the vehicle speed of the host vehicle from the vehicle speed of the preceding vehicle. The TTC is an evaluation value indicating how many seconds later the vehicle collides with an object when traveling at the vehicle speed as it is. The smaller the TTC, the higher the risk of collision, and the larger the TTC, the higher the risk of collision. The nature becomes low. Note that TTC may be calculated in consideration of relative acceleration.

Then, the ECU 10 compares the TTC and the operation timing of each controlled object 30, and if the TTC is equal to or less than the operation timing, the ECU 10 operates the corresponding controlled object 30.

As the controlled object 30, a speaker, a seat belt, a brake, and the like are provided, and a predetermined operation timing is set for each controlled object 30. Therefore, the ECU compares the TTC and the operation timing of each controlled object 30, and operates the corresponding controlled object 30 when the TTC is equal to or lower than the operation timing.

Specifically, if the TTC falls below the speaker operation timing, an alarm is sent to the driver by the operation of the speaker. If the TTC is equal to or lower than the seat belt operation timing, the seat belt is wound up. If the TTC is equal to or less than the brake operation timing, the automatic brake is operated to reduce the collision speed. As described above, the collision between the host vehicle and the object is avoided or alleviated.

By the way, when the host vehicle approaches the object from a state in which the object is recognized as a fusion target, the object may not be recognized as a fusion target because the lower end of the object moves out of the shooting range of the image sensor 21. Yes (the image target GT is lost).

FIG. 2 shows an explanatory diagram of the relationship between the distance between the host vehicle and the object and the lost image target GT. In the drawing, it is shown as a shooting angle of view θ1 of the image sensor 21. First, when the distance between the host vehicle M1 and the object (preceding vehicle M2) is d1, the entire rear end portion of the preceding vehicle M2 is included in the shooting field angle θ1 of the image sensor 21, so that the image object is displayed in the whole body dictionary. The type of the target GT is specified, and a fusion target can be generated. However, when the distance between the host vehicle M2 and the preceding vehicle M2 approaches d2 (<d1), the lower end side of the rear end of the preceding vehicle M2 deviates from the shooting angle of view θ1 of the image sensor 21, and thus the image target GT in the whole body dictionary. It becomes impossible to specify the type of the target, and the fusion target cannot be generated.

However, since an object once recognized as a fusion target has a high degree of reliability that the object exists, it is preferable that vehicle control can be performed on the object even after the object is no longer recognized as a fusion target.

Therefore, when the image target GT is lost from the state where the object is recognized by the fusion target, that is, when the object is recognized only by the radar target LT, the vehicle and the object On the condition that the distance is a predetermined short distance, vehicle control is performed on an object recognized only by the radar target LT.

That is, when the image target GT is lost, if the distance between the host vehicle and the object is a predetermined short distance, the TTC for the object recognized only by the radar target LT is the operation timing of the controlled object 30. The controlled object 30 is actuated. On the other hand, if the distance between the host vehicle and the object is not a predetermined short distance, the controlled object 30 is operated even if the TTC for the object recognized only by the radar target LT is the operation timing of the controlled object 30. I won't let you.

The predetermined short distance is a distance at which the lower end of the object cannot be seen, and may be set for each vehicle type in consideration of the mounting height, the mounting angle, and the like of the image sensor 21. As described above, even when the image target GT is lost, vehicle control can be performed on an object with high reliability that has been detected as a fusion target.

Note that if the relative speed between the object and the host vehicle is large, the TTC becomes a small value. Therefore, there is a possibility that the operation of the controlled object 30 by the vehicle control has already been started before the image target GT is lost. high. In other words, when the image target GT is lost, the situation where the operation of the controlled object 30 by the vehicle control is not started is limited to the situation where the relative speed between the object and the host vehicle is small and the TTC is a large value. It is done.

Here, the relationship between relative speed and TTC will be described in detail with reference to FIG. In the figure, the vertical axis represents relative speed and the horizontal axis represents TTC. As shown in the figure, since the object can be detected as a fusion target up to a smaller value of TTC as the relative speed increases, the vehicle is detected with respect to the object detected as the fusion target before the image target GT is lost. The possibility that the controlled object 30 is activated by the control is increased. On the other hand, when the relative speed decreases, the object can be detected as a fusion target only up to a larger value of TTC. Therefore, there is a high possibility that the image target GT is lost before the controlled object 30 is activated by vehicle control. Become.

Therefore, in the present embodiment, the radar target LT is changed from the fusion target on the condition that the relative speed between the object and the host vehicle is smaller than a predetermined distance in addition to the predetermined short distance between the object and the host vehicle. The vehicle control is performed on the object that has been switched to the recognition only by the user.

