JP2020169016A - Collision detection apparatus - Google Patents

Abstract

To enable detection of a collision between the vehicle itself and a pedestrian.SOLUTION: A collision detection ECU 10 performs: starting measuring a timer value t when a collision prediction time TTC becomes equal to or less than an emergency brake start threshold TTCref, and storing the then collision prediction time TTC; determining that a vehicle itself has collided with an object (a pedestrian or bicycle) (S20) in a case where a timer value t is within an object collision prediction permissible period ((TTC-α)-(TTC+α), the timer value that is read when a front Gf detected by a frontal collision occupant-protecting acceleration sensor exceeds a collision determining threshold Gfref (Yes for S19). Here, the collision determining threshold Gfref is set at a value in accordance with a vehicular speed V.SELECTED DRAWING: Figure 4

Description

The present invention relates to a collision detection device that detects a collision between the own vehicle and a pedestrian and a bicycle.

Conventionally, for example, as proposed in Patent Document 1, a collision detection device that detects a collision between a own vehicle and a pedestrian and notifies the outside is known. The collision detection device (called a conventional device) proposed in Patent Document 1 detects a pedestrian located in front of the own vehicle by a camera, and then a collision is detected based on the outputs of the bumper sensor and the acceleration sensor. At that time, it is determined that the own vehicle and the pedestrian have collided, and the fact that the own vehicle and the pedestrian have collided is notified to the outside via the communication device. This notification information includes information indicating the vehicle position and information for identifying the vehicle. As a result, measures for protecting pedestrians can be taken promptly.

Japanese Unexamined Patent Publication No. 2013-156688

However, the conventional device requires a bumper sensor to determine a collision between the own vehicle and a pedestrian. The bumper sensor outputs a signal indicating the amount of displacement that the bumper is displaced by a collision, but it is not installed in a general vehicle. Therefore, the conventional device cannot determine the collision between the own vehicle and the pedestrian in a general vehicle.

The present invention has been made to solve the above problems, and an object of the present invention is to enable detection of a collision between an own vehicle and a pedestrian even in a general vehicle not provided with a bumper sensor. To do.

In order to achieve the above object, the features of the present invention are:
Object detection means (20, 30, S11) that detects pedestrians and bicycles located in front of the own vehicle as objects, and
An acceleration sensor (40) that detects the deceleration of the own vehicle caused by a collision from the front, and
The collision prediction time acquisition means (S12) for acquiring the collision prediction time, which is the prediction time until the object detected by the object detection means collides with the own vehicle,
Timing for determining whether or not the timing at which the deceleration detected by the acceleration sensor exceeds the collision determination threshold is the same as the collision prediction timing determined from the collision prediction time acquired by the collision prediction time acquisition means. When the determination means (S19) and the timing determination means determine that the timing at which the deceleration exceeds the collision determination threshold is the same as the collision prediction timing, the own vehicle becomes the object. It is equipped with a collision determination means (S20) for determining a collision.

The collision detection device of the present invention is a device that detects a collision between a pedestrian and a bicycle as an object and the object and the own vehicle. The collision detection device includes an object detection means, an acceleration sensor, a collision prediction time acquisition means, a timing determination means, and a collision determination means.

The object detecting means detects pedestrians and bicycles located in front of the own vehicle as objects. For example, the object detecting means includes a camera sensor and a radar sensor, and detects an object located in front of the own vehicle.

The acceleration sensor detects the deceleration of the own vehicle caused by a collision from the front. As the acceleration sensor, for example, an acceleration sensor for protecting front occupants for detecting a head-on collision and operating an airbag or the like may be used. Such accelerometers are generally standard equipment on all vehicles.

The collision prediction time acquisition means acquires the collision prediction time, which is the prediction time until the object detected by the object detection means collides with the own vehicle. For example, the collision prediction time is a time calculated by dividing the distance between the object and the own vehicle by the relative speed of the object with respect to the own vehicle.

The timing determination means determines whether or not the timing at which the deceleration detected by the acceleration sensor exceeds the collision determination threshold is the same as the collision prediction timing determined by the collision prediction time acquired by the collision prediction time acquisition means. judge.

The collision detection threshold is set when the deceleration detected by the acceleration sensor does not exceed the collision detection threshold when the own vehicle and the object do not collide, and when the own vehicle and the object collide. , The deceleration detected by the acceleration sensor is set to a value that exceeds the collision determination threshold. Therefore, the collision between the own vehicle and the object can be detected by comparing the deceleration detected by the acceleration sensor with the collision determination threshold value. However, there are cases where the deceleration detected by the acceleration sensor exceeds the collision determination threshold value for reasons other than the collision between the own vehicle and the object. In such a case, it is erroneously determined that the own vehicle has collided with the object.

In order to exclude the case where such an erroneous determination is made, the present invention includes a timing determination means and a collision determination means. The timing determination means determines whether or not the timing at which the deceleration exceeds the collision determination threshold value is the same as the collision prediction timing determined from the collision prediction time acquired by the collision prediction time acquisition means. If the timings of both are the same, it can be estimated that the own vehicle and the object collided.

Therefore, the collision determination means determines that the own vehicle has collided with the object when the timing determination means determines that the timing at which the deceleration exceeds the collision determination threshold value is the same as the collision prediction timing. .. Thereby, according to the present invention, it is possible to detect a collision between the own vehicle and a pedestrian and a collision between the own vehicle and a bicycle by using an acceleration sensor which is generally equipped as standard equipment in all vehicles. Therefore, it is possible to carry out the operation even if the bumper sensor is not provided as in the conventional device.

The "same period" does not mean exactly the same timing, but means a timing that is considered to be substantially the same in consideration of the effects of measurement error, calculation error, fluctuation of vehicle condition, and the like. ..

A feature of one aspect of the present invention is
The collision determination threshold value is set to a value corresponding to the vehicle speed of the own vehicle.

When traveling on a rough road, the higher the vehicle speed of the own vehicle, the greater the deceleration detected by the acceleration sensor due to the input from the road surface vibrates. Therefore, the deceleration detected by the acceleration sensor may exceed the collision determination threshold value, that is, a collision erroneous determination may be performed. Therefore, in one aspect of the present invention, the collision determination threshold value is set to a value corresponding to the vehicle speed of the own vehicle. For example, the collision determination threshold value is set to increase as the vehicle speed of the own vehicle increases in the collision determination speed region. Therefore, according to one aspect of the present invention, it is possible to improve the collision determination accuracy when traveling on a rough road.

