JP2010070047A - Collision forecasting device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a collision forecasting device for accurately forecasting a risk of collision between an object existing in front of one's own vehicle and the own vehicle. <P>SOLUTION: The collision forecasting device includes an object detection means for detecting a position of the object in front of the vehicle, an observation point setting means for setting an observation point for observing the object in an inner part of the vehicle or on an outer circumference surface of the vehicle and at more rear side than the front end of the vehicle, a detection angle calculation means for calculating a detection angle indicating a detection direction of the object viewing from the observation point, a variation amount calculation means for calculating the variation amount of the detection angle per unit time, and a collision determination means for forecasting the risk of collision between the vehicle and the object based on the variation amount value. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to a collision prediction apparatus, and more particularly to a collision prediction apparatus that predicts the risk of an object approaching a vehicle colliding with the vehicle.

  Conventionally, an apparatus (hereinafter referred to as a collision prevention apparatus) that predicts the risk of a collision between the host vehicle and an object around the vehicle and executes vehicle control for the purpose of avoiding a collision or assisting passenger protection based on the prediction result. Have been developed). In a vehicle equipped with such a collision prevention device, if the risk of collision cannot be accurately predicted by the device, unnecessary vehicle control may be executed, and the driver of the vehicle may feel uncomfortable.

A vehicle collision warning system that solves such a problem is disclosed in Patent Document 1. The vehicle collision warning system disclosed in Patent Document 1 calculates the rate of change in the detected angular position of an object using the position of a radar receiver as an observation point. When the detected object comes into contact with the own vehicle, the rate of change increases. Therefore, the vehicle collision warning system can predict whether or not the object and the vehicle collide according to whether or not the value of the change rate exceeds a threshold value.
JP 2000-504828 A

  In the vehicle collision warning system disclosed in Patent Document 1, it is assumed that the angular position where the detected object is present is calculated using the radar receiver as an observation point. Here, when the radar receiver is attached to the front end of the vehicle, the observation point is located at the front end of the vehicle.

  FIG. 16 shows how the change rate of the angular position of the detected object changes with time when the observation point of the angular position is set at the front end of the vehicle as described above and the detected object collides with or passes by the own vehicle. FIG. 16 is a graph showing the relationship between the estimated time until the vehicle collides with the detected object and the change rate of the angular position when the observation point is set at the front end of the vehicle. In the graph shown in FIG. 16, the item p indicates the rate of change in the angular position of the detected object when the vehicle and the detected object pass each other. The item fL indicates the rate of change of the angular position of the detected object when the detected object approaches from the right front of the own vehicle and collides with the front left side of the own vehicle. The item fR indicates the rate of change of the angular position when the detected object approaches from the right front of the host vehicle and collides with the front right side of the host vehicle. The item sF indicates the rate of change of the angular position when the detected object approaches from the right front side of the own vehicle and collides with the front side of the right side of the own vehicle. The item sB indicates the rate of change of the angular position when the detected object approaches from the right front side of the host vehicle and collides with the front side of the right side of the host vehicle.

  As shown in FIG. 16, the transition state of the change rate of the angular position differs depending on which position of the own vehicle the detected object collides with. The change rate of the angular position may change in the same manner regardless of whether the detected object passes the own vehicle or when the detected object collides with the own vehicle. Specifically, as shown in FIG. 16, the items p, fR, sF, and sB all change exponentially as the time until the collision decreases. Here, as shown in Patent Document 1, when the change rate of the angular position exceeds a predetermined threshold, it is determined that there is no risk of collision between the own vehicle and the detected object. In order to collide with the threshold value, it is necessary to set a threshold value smaller than the item p. However, since the item sB always keeps a larger value than the item p in spite of actually colliding with the own vehicle, it is determined that the own vehicle and the detected object do not collide. That is, in the prior art, it may not be possible to accurately predict whether or not the detected object collides with the vehicle.

  The present invention has been made in view of the above, and an object of the present invention is to provide a collision prediction apparatus capable of accurately predicting the risk of collision between an object existing ahead of the host vehicle and the host vehicle.

  The present invention employs the following configuration in order to solve the above problems. That is, the first invention is an object detection means for detecting the position of an object in front of the vehicle, and an observation for observing the object in the vehicle or on the outer peripheral surface of the vehicle and behind the front end of the vehicle. Observation point setting means for determining a point, detection angle calculation means for calculating a detection angle indicating the detection direction of the object viewed from the observation point, change amount calculation means for calculating a change amount of the detection angle per unit time, vehicle And a collision determination unit that predicts the risk of collision between the object and the object based on the value of the change amount.