Also, after a predetermined time has elapsed since the object could not be specified with the fusion target (after the image target GT was lost), the reliability of the object specified with the radar target only gradually decreases. As described above, when the image target GT is lost, the vehicle control is performed on the condition that the host vehicle and the object are close to each other and the relative speed is small. There is a high possibility that the vehicle has already stopped before the predetermined time elapses after the loss. Therefore, if the elapsed time after the object is no longer specified by the fusion target is within a predetermined time, the vehicle control is performed, and if the elapsed time exceeds the predetermined time, the vehicle control is not performed.

Furthermore, in this embodiment, when the image target GT is lost and the state is changed to a state where the object is recognized only by the radar target LT, the radar is used instead of the lateral position of the object obtained by the fusion target. The collision determination is performed using the lateral position of the object obtained by the target LT.

At this time, immediately after the image target GT is lost, the object is identified by the radar target LT immediately after the transition after switching to the detection of only the lateral position of the object identified by the fusion target and the radar target LT. It is determined whether or not the difference from the horizontal position is less than a predetermined value. If the difference between the lateral position of the object specified by the radar target LT and the lateral position of the object specified by the fusion target is large, vehicle control is not performed. That is, if the difference between the horizontal position of the object specified by the radar target LT and the horizontal position of the object specified by the fusion target is large, the reliability of the object specified by the fusion target is low. In this case, vehicle control using a radar target is not performed.

Note that if the lateral position of the object is in a predetermined approaching state with respect to the host vehicle, the possibility of collision between the object and the host vehicle increases. On the other hand, if the lateral position of the object is not in a predetermined approaching state with respect to the own vehicle, the possibility of a collision between the object and the own vehicle is reduced. Therefore, in the present embodiment, vehicle control is performed on the condition that the lateral position of the object is in a predetermined approaching state with respect to the host vehicle.

As described above, even if the image target GT is lost, it is possible to appropriately perform vehicle control on a highly reliable object detected as a fusion target.

Next, an execution example of the above process by the ECU 10 will be described with reference to the flowchart of FIG. The following processing is repeatedly performed by the ECU 10 at a predetermined cycle for each radar target LT when the radar target LT is acquired.

First, the ECU 10 determines whether or not the radar target LT is in a fusion state in step S11. For example, the ECU 10 affirms when the radar target LT and the image target GT are in a fusion state because the image target GT is included in a predetermined range in the coordinate system of the radar target LT.

In step S12, in the fusion state, the ECU 10 determines whether or not the lateral position of the object specified by the fusion target is equal to or less than a predetermined first threshold value Th1. Specifically, the ECU 10 determines whether or not the distance between the host lane O and the lateral position is equal to or less than a predetermined first threshold Th1 when the center position in the vehicle width direction of the host vehicle M1 is the axis (own lane O). Determine.

ECU10 determines whether TTC is below the operation timing Th2 of the controlled object 30 in step S13, when step S12 is affirmed. When step S13 is affirmed, the ECU 10 operates the controlled object 30 in step S14. When step S12 or 13 is denied, ECU10 progresses to step S22 and does not operate the controlled object 30.

On the other hand, if step S11 is negative, the ECU 10 determines whether or not the image target GT is lost (image lost FSN) from the fusion state in step S15. If the determination in step S15 is affirmative, the ECU 10 determines whether or not the distance of the object specified by the radar target in step S16 is equal to or less than a third threshold Th3. If the determination in step S16 is affirmative, the ECU 10 determines in step S17 whether or not the relative speed between the host vehicle and the object specified by the radar target is equal to or less than a fourth threshold Th4.

If the determination in step S17 is affirmative, the ECU 10 determines in step S18 whether the state in which the image target GT has been lost continues for a predetermined number of times (or cycles). This process is affirmed when the state in which the object is detected only by the radar target LT is repeated a predetermined number of times or less after switching from the fusion target to the detection of the radar target LT. By using as a determination condition whether or not the state in which the image target GT is lost continues, it can be distinguished from the case in which the image target GT is lost due to the influence of disturbance or the like.