A feature of one aspect of the present invention is
In the collision determination means, the timing determination means determines that the timing at which the deceleration exceeds the collision determination threshold is the same as the collision prediction timing (S19: Yes), and the deceleration is When the stop of the own vehicle is detected within a predetermined period after the collision determination threshold is exceeded (S51 to S53), it is determined that the own vehicle has collided with the object. There is.

For example, when the object detecting means cannot correctly recognize the object, the own vehicle collides with the object even if the deceleration exceeds the collision determination threshold value at the same time as the collision prediction timing. It is possible that this is not the case. On the other hand, when the own vehicle collides with the object and the deceleration exceeds the collision determination threshold, and when the own vehicle does not collide with the object and the deceleration exceeds the collision determination threshold, the subsequent cases are followed. The behavior of the own vehicle is different. That is, the driver immediately stops the own vehicle when the own vehicle collides with the object, but otherwise, for example, when the input from the road surface acts on the own vehicle, the deceleration determines the collision. Even if an impact that exceeds the threshold value is generated, the own vehicle is driven as it is.

Therefore, in the collision determination means, after the timing determination means determines that the timing at which the deceleration exceeds the collision determination threshold is the same as the collision prediction timing, and the deceleration exceeds the collision determination threshold. When the stop of the own vehicle is detected within a predetermined period, it is determined that the own vehicle has collided with the object. That is, in addition to the condition that the timing at which the deceleration exceeds the collision determination threshold is the same as the collision prediction timing, the stop of the own vehicle is detected within a predetermined period after the deceleration exceeds the collision determination threshold. It is set as a requirement that it is determined that the own vehicle has collided with an object that the condition of being carried out is satisfied. Thereby, according to one aspect of the present invention, the collision determination accuracy can be further improved.

A feature of one aspect of the present invention is
In the collision determination means, the timing determination means determines that the timing at which the deceleration exceeds the collision determination threshold is the same as the collision prediction timing (S19: Yes), and the deceleration is When a specific operation of the driver is detected within a predetermined period before and after the collision determination threshold is exceeded (S61 to S66), it is configured to determine that the own vehicle has collided with the object. It is in.

For example, the driver may perform a horn operation (an operation of ringing the horn) immediately before the own vehicle collides with an object. For example, when the own vehicle collides with an object, it can be estimated that the driver immediately stops the own vehicle and then performs a shift change operation, a parking brake operation, a door opening operation, and the like.

Therefore, the collision determination means determines by the timing determination means that the timing at which the deceleration exceeds the collision determination threshold is the same as the collision prediction timing, and before and after the deceleration exceeds the collision determination threshold. When a specific operation of the driver is detected within a predetermined period, it is determined that the own vehicle has collided with the object. That is, in addition to the condition that the timing at which the deceleration exceeds the collision determination threshold is the same as the collision prediction timing, the driver's specific operation is detected within a predetermined period before and after the deceleration exceeds the collision determination threshold. It is set as a requirement that it is determined that the own vehicle has collided with an object that the condition of being carried out is satisfied. Thereby, according to one aspect of the present invention, the collision determination accuracy can be further improved. The specific operation is an operation for calling attention to the object (for example, a horn operation) or an operation that can be estimated to be performed when the own vehicle is stopped (for example, a shift operation, a side brake operation, etc.). It is advisable to adopt at least one of (door opening operation, etc.).

In the above description, in order to help the understanding of the invention, the reference numerals used in the embodiments are attached in parentheses to the constituent elements of the invention corresponding to the embodiment. It is not limited to the embodiment defined by.

It is a schematic system configuration diagram of the collision detection device which concerns on this embodiment. It is a graph which shows an example of the waveform of the front Gf detected by the acceleration sensor for front occupant protection. It is a graph which shows the collision determination threshold value map. It is a flowchart which shows the collision determination routine which concerns on embodiment. It is explanatory drawing which shows the image of the collision determination operation. It is a flowchart which shows the part which concerns on the modification 1 in the collision determination routine. It is a flowchart which shows the part which concerns on the modification 2 in the collision determination routine. It is a graph which shows the collision determination threshold value map which concerns on modification 2. It is a top view explaining the case which erroneously determines that the own vehicle collided with a pedestrian. It is a top view explaining another case which erroneously determines that the own vehicle collided with a pedestrian. It is a top view explaining the difference in the behavior of the own vehicle between the case where the own vehicle collides with a pedestrian (a) and the case where the own vehicle passes through a rough road portion (b). It is a flowchart which shows the part which concerns on the modification 3 in the collision determination routine. It is a flowchart which shows the part which concerns on the modification 4 in the collision determination routine.

Hereinafter, the collision detection device according to the embodiment of the present invention will be described with reference to the drawings.

The collision detection device according to the embodiment of the present invention is applied to a vehicle (hereinafter, may be referred to as a "own vehicle" to distinguish it from other vehicles). As shown in FIG. 1, the collision detection device 1 includes a collision detection ECU 10. A camera sensor 20, a radar sensor 30, a vehicle speed sensor 40, a front occupant protection acceleration sensor 50, and another occupant protection acceleration sensor 60 are connected to the collision detection ECU 10.

The collision detection ECU 10 is an electric control unit (Electric Control Unit) including a microcomputer as a main part, and is connected to other ECUs so as to be able to transmit and receive information via a CAN (Controller Area Network) (not shown). Has been done. In the present specification, the microcomputer includes a CPU, ROM, RAM, non-volatile memory, interface I / F, and the like. The CPU realizes various functions by executing instructions (programs, routines) stored in ROM.

The camera sensor 20 photographs the front of the own vehicle from the vehicle interior, detects a three-dimensional object existing in front of the own vehicle by processing the captured image, and detects pedestrians and bicycles from the three-dimensional objects. It is extracted and information (camera information) such as their position and size is supplied to the collision detection ECU 10.

The radar sensor 30 is, for example, a millimeter-wave radar sensor, which transmits radio waves in the millimeter-wave band in front of the own vehicle and receives the reflected wave, so that the own vehicle is based on the transmitted wave and the received wave. The relative position (distance, direction) with the three-dimensional object existing in front of the is measured, and the relative speed with the three-dimensional object is measured. The radar sensor 30 supplies the three-dimensional object information (radar information), which is the measurement result, to the collision detection ECU 10.