  According to a second aspect of the present invention, in the first aspect of the present invention, a collision part estimation unit that estimates an outer peripheral surface of an outer periphery of a vehicle that may collide with an object as a predicted collision part and a collision part estimation unit are estimated. Threshold setting means for setting a threshold used for predicting the risk of collision between the vehicle and the object according to the predicted collision portion is further provided, and the collision determination means determines whether or not the change amount satisfies the threshold. Thus, the collision between the vehicle and the object is predicted.

  According to a third invention, in the second invention, a host vehicle travel route calculation unit that calculates a travel route of the vehicle, and a detected object travel route calculation unit that calculates a travel route of an object that moves relative to the vehicle, And the collision site estimation means estimates the predicted collision site based on the traveling path of the vehicle and the traveling path of the object.

  In a fourth aspect based on the third aspect, when the traveling path of the vehicle and the traveling path of the object are parallel to each other, the front surface of the vehicle is set as a predicted collision portion.

  According to a fifth invention, in the second invention, the vehicle further comprises collision point estimation means for estimating an intersection point between a traveling path of the vehicle and a traveling path of an object moving relative to the vehicle as a predicted collision point. The part estimation means determines which of the vehicle or the object first reaches the predicted collision point, and estimates the predicted collision part according to the determination result.

  In a sixth aspect based on the fifth aspect, when it is determined that the object reaches the predicted collision point before the vehicle, it is determined that the front surface of the vehicle is the predicted collision site and the vehicle reaches the predicted collision point before the object. In this case, the side of the vehicle is set as the predicted collision site.

  According to a seventh aspect, in the second aspect, when the vehicle and the object collide, the seventh aspect further includes a collision time calculation unit that calculates an expected collision time required until the collision, and the threshold setting unit includes the collision part estimation unit. The threshold is set in accordance with the estimated collision portion estimated in (1) and the estimated collision time calculated by the collision time calculation means.

  In an eighth aspect based on the seventh aspect, the threshold value setting means sets the threshold value to a stepwise larger value as the predicted collision time becomes shorter when the side surface of the vehicle is a predicted collision part. When the amount of change is larger than the threshold, the object and the vehicle are predicted to collide, and when the amount of change is equal to or less than the threshold, the object and the vehicle are predicted not to collide.

  In a ninth aspect based on the first aspect, the observation point setting means sets the observation point at the center of the rear end of the vehicle.

  In a tenth aspect based on the first aspect, the observation point setting means sets the observation point at the rear left corner of the vehicle.

  In an eleventh aspect based on the first aspect, the observation point setting means sets the observation point so as to be positioned at the right corner of the rear end of the vehicle.

  According to the first invention, when predicting a collision with an object approaching the vehicle, the amount of change per unit time of the detected angle when the vehicle and the object collide (hereinafter referred to as the detected angle change amount). The value of (referred to as “calling”) is a value larger than the value of the detected angle change amount when the vehicle does not collide with the object, regardless of where the object collides. Therefore, the risk of collision between the vehicle and the object can be accurately determined based on the observed change amount.

  According to the second invention, the threshold for determining the risk of collision can be changed according to which part of the vehicle the object collides with. Therefore, the risk of collision between the vehicle and the object can be accurately determined based on the appropriately set threshold value.

  According to the third aspect, it is possible to estimate a part of a vehicle where an object may collide with a simple process.

  According to the fourth invention, it is possible to estimate that an object may collide with the front of the vehicle by a simple process.

  According to the fifth aspect of the present invention, it is possible to estimate a vehicle part where an object may collide with a simple process.

  According to the sixth aspect of the present invention, it is possible to estimate whether the part of the vehicle where the object may collide with the simple process is the front surface of the vehicle or the side surface of the vehicle.

  According to the seventh aspect, it is possible to set an optimum threshold value according to the object and the situation of the vehicle that change every moment.

  According to the eighth invention, for example, when the object collides with the rear side of the vehicle, such as when the object approaches the vehicle on a trajectory where the amount of change in the detected angle per unit time tends to increase. Since an appropriate threshold is set, a collision between the object and the vehicle can be accurately predicted.

  According to the ninth aspect, the risk of collision with an object approaching from the front of the vehicle can be determined more accurately.