If the ECU 10 affirms Step S18, in Step S19, the ECU 10 determines whether or not the lateral position of the object specified by the radar target LT is equal to or less than the fifth threshold Th5. The fifth threshold Th5 is set as a determination value for the approaching state between the vehicle and the object. If the determination in step S19 is affirmative, the ECU 10 determines in step S20 whether or not the lateral position of the object specified by the fusion target is equal to or less than the sixth threshold Th6 immediately before the image target GT is lost. To do. Note that the difference between the fifth threshold Th5 and the sixth threshold Th6 is set to be less than a predetermined value. Therefore, when both steps S19 and S20 are positive, the difference between the lateral position of the object specified by the radar target LT and the lateral position of the object specified by the fusion target is less than a predetermined value. become. When step S20 is affirmed, the ECU 10 determines whether or not TTC is equal to or lower than the operation timing Th2 of the controlled object 30 in step S21. When the ECU 10 affirms Step S21, the ECU 10 proceeds to Step S14 and operates the controlled object 30. If any of steps S15 to S21 is denied, the ECU 10 proceeds to step S22 and does not actuate the controlled object 30.

FIG. 1B shows functional blocks representing the functions of the ECU 10. The ECU 10 includes a state determination unit 101, a distance determination unit 102, a lateral position acquisition unit 103, a lateral position determination unit 104, a relative speed determination unit 105, and a vehicle control unit 106.

The state determination unit 101 is a functional block that executes Step S15 of the flowchart of FIG. 4. Whether the object has transitioned from a state in which the object is detected by the fusion target to a state in which the object is detected only by the first target information. Determine whether or not.

The distance determination unit 102 is a functional block that executes Step S <b> 16, and the distance to the object when the state determination unit 101 determines that the state has been changed to a state in which the object is detected using only the first target information. Is determined to be a predetermined short distance.

The lateral position acquisition unit 103 is a functional block that acquires the lateral position of the object in the vehicle width direction, and the object is detected only from the first target information from the state in which the object is detected as a fusion target. If it is determined that the state has been changed, the lateral position of the object in the vehicle width direction is acquired using the first target information. In addition, when the horizontal position acquisition unit 103 transitions from a state in which the object is detected as a fusion target to a state in which the object is detected only with the first target information, the lateral position acquisition unit 103 selects the fusion target immediately before the transition. To obtain the first lateral position of the object, and obtain the second lateral position of the object using the first target information immediately after the transition.

The lateral position determination unit 104 is a functional block that executes steps S19 and S20. When the distance determination unit 102 determines that the distance to the object is a predetermined short distance, the lateral position acquisition unit 104 Whether or not the acquired lateral position of the object in the vehicle width direction is in a predetermined approach state with respect to the host vehicle (that is, the lateral position of the object specified by the radar target LT is equal to or less than a fifth threshold Th5). If it is in a predetermined approaching state with respect to the host vehicle, it is determined whether or not the difference between the first lateral position and the second lateral position is greater than or equal to a predetermined value.

The relative speed determination unit 105 is a functional block that executes Step S17, and determines whether the relative speed between the object and the host vehicle is smaller than a predetermined value (Th4).

The vehicle control unit 106 is a functional block that executes steps S14 and S22, and performs vehicle control on the object according to the flowchart of FIG.

According to the above, the following effects can be achieved.

(1) By detecting an object using a fusion target generated by the fusion of the radar target LT and the image target GT, the detection accuracy of the object can be improved. However, when the host vehicle and the object approach, a part of the object (for example, the lower end) may be removed from the captured image, and the image target GT is not acquired. In this case, since the fusion target is not generated, the vehicle control for the object using the fusion target may not be performed.

On the other hand, since the reliability of an object (object) detected as a fusion target before the host vehicle and the object approach each other is high, the fusion target is not detected as the host vehicle approaches the object. However, it is preferable that vehicle control is performed on the object.

Therefore, when the image target GT cannot be acquired from the state in which the object is detected with the fusion target, and the transition is made to the state in which the object is detected only with the radar target LT, the distance to the object is Vehicle control over an object is performed on condition that the distance is a predetermined short distance. Therefore, even after the object is no longer detected as a fusion target, it is possible to perform vehicle control on the object with high reliability.

(2) Transition from a state in which an object is detected as a fusion target to a state in which the object is detected only by the radar target LT, and it is determined that the distance between the host vehicle and the object is a predetermined short distance. If the lateral position of the object at that time is in a predetermined approaching state with respect to the own vehicle, the possibility of collision between the object and the own vehicle is increased, and the lateral position of the object is If it is not an approaching state, the possibility of a collision between the object and the host vehicle is reduced.