Hereinafter, the camera sensor 20 and the radar sensor 30 may be collectively referred to as a peripheral monitoring sensor.

The collision detection ECU 10 recognizes pedestrians and bicycles existing in front of the own vehicle as objects based on the camera information supplied from the camera sensor 20 and the radar information supplied from the radar sensor 30, and also targets the target. Grasp the relative position (distance, direction) of the object with respect to the own vehicle, and the relative speed with respect to the own vehicle. Based on such relative position and relative speed, the collision detection ECU 10 extracts objects having a high possibility of colliding with the own vehicle from the objects existing in front of the own vehicle, and sets them as the objects for collision determination. ..

The collision detection ECU 10 acquires the collision prediction time TTC for the object. Collision prediction time TTC is the predicted time from the present time until the own vehicle collides with the object, and is the distance d between the current object and the own vehicle and the relative velocity Vr of the object with respect to the own vehicle. Based on this, it is calculated by the following equation (1). The relative velocity Vr is the velocity in the direction in which the own vehicle and the object approach each other.
TTC = d / Vr ・ ・ ・ (1)

This collision prediction time TTC is used as an index showing the high possibility that the own vehicle collides with the object, and the smaller the value, the higher the possibility (risk) that the own vehicle collides with the object. .. The collision prediction time TTC changes from moment to moment.

The camera sensor 20 and the radar sensor 30 are, for example, the camera sensor 20 and the radar used in collision avoidance support control (pre-crash safety control: PCS control) that assists the own vehicle to avoid colliding with an obstacle. The sensor 30 may be shared.

For example, the vehicle provided with the collision detection device 1 of the present embodiment includes a driving support ECU (not shown). The driving support ECU detects an obstacle (any three-dimensional object, not limited to pedestrians and bicycles) based on the information from the camera sensor 20 and the radar sensor 30, and has a very short collision prediction time TTC related to the obstacle. Calculate in the calculation cycle. The driving support ECU alerts the driver by sounding a buzzer when the collision prediction time TTC falls below the warning threshold, and further, when the collision prediction time TTC falls below the brake assist threshold, the brake hydraulic pressure subsidy is provided. Then, when the collision prediction time TTC falls below the emergency braking threshold value, the automatic brake is activated. As a result, it is possible to assist the driver's driving operation so that the own vehicle does not collide with an obstacle or the damage is reduced even if the vehicle collides.

The collision detection ECU 10 acquires information representing the collision prediction time TTC calculated by the driving support ECU. The division of roles between the collision detection ECU 10 and the driving support ECU can be arbitrarily set. For example, it is possible to incorporate the function of the collision detection ECU 10 into the driving support ECU. Further, the collision detection ECU 10 may calculate the collision prediction time TTC of the object.

In the present embodiment, the collision detection ECU 10 targets pedestrians and bicycles whose relative speed Vr with respect to the own vehicle is a predetermined speed (for example, 25 km / h) or more as a collision determination target. In this case, finally, the target object for collision determination is the target for which the driving support ECU operates the automatic brake when the collision prediction time TTC falls below the emergency braking start threshold value (for example, 600 milliseconds). Objects (pedestrians and bicycles). Therefore, pedestrians and bicycles that are very likely to collide are the objects of collision determination by the collision detection ECU 10.

The vehicle speed sensor 40 detects the vehicle speed V of the own vehicle and supplies the vehicle speed signal representing the vehicle speed V to the collision detection ECU 10.

The front collision occupant protection acceleration sensor 50 is, for example, a sensor for detecting a head-on collision (front collision) of the own vehicle and operating a front collision occupant protection device such as an airbag or a seatbelt pretensioner. Outputs a detection signal that represents the deceleration caused by a head-on collision. The front thrust occupant protection acceleration sensor 50 supplies a detection signal to the airbag ECU (not shown). The airbag ECU operates the front collision occupant protection device when the deceleration detected by the front collision occupant protection acceleration sensor 50 exceeds the front collision occupant protection threshold value.

The detection signal output by the front occupant protection acceleration sensor 50 is also supplied to the collision detection ECU 10.

The other occupant protection acceleration sensor 60 is an occupant protection acceleration sensor different from the front occupant protection acceleration sensor 50. For example, it detects a side collision (side collision) of the own vehicle and detects a side airbag or a seat. It is a sensor for operating a side collision occupant protection device such as a belt pretensioner. The other occupant protection acceleration sensor 60 outputs a detection signal indicating the acceleration generated by the side collision. The other occupant protection acceleration sensor 60 outputs a detection signal to the airbag ECU. The airbag ECU operates the side collision occupant protection device when the acceleration detected by the other occupant protection acceleration sensor 60 exceeds the side collision occupant protection threshold value.

The detection signal output by the other occupant protection acceleration sensor 60 is also supplied to the collision detection ECU 10.

The acceleration or deceleration detected by the occupant protection acceleration sensors 50 and 60 both represent the magnitude of the impact force at the time of a collision. Hereinafter, since it is not necessary to distinguish between acceleration and deceleration, they are collectively referred to as acceleration. In addition, when discussing the magnitude relationship of acceleration including deceleration, it is assumed that those absolute values are discussed. Further, the acceleration (deceleration) detected by the front occupant protection acceleration sensor 50 is referred to as a front Gf, and the acceleration detected by another occupant protection acceleration sensor 60 is referred to as another Gs.

<Collision detection method>
The collision detection ECU 10 determines a collision between the own vehicle and the object with the pedestrian and the bicycle as the objects. This collision determination is performed based on the size of the front Gf represented by the detection signal of the front occupant protection acceleration sensor 50. For example, in a situation where an object (pedestrian or bicycle) with a high possibility of collision is detected in front of the own vehicle, the front Gf is compared with the collision determination threshold value Gfref, and the size of the front Gf determines the collision. When the threshold value Gfref is exceeded, it can be determined that the object has collided with the own vehicle. In this case, the collision determination threshold Gfref does not exceed the collision determination threshold when the own vehicle and the object do not collide, and when the own vehicle and the object collide, the front It is necessary to set a value such that Gf exceeds the collision determination threshold value.