  According to the tenth aspect, for example, the risk of a collision with an object approaching from the right front of the vehicle can be determined more accurately. Specifically, when the observation point is set at the right corner of the rear end of the vehicle, the object approaching from the right side of the vehicle and the vehicle are closer than when the observation point is set at another position of the vehicle. The difference between the value of the observation angle change amount when the vehicle collides and the value of the observation angle change amount when the object and the vehicle do not collide become larger. Therefore, the risk of collision with an object approaching from the right side of the vehicle can be more accurately determined based on the value of the observation angle change amount.

  According to the eleventh aspect, for example, the risk of collision with an object approaching from the left front of the vehicle can be determined more accurately. Specifically, when the observation point is set at the rear left corner of the vehicle, the object approaching from the pleat side of the vehicle and the vehicle are less than when the observation point is set at another position of the vehicle. The difference between the value of the observation angle change amount when the vehicle collides and the value of the observation angle change amount when the object and the vehicle do not collide become larger. Therefore, it is possible to more accurately determine the risk of collision with an object approaching from the left side of the vehicle based on the value of the observation angle change amount.

(First embodiment)
Hereinafter, a collision prediction apparatus according to a first embodiment of the present invention will be described with reference to FIGS. In the embodiment described below, an example in which the collision prediction device 13 is mounted on the vehicle 1 will be described.

  First, the functional configuration of the collision prediction apparatus will be described with reference to FIG. FIG. 1 is a block diagram illustrating a functional configuration of the collision prediction apparatus. As shown in FIG. 1, the collision prediction device 13 is mounted on the vehicle 1. The vehicle 1 includes a radar device 10, a vehicle speedometer 11, a vehicle position detection device 12, and an alarm device 14 in addition to the collision prediction device 13. The collision prediction device 13 is electrically connected to the radar device 10, the vehicle speedometer 11, the vehicle position detection device 12, and the alarm device 14. Note that the radar device 10, the vehicle speedometer 11, the vehicle position detection device 12, the collision prediction device 13, and the warning device 14 all operate when the IG power supply of the vehicle 1 is on.

  The radar device 10 is typically a radar device that radiates electromagnetic waves, receives a reflected wave from an object existing around the vehicle, and detects the position of the object. The radar device 10 is disposed inside the front grill of the vehicle 1 and detects an object existing in front of the vehicle 1. The radar device 10 measures the distance Lr from the vehicle 1 to the detected object, the horizontal direction of the detected object viewed from the radar device 10, and the relative velocity Vr between the detected object and the vehicle 1. The radar apparatus 10 outputs data indicating the distance from the vehicle 1 to the detected object, the detection direction of the detected object viewed from the radar apparatus 10, and the relative speed Vr of the detected object with the vehicle 1 to the collision prediction apparatus 13. .

  The vehicle speedometer 11 is a device that measures the traveling speed of a vehicle generally mounted on the vehicle. The vehicle speedometer 11 measures the traveling speed of the vehicle 1 (hereinafter referred to as the host vehicle speed Vs) and outputs the measured value to the collision prediction device 13.

  The own vehicle position detection device 12 is typically a device having a GPS (Global Positioning System) such as a generally known car navigation system. The own vehicle position detection device 12 stores map information in advance, and based on a signal transmitted from a satellite or the like, absolute position coordinates of the vehicle 1 on the map (hereinafter referred to as own vehicle position Ps). taking measurement. Strictly speaking, the own vehicle position Ps is a point indicating the position where the own vehicle position detection device 12 is mounted in the vehicle 1. The own vehicle position detection device 12 calculates the traveling direction D of the vehicle 1 based on the change direction of the own vehicle position Ps and the like. The vehicle position detection device 12 outputs data indicating the measured vehicle position Ps, the traveling direction D, and map information around the vehicle position Ps (hereinafter referred to as a surrounding map) to the collision prediction device 13. To do.

  The collision prediction device 13 is typically a processing device including an information processing device such as a microcomputer, a storage device such as a memory, and an interface circuit. Although detailed description of the process will be described later, the collision prediction device 13 predicts a collision between the vehicle 1 and an object ahead based on the data output from the radar device 10. Here, the collision prediction device 13 observes a change in the position of the object with the rear end of the vehicle 1 as a reference point, and predicts the danger of a collision between the vehicle 1 and the object ahead based on the observation result. When the collision prediction device 13 predicts that there is a high risk of collision between the vehicle and a surrounding object, the collision prediction device 13 outputs an instruction signal for issuing an alarm to the alarm device 14. By determining the observation reference point at the rear end of the vehicle 1 as described above, it is possible to predict the risk of collision between the vehicle 1 and the object with higher accuracy than in the past.