Therefore, when it is determined that the object is detected only by the radar target LT and the distance between the host vehicle and the object is determined to be a predetermined short distance, the lateral position of the object is determined relative to the host vehicle. The vehicle control for the object is executed on condition that the vehicle is in a predetermined approach state. In this case, if the possibility of collision between the object and the host vehicle is low, vehicle control is not performed. Therefore, after the object is no longer detected as a fusion target, it is highly reliable while suppressing unnecessary vehicle control. It becomes possible to perform vehicle control on the object.

(3) After transitioning from a state in which an object is detected as a fusion target to a state in which an object is detected only with a radar target LT, the lateral position of the object in the vehicle width direction is obtained using the radar target LT. Therefore, the lateral position of the object at that time can be acquired with high accuracy.

(4) When a transition is made from a state in which an object is detected as a fusion target to a state in which the object is detected only by a radar target LT, the lateral position of the object acquired as a fusion target and the radar target LT If there is a large difference from the lateral position of the acquired object, the possibility that the object has moved is high. Therefore, in this case, unnecessary vehicle control is suppressed by not performing vehicle control of the host vehicle.

(5) When the relative speed between the object and the host vehicle increases, the collision margin time TTC calculated by dividing the distance by the relative speed decreases, so vehicle control starts before the fusion target is not detected. Is likely to be. In other words, the situation in which the image target GT cannot be detected before the vehicle control is started is limited to the situation where the relative speed between the object and the host vehicle is small and the collision margin time TTC is large. Therefore, when it is determined that the object is located at a predetermined short distance, the vehicle control of the host vehicle is performed on the condition that the relative speed is low, so that the object is not detected by the fusion target. After that, unnecessary vehicle control can be suppressed, and vehicle control for an object with high reliability can be performed.

(6) The reliability of the object decreases after a predetermined time elapses after the transition to the state in which the object is detected only by the radar target LT. Therefore, in this case, the vehicle control is not performed on the object regardless of whether the distance to the object is a predetermined short distance, so that unnecessary vehicle control is performed on the object with reduced reliability. Can be avoided.

(7) When the object is no longer detected as a fusion target due to the fact that the distance to the object is a short distance, vehicle control is not performed, so the reliability of object detection decreases. It is possible to avoid unnecessary vehicle control from being performed in a situation where there is a possibility of being.

(Second Embodiment)
Next, the second embodiment will be described focusing on the differences from the first embodiment. In the first embodiment, when the image target GT is lost and the distance from the object is a short distance, the speaker, the seat belt, the brake, etc. are operated as vehicle control. However, the second embodiment In such a case, the brake is operated to avoid collision with an object.

As shown in FIG. 5B, the ECU 10a (see also FIG. 1A) of the second embodiment further includes a pedestrian determination unit 107 that determines that the object is a pedestrian, but includes a state determination unit 101 and a distance determination unit 102. The lateral position acquisition unit 103, the lateral position determination unit 104, and the relative speed determination unit 105 are common to the first embodiment. As will be described in detail later, the vehicle control unit 106a of the present embodiment performs primary brake control based on the first margin, which is a margin of collision against an object, in addition to the function of the vehicle control unit 106 of the first embodiment. At the same time, the secondary brake control is performed based on the second margin having a collision margin smaller than the first margin.

Here, as a brake control for avoiding a collision performed by the vehicle control unit 106a, a brake control in two stages of a pre-braking (FPB) and an intervention braking (PB) has been proposed. Specifically, when there is a possibility of collision with an object, first, weak braking is performed as FPB. This FPB operation speeds up the PB braking force and alerts the driver. Then, if the avoidance operation by the driver is not performed even by the operation of FPB, a stronger brake is implemented as PB. In other words, the PB is performed under a situation where the possibility of collision is higher than that of the FPB, and the braking force by the PB is generally set larger than that of the FPB. As described above, since the PB operates in a more limited situation, the operating condition of the PB is set to be stricter than that of the FPB.

In the present embodiment, the FPB is activated when the lateral position of the object specified by the fusion target is equal to or less than a predetermined first threshold Th1 and the TTC is equal to or less than the FPB activation timing Th7. Set to do. As the condition is satisfied, the FPB flag is set to “1”. However, the FPB is set so as not to operate at an operation timing Th8 or less of the PB. In other words, PB is preferentially operated under the circumstances.

On the other hand, the PB operation condition is set so that the PB operates when the lateral position of the object is equal to or less than a predetermined fifth threshold Th5 and TTC is equal to or less than the PB operation timing Th8. As the condition is satisfied, the PB flag is set to “1”.

Here, the threshold of TTC is set such that the FPB operation timing Th7 is larger than the PB operation timing Th8. That is, the TTC threshold is set so that FPB is performed in a situation where the collision margin is high, in other words, in a situation where the possibility of collision is low. Note that after the FPB is activated, the FPB operation is stopped by a driver operation (for example, turning of the steering wheel or a brake operation). Not.