In order to detect a head-on collision with a pedestrian or a bicycle, the collision detection threshold Gfref must be set to a value smaller than the front Gf detected when the own vehicle collides with a building or another vehicle. is there. When the collision determination threshold value Gfref is set to a small value in this way, there may be a case where the front Gf exceeds the collision determination threshold value Gfref due to an event other than the collision between the own vehicle and the object.

(Case 1) For example, during braking, the front Gf may exceed the collision determination threshold value Gfref. In addition, the front Gf may exceed the collision determination threshold value Gfref due to electrical noise or road noise (noise that rides on the front Gf due to input from the road surface). Therefore, there is a risk that the collision determination between the own vehicle and the object may be erroneous.

(Case 2) Even when driving on a rough road, the front Gf may exceed the collision determination threshold value Gfref. Therefore, there is a risk that the collision determination between the own vehicle and the object may be erroneous.

(Case 3) There is a possibility that a collision between an obstacle other than the object and the own vehicle may be mistakenly determined as a collision between the own vehicle and the object.

(For case 1)
The deceleration generated during braking is superimposed on the front Gf as a low frequency component. In addition, electrical noise is also superimposed on the front Gf as a low frequency component. Therefore, the collision detection ECU 10 reduces the influence of deceleration at the time of braking and the influence of electrical noise at the time of collision determination by removing the low frequency component from the front Gf by using a bandpass filter. Further, the road noise is superimposed on the front Gf as a high frequency component. Therefore, the collision detection ECU 10 reduces the influence of road noise at the time of collision determination by removing a high frequency component from the front Gf by using a bandpass filter.

Therefore, the influence of deceleration, the influence of electrical noise, and the influence of road noise detected during braking can be reduced by performing bandpass filter processing.

(For case 2)
Bandpass filtering alone cannot solve problems when driving on rough roads. FIG. 2 shows an example of the waveform of the front Gf detected by the front occupant protection acceleration sensor 50. Line L1 (thin solid line) represents the waveform of the front Gf when traveling on a continuous rough road at 25 km / h. The line L2 (thin dotted line) represents the waveform of the front Gf when traveling on a continuous rough road at 60 km / h. The line L3 (thick solid line) represents the waveform of the front Gf when an object (pedestrian or bicycle) collides with the own vehicle while traveling on a flat road at 25 km / h. The line L4 (thick broken line) represents the waveform of the front Gf when the object collides with the own vehicle while traveling on a flat road at 60 km / h.

In order to detect that the front Gf exceeds the collision determination threshold value Gfref and determine that the object has collided with the own vehicle, the collision determination threshold value Gfref is set to the peak value (peak value) of the lines L3 and L4. ) Must be set to a value smaller than (on requirement). On the other hand, in order to prevent an erroneous determination that the object has collided with the own vehicle when traveling on a rough road, the collision determination threshold value Gfref is set to the peak value (peak value) of the lines L1 and L2. ) Must be set to a value greater than (off requirement).

However, as can be seen from FIG. 2, there is no collision determination threshold value Gfref that satisfies both the on requirement and the off requirement. Therefore, in the present embodiment, the collision determination threshold value Gfref is set according to the vehicle speed V. FIG. 3 represents a collision determination threshold map. The collision determination threshold map is a map for setting a collision determination threshold Gfref according to the vehicle speed V. This collision determination threshold map has a characteristic of setting a collision determination threshold value Gfref that increases as the vehicle speed V increases when the vehicle speed V exceeds a predetermined vehicle speed in the collision determination vehicle speed region. In the low vehicle speed range where collision determination is not performed, a high collision determination threshold Gfref (collision determination threshold Gfref that does not exceed the front Gf) is set, so that the collision determination of the object (target) is substantially achieved. Judgment that an object has collided with the own vehicle) is not performed. The collision detection ECU 10 stores the collision determination threshold map, and sets the collision determination threshold Gfref according to the current vehicle speed V.

Therefore, it is possible to set the collision determination threshold value Gfref that satisfies both the on requirement and the off requirement. As a result, the front Gf can be prevented from exceeding the collision determination threshold value Gfref even when traveling on a rough road.

(For case 3)
If the collision prediction timing determined by the collision prediction time TTC of the object and the timing when the front Gf exceeds the collision determination threshold value Gfref (called the actual detection timing) are the same time, it is said that the own vehicle has collided with the object. Can be estimated. On the other hand, if the collision prediction timing of the object and the actual detection timing are not the same, it can be estimated that the own vehicle has not collided with the object. Therefore, by determining the matching between the collision prediction timing and the actual collision detection timing, the collision determination between the own vehicle and the object can be appropriately performed.

In the present embodiment, the time t required for the front Gf to exceed the collision determination threshold Gfref from the time when the collision prediction time TTC is determined to be below the emergency braking start threshold value (for example, 600 milliseconds). , When the collision prediction time TTC at the time when it is determined that the value has dropped below the emergency braking start threshold value is almost equal ((TTC-α) ≤ t ≤ (TTC + α)), the collision prediction timing and the actual detection timing are at the same time. Is determined to be. When such a determination is made, it is determined that the own vehicle has collided with the object. In addition, α is an error tolerance set in consideration of the influence of measurement error, calculation error, fluctuation of vehicle condition, and the like.

<Collision detection routine>
Next, a specific process for determining the collision between the own vehicle and the object will be described. FIG. 4 shows a collision determination routine executed by the collision detection ECU 10. The collision detection ECU 10 starts a collision determination routine when the ignition switch is turned on.

When the collision determination routine is started, the collision detection ECU 10 determines in step S11 whether or not an object (pedestrian and bicycle) is detected. That is, it is determined whether or not a pedestrian or a bicycle is detected as a collision avoidance target of the PCS control performed by the driving support ECU.

When the object is not detected (S11: No), the collision detection ECU 10 returns the process to step S11 and repeats the above-described process.

When such a process is repeated and an object is detected (S11: Yes), the collision detection ECU 10 advances the process to step S12 to acquire the collision prediction time TTC. The collision prediction time TTC may be acquired by reading the value calculated by the driving support ECU, or may be acquired by the collision detection ECU 10 by calculation.

Subsequently, in step S13, the collision detection ECU 10 determines whether or not the collision prediction time TTC is equal to or less than the emergency braking start threshold value TTCref. When the collision prediction time TTC is larger than the emergency brake start threshold value TTCref (S13: No), the collision detection ECU 10 returns the process to step S11 and repeats the above process.