  The alarm device 14 is typically an audio output device such as a buzzer. In response to an instruction from the collision prediction device 13, the warning device 14 emits a warning sound when there is a high risk of collision between the vehicle 1 and a surrounding object.

  Next, details of the process executed by the collision prediction apparatus 13 will be described with reference to FIG. FIG. 2 is an example of a flowchart illustrating processing executed by the collision prediction device 13. The collision prediction apparatus 13 starts the process shown in FIG. 2 when the IG power supply of the vehicle is turned on.

  In step S1, the collision prediction apparatus 13 determines whether or not the IG power supply is in an on state. When the IG power supply is on, the collision prediction device 13 advances the process to step S2. On the other hand, the collision prediction device 13 ends the process when the IG power supply is in a state other than on.

  By the process of step S1, the collision prediction apparatus 13 repeatedly loops and executes the processes from step S1 to step S14 described below while the IG power supply of the vehicle 1 is on.

  In step S2, the collision prediction device 13 detects the host vehicle position Ps and the traveling direction D. Specifically, the collision prediction device 13 acquires data indicating the vehicle position Ps, the traveling direction D, and the surrounding map output from the vehicle position detection device 12, and stores the data in the storage device. When the process of step S2 is completed, the collision prediction apparatus 13 advances the process to step S3.

  By repeatedly executing the process of step S2 in a loop process, the storage device of the collision prediction device 13 stores the values of the vehicle position Ps and the traveling direction D at each time.

  In step S3, the collision prediction device 13 selects one of the detected objects. Hereinafter, the detection object selected in step S3 is referred to as a selection detection object. When the process of step S3 is completed, the collision prediction apparatus 13 advances the process to step S4.

  By the process of step S3, the collision prediction apparatus 13 executes the processes from step S4 to step S9 described below for each detected object detected by the radar apparatus 10.

  In step S4, the collision prediction device 13 maps the detected object position Pt. The detected object position Pt is an absolute position coordinate on the surrounding map of the detected object detected by the radar device 10. Specifically, first, the collision prediction device 13 obtains the distance to the detected object and the detection direction since it is output by the radar device 10. Next, the collision prediction device 13 calculates the absolute position of the selected detection object on the surrounding map. More specifically, the collision prediction device 13 uses the vehicle position Ps and the traveling direction D indicated by the absolute coordinates on the surrounding map as a reference, and detects the selected detection object from the vehicle position Ps. The detected object position Pt on the surrounding map is calculated based on the detected distance, the detection direction of the selected detected object viewed from the radar device 10, and the detected distance, and stored in the storage device. It is assumed that the distance from the vehicle position Ps to the radar device 10 that has detected the selection detection object is known, and the collision prediction device 13 stores the distance in advance. In addition, as a method of calculating the absolute coordinates of a point whose relative positional relationship is known with reference to the own vehicle position Ps, any conventionally known method may be used. When the process of step S4 is completed, the collision prediction apparatus 13 advances the process to step S5.

  By repeatedly executing the process of step S4 in a loop process, the value of the detected object position Pt at each time is stored in the storage device of the collision prediction device 13.

  In step S5, the collision prediction device 13 calculates and stores the detection angle θ. Hereinafter, the definition of the detection angle θ will be described with reference to FIG. FIG. 3 is a plan view showing the position of the vehicle 1, the position of the vehicle 2 that is the selection detection object, and the detection angle of the vehicle 2 in the surrounding map coordinate system. As shown in FIG. 3, in the present embodiment, the process of the collision prediction device 13 will be described by taking as an example the case where the vehicle 2 is selected as the selection detection object.

  First, the collision prediction device 13 calculates the absolute coordinates of the observation point Po on the surrounding map based on the own vehicle position Ps. The observation point Po is a point for determining the detection angle θ and is set at the center of the rear end of the vehicle 1. The collision prediction device 13 stores in advance a relative positional relationship between the mounting position of the vehicle position detection device 12 and the rear end of the vehicle 1 in a storage device. Therefore, the collision prediction device 13 can calculate the absolute coordinates of the observation point Po on the surrounding map based on the absolute coordinates of the vehicle position Ps and the relative positional relationship.