In this embodiment, FPB corresponds to “primary brake control”, PB corresponds to “secondary brake control”, operation timing Th7 corresponds to “first margin”, and operation timing Th8 corresponds to “first brake control”. This corresponds to “2 margin”.

By the way, when the image lost FSN occurs and the PB operating condition is satisfied, if the FPB is already operating, the object has been detected as a contact avoidance target before the PB operating. become. That is, it can be said that the object is continuously detected, but the reliability of the detected object is considered high. On the other hand, when the image lost FSN occurs and the PB operating condition is satisfied, if the FPB is not operating, the object is not detected as a contact avoidance target before the PB operating. become. That is, it is considered that the object is not detected in advance and the reliability of the detected object is low. As a case where such a situation occurs, for example, erroneous detection by the radar sensor 22 or detection due to an abrupt approach of an object from the side can be considered, and the operation of the PB based on these may become an unnecessary operation. is there.

Therefore, in this embodiment, the PB is activated on the condition that the FPB is activated when it is determined that the image lost FSN has occurred at a short distance. That is, at the time when the PB is activated, if the FPB has been activated in advance, the PB is activated assuming that there is a possibility of collision. If the FPB has not been activated in advance, the possibility of collision is low and the PB is not activated. That is, when the FPB is operating, the operation of the PB is permitted, and when the FPB is not operating, the operation of the PB is prohibited.

On the other hand, when the FPB is operating, the FPB operating condition is not satisfied and the FPB ends due to a change in the temporary behavior of the object (for example, the lateral position of the object temporarily exceeds the threshold). There is. In this case, it is possible that the PB operating condition is satisfied when the lateral position of the object thereafter becomes equal to or less than the threshold value again. In such a case, if the object is within a predetermined time T from the end of FPB, it is considered that the reliability of the detected object is high.

Therefore, in this embodiment, even after the operation of the FPB is stopped, a predetermined delay is provided to allow the execution of the PB. Specifically, when it is determined that the image lost FSN has occurred at a short distance, even if FPB is not performed, PB is set on the condition that there is a history that FPB was completed within the predetermined time T immediately before. I tried to do it. In the present embodiment, the brake experience flag is used as the history. The brake experience flag is set to “1” when the operation of the FPB is finished, and is reset to “0” after a predetermined time T has elapsed.

FIG. 5A shows a logic circuit 40 for determining permission and prohibition of PB operation in the present embodiment. The logic circuit 40 includes an NAND circuit C1 that inputs an FPB flag and a signal of the PB flag, an AND circuit C2 that inputs an output signal of the NAND circuit C1 and an inverted signal of a brake experience flag, an output signal of the AND circuit C2, an image lost And an AND circuit C3 for inputting an FSN control signal and a short distance determination signal. If the signal output from the AND circuit C3 is “1”, the operation of the PB is prohibited, and if the signal is “0”, the operation of the PB is permitted.

In the logic circuit 40, the NAND circuit C1 outputs “0” when both the FPB flag and the PB flag are “1”, and “1” when at least one of the FPB flag and the PB flag is “0”. Is output. The AND circuit C2 outputs “1” only when the output signal of the NAND circuit C1 is “1” and the brake experience flag is “0”.

Here, as described above, “1” is input as the control signal for the image lost FSN when the image target GT is lost from the fusion state (image lost FSN). When the image lost FSN is not in the image lost FSN state, “1” is input. "0" is input. In addition, as described above, the short distance determination signal is input when “1” is input when the distance of the object specified by the radar target LT is equal to or smaller than the third threshold Th3 and is larger than the third threshold Th3. “0” is input.

The AND circuit C3 outputs “1” only when the output signal of the AND circuit C2 is “1”, the image is in the lost FSN state, and the distance from the object is a short distance. That is, when the image lost state occurs at a short distance, if the FPB is not operating and there is no braking experience, the operation of the PB is prohibited.

Next, the case where the operation of PB is permitted and the case where it is prohibited will be described with reference to FIGS. here,
(1) After the FPB is activated, the scene where the PB is activated under the FPB operating state;
(2) A scene in which FPB is once terminated after FPB is activated, and then PB is activated,
(3) The state of the PB operation will be described for each of the scenes where the PB operation is prohibited due to a sudden approach of a pedestrian or the like from the side. Among these, FIGS. 6, 8, and 10 are plan views in which the TTC (collision margin time) is the vertical axis and the horizontal position with respect to the host vehicle in the vehicle width direction is the horizontal axis. On the vertical axis, FPB operation timing Th7 and PB operation timing Th8 are provided. In this case, when the TTC for the object falls below the respective threshold values of PB or FPB, it is determined that each operation timing is reached.