When such a process is repeated and the collision prediction time TTC becomes equal to or less than the emergency brake start threshold value TTCref (S13: Yes), the collision detection ECU 10 advances the process to step S14 and starts timing by the timer. This timed timer value is called a timer value t.

Subsequently, in step S15, the collision detection ECU 10 acquires the vehicle speed V detected by the vehicle speed sensor 40 and the front Gf detected by the front occupant protection acceleration sensor 50. All of these sensor values are the latest values (current sensor values).

Subsequently, in step S16, the collision detection ECU 10 sets a collision determination threshold value Gfref according to the vehicle speed V with reference to the collision determination threshold value map. Subsequently, in step S17, the collision detection ECU 10 determines whether or not the front Gf is larger than the collision determination threshold value Gfref. In this case, the collision detection ECU 10 applies bandpass filter processing to the detection signal of the front occupant protection acceleration sensor 50, and uses the deceleration value represented by the signal after the bandpass filter processing as the front Gf.

When the front Gf is equal to or less than the collision determination threshold value Gfref (S17: No), the collision detection ECU 10 advances the process to step S18 and determines whether or not the timer value t is larger than the determination end time (TTC + α). To do. Here, the collision prediction time TTC is a value when it is determined as "Yes" in step S13, that is, a value when it is determined that the collision prediction time TTC has fallen below the emergency braking start threshold value TTCref (automatic braking is activated). Value at the timing of the operation). The collision detection ECU 10 stores the collision prediction time TTC when it is determined in step S13 that the collision prediction time TTC has fallen below the emergency braking start threshold TTCref, and the stored collision prediction time TTC is used in this step S18. And it is used in the process of step S19 described later.

When the timer value t is equal to or less than the determination end time (TTC + α) (S18: No), the collision detection ECU 10 returns the process to step S15 and repeats the above-described process. Therefore, the comparison between the collision determination threshold value Gfref set according to the vehicle speed V and the front Gf is repeated.

When such a process is repeated and the front Gf becomes larger than the collision determination threshold value Gfref (S17: Yes), the collision detection ECU 10 advances the process to step S19, and the timer value t allows the object collision prediction. Judge whether the value is within the period. The object collision prediction allowable period is a period between the value obtained by subtracting the error tolerance α from the collision prediction time TTC and the value obtained by adding the error tolerance α to the collision prediction time TTC ((TTC-α) to (TTC + α). )).

When the own vehicle collides with an object, the front Gf exceeds the collision determination threshold value Gfref. At the timing when the front Gf exceeds the collision determination threshold value Gfref, it can be estimated that the timer value t indicates a value within the object collision prediction allowable period ((TTC-α) to (TTC + α)). On the contrary, if the front Gf exceeds the collision determination threshold value Gfref before the timer value t reaches the value within the object collision prediction allowable period ((TTC-α) to (TTC + α)), the own vehicle It can be presumed that the obstacle that collided with the object is not an object. Alternatively, it can be estimated that the front Gf exceeds the collision determination threshold value Gfref for reasons other than collision.

When the collision detection ECU 10 determines "Yes" in step S19, that is, the collision prediction timing determined by the collision prediction time TTC of the object and the actual detection timing when the front Gf exceeds the collision judgment threshold Gfref are at the same time. If it is determined that there is, the process proceeds to step S20, and it is determined that the own vehicle has collided with the object.

Subsequently, the collision detection ECU 10 proceeds to the process in step S21 to notify the service center of the collision occurrence information. The collision occurrence information includes information indicating that the own vehicle has collided with a pedestrian or a bicycle, the position information of the own vehicle (position information acquired by the GPS receiver), and vehicle information for identifying the own vehicle. There is. In this case, the collision detection ECU 10 transmits collision occurrence information to the service center via an in-vehicle communication device (not shown). When the service center receives the collision occurrence information, it requests the dispatch of an ambulance to the position of its own vehicle.

The collision detection ECU 10 ends the collision determination routine when the transmission of the collision occurrence information to the service center is completed.

On the other hand, when "No" is determined in step S19, that is, the collision prediction timing determined by the collision prediction time TTC of the object and the actual detection timing when the front Gf exceeds the collision judgment threshold Gfref are not at the same time. If it is determined, the collision detection ECU 10 proceeds to the process in step S22 and determines that the object and the own vehicle do not collide.

Further, when it is determined as "Yes" in step S18, that is, when the timer value t exceeds the determination end time (TTC + α) without the front Gf becoming larger than the collision determination threshold value Gfref, the object and the own vehicle. It can be estimated that there is no collision with. Therefore, even in this case, the collision detection ECU 10 proceeds to the process in step S22.

When the collision detection ECU 10 performs the process of step S22, the collision determination routine ends.

FIG. 5 is an explanatory diagram showing the collision determination operation using the passage of time. In situation 1, an object (pedestrian or bicycle) is detected in front of the own vehicle. When the own vehicle approaches the object and the collision prediction time TTC drops below the emergency braking start threshold TTCref, automatic braking is started (Situation 2). At this point, the collision prediction time TTC is stored and time counting by the timer is started. Further, the comparison between the front Gf and the collision determination threshold value Gfref is started.

Then, when the timer value t at the time when it is determined that the front Gf becomes larger than the collision determination threshold value Gfref is within the object collision prediction allowable period ((TTC-α) to (TTC + α)), the object. Is determined to have collided with the own vehicle (Situation 3).

According to the collision detection device of the present embodiment described above, since the matching between the collision prediction timing and the actual detection timing is determined, the collision determination between the object and the own vehicle can be appropriately performed. Further, since the collision determination threshold value Gfref that increases as the vehicle speed V increases is set using the collision determination threshold value map, the front Gf does not exceed the collision determination threshold value Gfref even when traveling on a rough road. It is possible to improve the collision determination accuracy when traveling on a rough road. Further, by performing bandpass filter processing on the detection signal of the front occupant protection acceleration sensor 50, the influence of noise superimposed on the front Gf can be reduced.

As a result, the acceleration sensor (acceleration sensor 50 for protecting front occupants), which is generally equipped as standard equipment on all vehicles, is used to detect a collision between the own vehicle and a pedestrian and a collision between the own vehicle and a bicycle. Can be done. Therefore, even if a special bumper sensor is not provided as in the conventional device, the collision determination with the object can be appropriately performed.