  Next, the collision prediction device 13 has a straight line connecting the detected object position Pt and the observation point Po on the surrounding map (hereinafter referred to as a detection axis) and a straight line passing through the vehicle 1 in the front-rear direction through the observation point Po ( Hereinafter, the angle formed by the axle line is calculated as the detected angle θ. For the sake of explanation, the value of the detection angle θ is assumed to be θ = 0 when the detection axis coincides with the axle, and the traveling direction of the vehicle 1 from the axle is centered on the observation point Po. It is assumed that the value increases as it turns to the right. The collision prediction device 13 stores the calculated value of the detected angle θ in the storage device. When the process of step S5 is completed, the collision prediction apparatus 13 advances the process to step S6.

  By repeatedly executing the process of step S5, the value of the detected angle θ at each time is stored in the storage device of the collision prediction device 13. Each time the collision prediction device 13 newly calculates the value of the detection angle θ, the data indicating the value of the detection angle θ is determined by overwriting the oldest data of the detection angle θ stored at that time. It is assumed that only the number is always stored.

In step S6, the collision prediction apparatus 13 calculates the detected angle change amount θv. The detected angle change amount θv is a numerical value indicating the change amount per unit time of the detected angle θ. Specifically, first, the collision prediction device 13 reads the detection angle value θ1 that is most recently stored in the storage device and the detection angle value θ2 that is newly stored secondly from the storage device. The collision prediction device 13 calculates the detected angle change amount θv by the following equation (1) based on the values of the detected angle θ1 and the detected angle θ2. Note that Δt in Equation (1) indicates the time required from when the detected angle θ2 is stored until the detected angle θ2 is stored. As the value of Δt, the processing time required until the process of step S6 is looped and executed again may be obtained experimentally, and the time may be stored in advance in the collision prediction device 13 as the value of Δt. The collision prediction device 13 may be provided with a timer function, and measurement may be performed using the function.
θv = (θ2−θ1) / Δt (1)
When the collision prediction apparatus 13 completes the process of step S6, the collision prediction apparatus 13 proceeds with the process to step S7.

  In step S7, the collision prediction device 13 executes a subroutine process of a threshold setting process. The threshold value setting process is a process for setting a threshold value θvth of the detected angle change amount. Although details will be described in the process of step S8 described later, the collision prediction device 13 predicts whether or not the selected detection object and the vehicle collide according to whether or not the value of the detected angle change amount θv exceeds the threshold value θvth. To do. Hereinafter, the threshold setting process will be described with reference to FIG. FIG. 4 is an example of a flowchart showing threshold setting processing.

In step S70, the collision prediction device 13 calculates a collision prediction time TTC. The predicted collision time TTC is a time estimated to be required until the vehicle 1 and the selected detection object collide. The collision prediction device 13 first obtains the distance Lr from the vehicle 1 to the detected object and the relative speed Vr between the detected object and the vehicle 1 from the radar device 10. Then, the collision prediction device 13 calculates the value of the predicted collision time TTC based on the following equation (2).
TTC = Lr / Vr (2)
When the process of step S70 is completed, the collision prediction apparatus 13 advances the process to step S71.

  In step S71, the collision prediction device 13 calculates the advance angle α. The travel angle α is an angle formed by a straight line indicating the travel path of the vehicle 1 and a straight line indicating the travel path of the selected detection object. Hereinafter, a straight line indicating the travel path of the vehicle 1 is referred to as a host vehicle travel path line, and a straight line indicating the travel path of the selected detected object is referred to as a detected object travel path line.

  Specifically, first, the collision prediction device 13 calculates the own vehicle travel route line. For example, in the coordinate system of the surrounding map, the collision prediction device 13 is a straight line connecting the latest vehicle position Ps stored in advance in the process of step S2 and the vehicle position Ps stored immediately before the latest vehicle position Ps. Is calculated as the own vehicle travel route line. Next, the collision prediction device 13 calculates a detected object travel route line. For example, in the coordinate system of the surrounding map, the collision prediction device 13 connects the latest detected object position Pt stored in advance in the process of step S4 and the detected object position Pt stored immediately before the latest detected object position Pt. Is calculated as a detection object travel path line. In addition, the calculation method of said own vehicle advancing route line and detected object advancing route line is an example, and the method of calculating the advancing route of the vehicle 1 may use the conventionally well-known arbitrary methods. The collision prediction device 13 calculates the angle formed by the calculated own vehicle travel route line and the detected object travel route line as a travel angle α based on a linear expression representing each line.