The horizontal axis is provided with an operating width W which is an operating range in the vehicle width direction of FB and PB. The operating width W is set by adding a predetermined length to the width of the host vehicle. 6, 8, and 10, when a pedestrian as an object is located within the operating width W, the predetermined lateral position (Th1, Th5) is satisfied. 6, 8, and 10 assume a scene in which a pedestrian approaches the vehicle along a broken arrow when the vehicle travels in the traveling direction. 7, 9, and 11 show timing charts corresponding to FIGS. 6, 8, and 10, respectively.

6 and 7 show a scene in which the PB operates under the FPB operating state after the FPB operates. In a state where a pedestrian is detected within the operating width W, when the TTC becomes the operating timing Th7 or less at the timing t11, the FPB flag is set to “1” and the FPB is operated. Thereafter, the distance between the pedestrian and the host vehicle further approaches, and at timing t12, the state changes to the image lost FSN state ("0" → "1"). At this time, the short distance determination signal is “1”, and the output signal of the AND circuit C3 is “1” (the same applies to timing t22 in FIG. 9 and timing t32 in FIG. 11). At time t13, when the TTC becomes equal to or less than the operation timing Th8, the PB flag is set to “1”, the FPB is ended, and the PB is operated.

At timing t13, the output signal of the NAND circuit C1 becomes “0” by setting the PB flag to “1” while the FPB flag is “1”. Accordingly, the output signals of the AND circuit C2 and the AND circuit C3 become “0”, so that the operation of the PB is permitted. That is, in the image lost FSN state, when it is determined that the distance to the object is a short distance, the PB is executed on the condition that the FPB is executed, that is, the FPB flag is “1”. .

8 and 9 show scenes in which the FPB is temporarily ended after the FPB is activated, and then the PB is activated. As in FIGS. 6 and 7, the FPB is activated at timing t21, and transitions to the image lost FSN state at timing t22 (“0” → “1”). Thereafter, when the pedestrian deviates from the operating width W at timing t23, the FPB flag is reset to “0” and the FPB ends. At this time, the brake experience flag is set to “1”. At time t24, the object is detected again within the operating width W. At this time, since the interval between the timing t23 and the timing t24 is within a predetermined time T (for example, 0.7 msec), the brake experience flag is maintained at “1”, whereby the PB is activated.

In such a case, in the logic circuit 40, when the PB flag is set to “1”, the brake experience flag is already “1”, and based on this, the output signals of the AND circuit C2 and the AND circuit C3 are “ “0”, the operation of PB is permitted. That is, in the image lost FSN state, when it is determined that the distance to the object is a short distance, there is a history that the FPB has been finished within the predetermined time T immediately before, that is, the brake experience flag is “1”. PB is performed on the condition.

10 and 11 show a scene in which the operation of the PB is prohibited due to a sudden approach of a pedestrian from the side. In this case, the TTC is equal to or less than the operation timing Th7 at the timing t31, but since the pedestrian is not detected within the operation width W, the FPB flag is maintained in the “0” state. At the subsequent timing t32, the state is changed to the image lost FSN state ("0" → "1"). At time t33, the TTC becomes equal to or less than the operation timing Th8. However, since the pedestrian is not detected within the operation width W, the PB flag is maintained at “0”. When a pedestrian is detected within the operating width W at timing t34, the PB flag is set to “1”. However, at this time, since the FPB is not operating (FPB flag is not “1”) and the brake experience flag is not set to “1”, the operation of the PB is prohibited. That is, in the image lost FSN state, when it is determined that the distance to the object is a short distance, FPB has not been performed and there is no history that FPB has been completed within the predetermined time T immediately before , FPB is not implemented.

In the case where an object suddenly enters from the side as shown in FIGS. 10 and 11, for example, a pedestrian can be considered as the object. Even if the pedestrian moves toward the host vehicle at a position near the host vehicle, the pedestrian can easily stop or change direction. For this reason, if PB is performed in such a case, there is a high possibility of unnecessary operation. For example, FIG. 12 shows a logic circuit 50 obtained by further adding a pedestrian determination signal to the logic circuit 40. Here, a pedestrian determination signal is input to the AND circuit C3. As the determination signal of the pedestrian, “1” is input when it is determined that the object is a pedestrian, and “0” is input when it is determined that the object is not a pedestrian. In other words, when the object is a pedestrian, the unnecessary operation of PB can be suppressed by acting so as to prohibit the operation of PB.