In the present embodiment, it is determined in real time whether or not the front Gf exceeds the collision determination threshold value Gfref during the automatic braking, and the timer value when the front Gf exceeds the collision determination threshold value Gfref. The collision determination is performed based on whether or not t is a value within the object collision prediction allowable period ((TTC-α) to (TTC + α)), but it is not always necessary to do so. For example, data representing the magnitude relationship between the front Gf and the collision determination threshold Gfref (data indicating whether Gf> Gfref or not) and the timer value t are sampled in association with each other, and the determination end time (TTC + α) is sampled. ) Is exceeded (for example, at the end of automatic braking), and based on the sampling data, the timer value t when the front Gf exceeds the collision determination threshold Gfref is the elephant collision prediction allowable period ((()). It may be configured to determine whether or not the value is within TTC-α) to (TTC + α)).

Further, in the present embodiment, the collision prediction time TTC at the time when the automatic braking is started is used, and the collision determination is performed by comparing the timer value t from that time point with the collision prediction time TTC. There is no need to do so. For example, the collision prediction time TTC at the time when a predetermined event occurs may be used, and the collision determination may be performed by comparing the timer value t measured from that time point with the collision prediction time TTC.

<Modification example 1>
Next, a modification 1 of the collision determination routine will be described. FIG. 6 shows a portion to be changed in the collision determination routine of the embodiment. In this modification 1, in the collision determination routine (FIG. 4) of the embodiment, the processing of steps S31 to S33 is incorporated between steps S14 and S15, and the other measures are the same as those of the embodiment. The same is true. Therefore, here, a process different from the embodiment will be described.

In step S31, the collision detection ECU 10 determines whether or not the own vehicle is traveling on a rough road. In this case, the collision detection ECU 10 sets the front Gf and the other Gs based on the front Gf detected by the front occupant protection acceleration sensor 50 and the other Gs detected by the other occupant protection acceleration sensor 60. When both are continuous, it is determined that the own vehicle is traveling on a rough road (determination method 1). Alternatively, the collision detection ECU 10 determines that the own vehicle is traveling on a rough road when the size of the other Gs detected by the other occupant protection acceleration sensor 60 is larger than the rough road determination threshold value (determination). Method 2). Alternatively, the determination method 1 and the determination method 2 may be combined (for example, OR condition or AND condition) for determination. Further, it may be determined whether or not the vehicle is traveling on a rough road based on the detection signal of a sensor used for vehicle motion control (for example, suspension stroke sensor, vertical acceleration sensor, etc.).

When the collision detection ECU 10 determines that the own vehicle is traveling on a rough road (S31: Yes), the collision detection ECU 10 selects a rough road traveling threshold map as the collision determination threshold map in step S32. On the other hand, when it is not determined that the own vehicle is traveling on a rough road (S31: No), the collision detection ECU 10 selects a normal traveling threshold map as the collision determination threshold map in step S33. When the collision detection ECU 10 selects the threshold map, the collision detection ECU 10 proceeds to the process in step S15. The normal running threshold map and the rough road running threshold map are maps having the same characteristics as the collision determination threshold map (FIG. 3) in the embodiment, and the rough road running threshold map is a normal running threshold map. In comparison, the magnitude of the collision determination threshold value Gfref with respect to the vehicle speed V is set larger. For example, the threshold map for rough road driving has a characteristic of setting the threshold value Gfref for collision determination, which is raised by a predetermined value ΔGfref as compared with the threshold map for normal driving.

Therefore, according to the collision determination routine of the first modification, the collision determination threshold map is switched between rough road driving and normal driving, so an appropriate collision determination threshold Gfref is set according to the driving road conditions. To. Thereby, the collision determination accuracy of the object can be further improved.

<Modification 2>
Next, a modification 2 of the collision determination routine will be described. FIG. 7 shows a portion to be changed in the collision determination routine of the embodiment. In this modification 2, in the collision determination routine (FIG. 4) of the embodiment, the processes of steps S41 to S43 are incorporated between steps S14 and S15, and the other measures are the same as those of the embodiment. The same is true. Therefore, here, a process different from the embodiment will be described.

In step S41, the collision detection ECU 10 acquires the relative velocity Vr of the object with respect to the own vehicle from the peripheral monitoring sensor, and determines whether or not the relative velocity Vr is larger than the map switching threshold value Vr1. When the relative velocity Vr is larger than the map switching threshold Vr1 (S41: Yes), the collision detection ECU 10 selects the relative velocity high threshold map as the collision determination threshold map in step S42. On the other hand, when the relative velocity Vr is equal to or less than the map switching threshold Vr1 (S41: No), the relative velocity small threshold map is selected as the collision determination threshold map in step S43. When the collision detection ECU 10 selects the threshold map, the collision detection ECU 10 proceeds to the process in step S15.

The relative velocity small threshold map is the collision determination threshold map shown in FIG. 8A, and has the same characteristics as the collision determination threshold map in the embodiment. On the other hand, the relative speed high threshold map is the collision determination threshold map shown in FIG. 8B, and the collision determination threshold Gfref is set to a small value even in the low vehicle speed range.

For example, in the collision determination threshold map (relative velocity small threshold map) of the embodiment, the collision determination threshold Gfref is set to a large value so as not to determine the collision between the own vehicle and the object in the low vehicle speed range. There is. That is, the vehicle speed range excluding the low vehicle speed range is set as the collision determination speed region for determining the collision between the own vehicle and the object. This is because the collision detection accuracy with the object at low speed is lower than the collision detection accuracy at high speed (it is easy to make a false judgment at low speed), and when the vehicle collides with the object at low speed. This is because the damage to the object is expected to be small.

When the relative speed Vr is large, high determination accuracy can be obtained even when traveling at low speed. Therefore, in this modification 2, when the relative speed Vr is larger than the map switching threshold value Vr1, the relative speed high use threshold value is set so that the collision judgment can be performed even in the low vehicle speed range. The map is selected.

Therefore, according to the collision determination routine of the second modification, it is possible to expand the vehicle speed range in which the collision determination can be performed while maintaining the collision determination accuracy.

<Modification example 3>
In the above embodiment, the presence or absence of a collision between the own vehicle and the object (pedestrian or bicycle) is determined based on the presence or absence of matching between the collision prediction timing and the actual detection timing, but even so, depending on the situation. May be misjudged. For example, as shown in FIG. 9, a pedestrian B (object) is detected in front of the own vehicle A by a peripheral monitoring sensor, and a rough road portion C such as a curb or a hole is next to the pedestrian B. Consider the situation formed on the road surface. In such a situation, when the driver performs a steering operation to avoid the pedestrian B, the wheels of the own vehicle A pass through the rough road portion C, and the collision determination threshold value Gfref is input from the rough road portion C to the wheels. Larger front Gf may be detected. In such a case, the collision prediction timing and the actual detection timing are at the same time even though the own vehicle A does not collide with the pedestrian B. When the three-dimensional object in front of the own vehicle A is switched immediately before the collision, the peripheral monitoring sensor cannot instantly recognize the three-dimensional object. Therefore, the collision detection ECU 10 erroneously determines that the own vehicle A has collided with the pedestrian B.

Further, for example, as shown in FIG. 10, the camera sensor 20 erroneously recognizes a non-object D such as a signboard or a trash can placed on the road as an object (pedestrian B), and the own vehicle A remains as it is. A case of collision with the non-object D is conceivable. In such a case, when the own vehicle A collides with the non-object D and a front Gf larger than the collision determination threshold value Gfref is detected, the collision prediction timing and the actual detection timing are the same. Therefore, the collision detection ECU 10 may erroneously determine that the own vehicle A has collided with the object (pedestrian B) even though the vehicle A has not collided with the object (pedestrian B).

Modification 3 also solves these problems. It is considered that the subsequent vehicle behavior (driver's operation) differs between the case where the own vehicle actually collides with an object (pedestrian or bicycle) and the case where the own vehicle collides with a non-object. For example, as shown in FIG. 11A, when the own vehicle A actually collides with the pedestrian B (object), it is considered that the driver immediately stops the own vehicle A (vehicle speed V = 0). .. On the other hand, as shown in FIG. 11B, when the own vehicle A passes through the rough road portion C, even if an impact occurs such that the front Gf becomes larger than the collision determination threshold value Gfref, the driver himself / herself It is considered that the vehicle A is not stopped and the traveling is continued as it is. Similarly, when the own vehicle collides with a non-object placed on the road, it is considered that the driver does not stop the own vehicle and continues the running as it is.

Therefore, in this modification 3, when it is determined that the front Gf becomes larger than the collision determination threshold value Gfref, the condition that the stop of the own vehicle is detected within a predetermined time from the determination is the collision determination condition. Has been added to.

FIG. 12 shows a portion to be changed in the collision determination routine of the embodiment. In the third modification, in the collision determination routine (FIG. 4) of the embodiment, the processes of steps S51 to S53 (shown by thick lines) are incorporated between steps S19 and S20, and other measures are taken. Is the same as in the embodiment. Therefore, here, a process different from the embodiment will be described.

When the collision detection ECU 10 determines in step S19 that the collision prediction timing determined by the collision prediction time TTC of the object and the actual detection timing when the front Gf exceeds the collision determination threshold value Gfref are the same time, the process is performed. Proceed to step S51. In step S51, the collision detection ECU 10 resets (zero clears) the timer value t of the timer and restarts the time counting by the timer. Therefore, the measurement of the elapsed time t after it is determined that the front Gf exceeds the collision determination threshold value Gfref is started.

Subsequently, in step S52, the collision detection ECU 10 determines whether or not the vehicle speed V detected by the vehicle speed sensor 40 is zero, that is, whether or not the own vehicle has stopped. When it is determined that the front Gf exceeds the collision determination threshold value Gfref, the own vehicle is not stopped. Therefore, the collision detection ECU 10 determines "No" in step S52, and proceeds to the process in step S53. In step S53, the collision detection ECU 10 determines whether or not the elapsed time t exceeds the threshold value tref. For example, the threshold tref is a few seconds. When the elapsed time t after the front Gf is determined to have exceeded the collision determination threshold value Gfref does not exceed the threshold value tref (S53: No), the collision detection ECU 10 returns the process to step S52 and performs the above-described process. repeat.

When the collision detection ECU 10 repeats such processing (S52, S53) and detects that the vehicle speed V becomes zero (V = 0: stop of the own vehicle) before the elapsed time t exceeds the threshold value tref, the collision detection ECU 10 detects that the vehicle speed V becomes zero. The process proceeds to step S20, and it is determined that the own vehicle has collided with the object.

On the other hand, when the elapsed time t exceeds the threshold value tref without the vehicle speed V becoming zero (S53: Yes), the collision detection ECU 10 proceeds to the process in step S22, and the object does not collide with the own vehicle. Is determined.

According to the collision determination routine of the third modification, the condition that the stop of the own vehicle is detected within a predetermined period after the front Gf exceeds the collision determination threshold value Gfref is added to the collision determination condition with the object. Has been done. Therefore, did the front Gf exceed the collision detection threshold Gfref due to a collision with an object (pedestrian or bicycle), or did the front Gf exceed the collision detection threshold Gfref due to a collision with an object other than the object or running on a rough road? It can be appropriately determined whether or not it exceeds. Thereby, the collision determination accuracy can be improved.

<Modification example 4>
In the above-described third modification, the condition that the stop of the own vehicle is detected within a predetermined period after the front Gf exceeds the collision determination threshold Gfref is added to the collision determination condition with the object. In this modification 4, the condition that the driver's specific operation is detected within a predetermined period before and after the front Gf exceeds the collision determination threshold Gfref is added to the collision determination condition with the object. This solves the problem that was tried to be solved in the modified example 3.

FIG. 13 shows a portion to be changed in the collision determination routine of the embodiment. In this modification 4, in the collision determination routine (FIG. 4) of the embodiment, the processes of steps S61 to S66 (shown by thick lines) are incorporated between steps S19 and S20, and other measures are taken. Is the same as in the embodiment. Therefore, here, a process different from the embodiment will be described.