Also, the operating width W can be made variable according to the type of object, the lateral speed of the object, and the like. For example, the operating width W and the lateral speed of the object have a relationship as shown in FIG. As shown in FIG. 13, the operating width W increases as the lateral speed increases until the lateral speed of the object is equal to or lower than a predetermined value. On the other hand, when the lateral velocity of the object becomes larger than a predetermined value, the operating width W becomes constant at the upper limit value.

According to the above, in addition to the effects of the first embodiment, the present embodiment can achieve the following effects.

When the image target GT is lost and the distance to the object is a short distance, PB is performed if FPB is performed in advance, and PB is not performed unless FPB is performed in advance. did. In this case, it is considered that the reliability of the detected object is high when the FPB is performed in advance, whereas the reliability is low when the FPB is not performed in advance. Therefore, according to this structure, when an object is no longer detected by a fusion target, unnecessary PB can be suppressed and brake control by PB can be appropriately performed.

(Other embodiments)
The present invention is not limited to the above, and may be implemented as follows. In the following description, the same components as those described above are denoted by the same reference numerals, and detailed description thereof is omitted.

(A1) When the relative speed between the object and the host vehicle is large, it is assumed that the image target GT is lost after the operation of the controlled object 30 by the vehicle control is started. In this case, the ECU 10 may continue the operation of the controlled object 30 by continuing the vehicle control even after the image target GT is lost.

When it is determined that the object has been detected only by the radar target LT under the situation where the object is detected as the fusion target and the vehicle control is being performed, the vehicle control of the own vehicle is continued. Thus, when the fusion target cannot be detected, the vehicle control is suddenly stopped and the occurrence of inconvenience such as the sudden stop of the operation of the controlled object 30 can be suppressed.

(A2) When the image target GT is lost and the object specified by the fusion target is in a predetermined approaching state with respect to the host vehicle, the ECU 10 may perform vehicle control. For example, when the distance between the own lane O and the lateral position at the time when the image target GT is lost is within a predetermined value, the vehicle control may be performed after the image target GT is lost.

(A3) In the flowchart of FIG. 4 described above, when it is determined that the image target GT is lost, vehicle control is performed based on at least whether the distance to the object is less than the threshold value. Whether or not each determination condition of S17 to S20 may be omitted.

(A4) The orientation of the imaging center axis of the image sensor 21 can be changed according to the change in the loaded weight in the host vehicle. Therefore, the short distance position where the image target GT is lost changes. Therefore, in the flowchart of FIG. 4, the third threshold Th3 of the distance used in the determination of S16 may be variably set according to a change in the orientation of the imaging center axis of the image sensor 21 in the host vehicle. The change in the imaging center axis may be obtained based on the detection value of a weight sensor provided in the vehicle. In this case, as the vehicle loading weight increases, the rear side of the vehicle sinks with respect to the front side of the vehicle, and the imaging central axis faces upward, so the third threshold Th3 is reduced. In this way, when the third threshold Th3 is set in consideration of the change in the orientation of the imaging center axis of the image sensor 21, the inter-vehicle distance at which the image target GT is lost can be determined more accurately. In addition to this, the inter-vehicle distance at which the image target GT is lost may be obtained in consideration that the mounting height of the image sensor 21 changes in accordance with the change in the loaded weight in the host vehicle.

(A5) In the second embodiment, TTC is used as the operation timing of PB and FPB. However, the TTC is not limited to this as long as it represents a collision margin. For example, a distance from an object based on TTC is set. It is good also as a structure to use.

(A6) In the second embodiment, the object to be controlled by the vehicle is a pedestrian. However, the object is not limited to the pedestrian, but may be another vehicle, an obstacle on the road, or the like.

Claims (12)