When the collision detection ECU 10 determines in step S19 that the collision prediction timing determined by the collision prediction time TTC of the object and the actual detection timing when the front Gf exceeds the collision determination threshold value Gfref are the same time, the process is performed. Proceed to step S61. In step S61, the collision detection ECU 10 determines whether or not the horn operation (the operation of ringing the horn) has been performed immediately before. For example, has the collision detection ECU 10 performed the horn operation by the driver during the period from a predetermined time (for example, several seconds) before the current time (when it is determined that the front Gf exceeds the collision determination threshold value Gfref) to the present time? Judge whether or not. The collision detection ECU 10 inputs a horn operation signal indicating that the horn operation has been performed, and holds the horn operation signal for a predetermined time. In step S61, the collision detection ECU 10 determines whether or not the horn operation signal is held.

If the front Gf exceeds the collision determination threshold value Gfref immediately after the driver operates the horn, it can be considered that the own vehicle has collided with an object (pedestrian or bicycle). Therefore, when the horn operation is detected (S61: Yes), the collision detection ECU 10 proceeds to step S20 and determines that the own vehicle has collided with the object.

On the other hand, when the horn operation is not detected (S61: No), the collision detection ECU 10 proceeds to the process in step S62. In step S62, the collision detection ECU 10 resets (zero clears) the timer value t of the timer and restarts the time counting by the timer. Therefore, the measurement of the elapsed time t after it is determined that the front Gf exceeds the collision determination threshold value Gfref is started.

Subsequently, in step S63, the collision detection ECU 10 reads a signal of a shift position sensor (not shown) and determines whether or not the shift position has been operated to "R" (reverse position) or "P" (parking position). When it is determined that the front Gf exceeds the collision determination threshold value Gfref, such a shift change operation is not performed. Therefore, the collision detection ECU 10 determines "No" in step S63, and proceeds to the process in step S64.

In step S64, the collision detection ECU 10 reads a signal of a side brake switch (not shown) and determines whether or not a side brake operation (parking brake operation) has been performed. When it is determined that the front Gf exceeds the collision determination threshold value Gfref, such a side brake operation is not performed. Therefore, the collision detection ECU 10 determines "No" in step S64, and proceeds to the process in step S65.

In step S65, the collision detection ECU 10 reads a signal of a door open / close switch (not shown) and determines whether or not a door opening operation has been performed. When it is determined that the front Gf exceeds the collision determination threshold value Gfref, such a door opening operation is not performed. Therefore, the collision detection ECU 10 determines "No" in step S65, and proceeds to the process in step S66.

In step S66, the collision detection ECU 10 determines whether or not the elapsed time t exceeds the threshold value tref. For example, the threshold tref is a few seconds. When the elapsed time t after the front Gf is determined to have exceeded the collision determination threshold value Gfref does not exceed the threshold value tref (S53: No), the collision detection ECU 10 returns the process to step S63 and performs the above-described process. repeat.

When the own vehicle collides with an object, the driver is considered to immediately stop the own vehicle and perform at least one of the shift operation, the side brake operation, and the door opening operation described above. Therefore, the collision detection ECU 10 determines that the shift operation is detected (S63: Yes) or the side brake operation is detected (S64: Yes) before the elapsed time t exceeds the threshold value tref, or the door. When the opening operation of is detected (S65: Yes), the process proceeds to step S20, and it is determined that the own vehicle has collided with the object.

On the other hand, if the elapsed time t exceeds the threshold value tref without detecting any of the shift operation, the side brake operation, and the door opening operation (S66: Yes), it is considered that the own vehicle has not stopped. Therefore, the collision detection ECU 10 advances the process to step S22 and determines that the object and the own vehicle do not collide.

According to the collision determination routine of the modification 4, it is added to the collision determination condition with the object that the driver's specific operation is detected within a predetermined period before and after the front Gf exceeds the collision determination threshold value Gfref. ing. This specific operation is an operation for calling attention to an object (for example, a horn operation) or an operation that can be estimated to be performed when the own vehicle is stopped (for example, a shift operation, a side brake operation, a door operation). At least one of the opening operations). Therefore, did the front Gf exceed the collision detection threshold Gfref due to a collision with an object (pedestrian or bicycle), or did the front Gf exceed the collision detection threshold Gfref due to a collision with an object other than the object or running on a rough road? It can be appropriately determined whether or not it exceeds. Thereby, the collision determination accuracy can be improved.

Although the collision detection device according to the present embodiment has been described above, the present invention is not limited to the above-described embodiment and modifications, and various modifications can be made without departing from the object of the present invention.

1 ... Collision detection device, 10 ... Collision detection ECU 10, 20 ... Camera sensor, 30 ... Radar sensor, 40 ... Vehicle speed sensor, 50 ... Front occupant protection acceleration sensor, 60 ... Other occupant protection acceleration sensor, TTC ... Collision Predicted time, Gfref ... Collision judgment threshold, Gf ... Detection deceleration, V ... Vehicle speed.

Claims (4)

An object detection means that detects pedestrians and bicycles located in front of the own vehicle as objects,
An acceleration sensor that detects the deceleration of the own vehicle caused by a collision from the front,
A collision prediction time acquisition means for acquiring a collision prediction time, which is a prediction time until an object detected by the object detection means collides with the own vehicle,
Timing for determining whether or not the timing at which the deceleration detected by the acceleration sensor exceeds the collision determination threshold is the same as the collision prediction timing determined from the collision prediction time acquired by the collision prediction time acquisition means. When the determination means and the timing determination means determine that the timing at which the deceleration exceeds the collision determination threshold is the same as the collision prediction timing, the own vehicle collides with the object. A collision detection device provided with a collision determination means for determination. In the collision detection device according to claim 1,
The collision detection device has the collision determination threshold value set to a value corresponding to the vehicle speed of the own vehicle. In the collision detection device according to claim 1 or 2.
In the collision determination means, the timing determination means determines that the timing at which the deceleration exceeds the collision determination threshold is the same as the collision prediction timing, and the deceleration is the collision determination threshold. A collision detection device configured to determine that the own vehicle has collided with the object when the stop of the own vehicle is detected within a predetermined period after exceeding the above. In the collision detection device according to claim 1 or 2.
In the collision determination means, the timing determination means determines that the timing at which the deceleration exceeds the collision determination threshold is the same as the collision prediction timing, and the deceleration is the collision determination threshold. A collision detection device configured to determine that the own vehicle has collided with the object when a specific operation of the driver is detected within a predetermined period before and after exceeding the above.

Images