1. The fusion target is obtained by fusing the first target information of the object ahead of the host vehicle acquired as a reflected wave of the carrier wave and the second target information of the object acquired by image processing of the captured image ahead of the host vehicle. A vehicle control device (10, 10a) for performing vehicle control of the host vehicle with respect to the object detected as the fusion target,
   A state determination unit (101) for determining whether or not the object has been changed from a state detected by the fusion target to a state detected only by the first target information;
   Distance determination for determining whether or not the distance from the object is a predetermined short distance when it is determined by the state determination unit that the object has been detected only by the first target information. Part (102),
   A vehicle control unit (106, 106a) for performing vehicle control on the object when the distance determination unit determines that the distance to the object is a predetermined short distance;
   A vehicle control device (10, 10a) comprising:
2. A lateral position acquisition unit (103) for acquiring a lateral position of the object in the vehicle width direction;
   When the distance determination unit (102) determines that the distance to the object is a predetermined short distance, the lateral position of the object acquired by the lateral position acquisition unit in the vehicle width direction is A lateral position determination unit (104) for determining whether or not the vehicle is in a predetermined approach state,
   When the lateral position determining unit (104) determines that the lateral position of the object in the vehicle width direction is in a predetermined approach state with respect to the host vehicle, the vehicle control unit (106, 106a) The vehicle control device (10, 10a) according to claim 1, which performs vehicle control on the vehicle.
3. The horizontal position acquisition unit (103) determines that the state has been detected from the state in which the object is detected as the fusion target to the state in which the object is detected only by the first target information. The vehicle control device (10, 10a) according to claim 2, wherein a lateral position of the object in the vehicle width direction is acquired using first target information.
4. When the lateral position acquisition unit (103) transitions from a state in which the object is detected as the fusion target to a state in which the object is detected only by the first target information, the lateral position acquisition unit (103) immediately before the transition The first lateral position of the object is acquired using a fusion target, and the second lateral position of the object is acquired using the first target information immediately after the transition,
   The lateral position determination unit (104) determines whether or not a difference between the first lateral position and the second lateral position is a predetermined value or more,
   The vehicle control unit (106, 106a) determines that the difference between the first lateral position and the second lateral position is greater than or equal to a predetermined value by the lateral position determination unit (104). The vehicle control device (10, 10a) according to claim 3, wherein vehicle control is not performed.
5. A relative speed determination unit (105) for determining whether a relative speed between the object and the host vehicle is smaller than a predetermined value;
   When the vehicle control unit (106, 106a) determines that the object is at a predetermined short-distance position on the condition that the relative speed between the object and the host vehicle is smaller than the predetermined value, The vehicle control device (10, 10a) according to any one of claims 1 to 4, wherein vehicle control is performed on the object.
6. The vehicle control unit (106, 106a) detects the object only by the first target information in a situation where the object is detected as the fusion target and vehicle control is performed on the object. The vehicle control device (10, 10a) according to any one of claims 1 to 5, wherein the vehicle control for the object is continued when it is determined that the state has been changed.
7. The vehicle control unit (106, 106a), after a lapse of a predetermined time from the transition from the state in which the object is detected as the fusion target to the state in which the object is detected only by the first target information The vehicle control device (10, 10a) according to any one of claims 1 to 6, wherein vehicle control is not performed on the object regardless of whether the distance to the object is a predetermined short distance.
8. The vehicle control unit (106, 106a) does not perform vehicle control on the object when the distance determination unit (102) determines that the distance to the object is not a predetermined short distance. The vehicle control device (10, 10a) according to any one of 1 to 7.
9. As the vehicle control, primary brake control is performed based on a first margin that is a collision margin with respect to the object, and a secondary brake is performed based on a second margin that is smaller than the first margin. A vehicle control device (10a) that performs control,
   The vehicle control unit (106a) performs the primary brake control when the distance determination unit (102) determines that the distance to the object is a predetermined short distance, or the primary control The vehicle control device (10a) according to any one of claims 1 to 8, wherein the secondary brake control is performed on the condition that there is a history that the brake control has been finished within a predetermined time immediately before.
10. The vehicle control unit (106a) does not perform the primary brake control when the distance determination unit (102) determines that the distance to the object is a predetermined short distance, and The vehicle control device (10a) according to claim 9, wherein the secondary brake control is not performed when there is no history.
11. A pedestrian determination unit (107) for determining that the object is a pedestrian;
    The vehicle control device (10a) according to claim 9 or 10, wherein the vehicle control unit (106a) permits the execution of the secondary brake control when it is determined that the object is a pedestrian.
12. The fusion target is obtained by fusing the first target information of the object ahead of the host vehicle acquired as a reflected wave of the carrier wave and the second target information of the object acquired by image processing of the captured image ahead of the host vehicle. And a step of performing vehicle control of the host vehicle with respect to the object detected as the fusion target,
    Determining whether or not the object has transitioned from a state in which the object is detected by the fusion target to a state in which the object is detected only by the first target information;
    Determining whether or not the distance to the object when it is determined that the object has been detected using only the first target information is a predetermined short distance;
    Performing vehicle control on the object when it is determined that the distance to the object is a predetermined short distance; and
    A vehicle control method comprising:

Images