  FIG. 5 shows the own vehicle travel route line Bs and the detected object travel route line Bt calculated by the above processing. FIG. 5 is a plan view showing the vehicle 1, the vehicle 2, the own vehicle travel route line Bs, and the detected object travel route line Bt in the surrounding map coordinate system. As shown in FIG. 5, when the own vehicle travel route line Bs and the detected object travel route line Bt are not parallel, the travel route lines intersect each other. The intersection of the own vehicle travel route line Bs and the detected object travel route line Bt is referred to as a predicted collision point Pc. The value of the traveling angle α is α = 0 when the own vehicle traveling route line Bs and the detected object traveling route line Bt are parallel, and the detected object traveling route line Bt is centered on the predicted collision point Pc. The value increases as the vehicle turns to the right from Bs in the traveling direction of the vehicle 1. When the detected object is a stationary object, the value of the advance angle α is calculated as 0. When the process of step S71 is completed, the collision prediction apparatus 13 advances the process to step S72.

  In step S72, the collision prediction device 13 determines whether or not the own vehicle travel route line Bs and the detected object travel route line Bt are parallel. Specifically, the collision prediction device 13 determines whether or not the travel angle α = 0 calculated in step S71. When the advance angle α = 0, the collision prediction device 13 determines that the host vehicle travel route line Bs and the detected object travel route line Bt are parallel, and the process proceeds to step S77. On the other hand, when the traveling angle α is not 0, the collision prediction device 13 determines that the own vehicle traveling route line Bs and the detected object traveling route line Bt are not parallel, and the collision predicting device 13 advances the process to step S73.

  According to the process of step S72, if it is determined that the host vehicle travel route line Bs and the detected object travel route line Bt are parallel, the vehicle 1 and the selected detected object are selected to collide with each other. It can be estimated that the detected object may collide with the front surface of the vehicle 1. In addition, according to the processing in step S72, the detection object that is relatively close to the vehicle 1 may collide with the front surface of the vehicle 1 regardless of whether the detection object is moving or stationary. You can estimate if there is.

  In step S73, the collision prediction device 13 calculates a predicted collision point Pc. Specifically, the collision prediction device 13 calculates, as the coordinates of the predicted collision point Pc, the coordinates of the intersection of the own vehicle travel route line Bs calculated in step S71 and the detected object travel route line Bt in the surrounding map coordinate system. To do. When the process of step S73 is completed, the collision prediction apparatus 13 advances the process to step S74.

In step S74, the collision prediction device 13 calculates the own vehicle collision point arrival time TAs. The own vehicle collision point arrival time TAs is the time required for the vehicle 1 to reach the predicted collision point Pc. Specifically, first, the collision prediction device 13 acquires the host vehicle speed Vs output from the vehicle speedometer 11. Next, a distance Ls (see FIG. 5) from the predicted collision point Pc to the host vehicle position Ps is calculated in the surrounding map coordinate system. Then, the collision prediction device 13 calculates the own vehicle collision point arrival time TAs based on the following equation (3).
TAs = Ls / Vs (3)
When the process of step S74 is completed, the collision prediction apparatus 13 advances the process to step S75.

In step S75, the collision prediction apparatus 13 calculates the detected object collision point arrival time TAt. The detected object collision point arrival time TAt is the time required for the selected detected object to arrive at the predicted collision point Pc. Specifically, first, the collision prediction device 13 calculates the detected object speed Vt. For example, the collision prediction device 13 calculates the detected object speed Vt by differentiating the time-series detected object position Pt stored in the process of step S4. Next, a distance Lt (see FIG. 5) from the predicted collision point Pc to the detected object position Pt in the surrounding map coordinate system is calculated. Then, the collision prediction device 13 calculates the detected object collision point arrival time TAt based on the following equation (4).
TAt = Lt / Vt (4)
When the process of step S75 is completed, the collision prediction apparatus 13 advances the process to step S76.

In step S76, the collision prediction device 13 determines whether or not the vehicle 1 arrives at the predicted collision point Pc from the detected object. Specifically, the collision prediction device 13 determines whether or not the value of the difference ΔTA between the own vehicle collision point arrival time TAs calculated by the following equation (5) and the detected object collision point arrival time TAt is smaller than zero. To do.
ΔTA = TAs−TAt (5)
If the collision prediction apparatus 13 determines that the value of ΔTA is smaller than 0, the collision prediction apparatus 13 proceeds with the process to step S77. If the collision prediction apparatus 13 determines that the value of ΔTA is equal to or greater than 0, the collision prediction apparatus 13 proceeds with the process to step S78.

  According to the processing from step S73 to step S76, it can be determined which of the vehicle 1 and the selected detection object first reaches the predicted collision point Pc. Here, when the vehicle 1 reaches the predicted collision point Pc first, it is considered that the own vehicle position Ps has passed the predicted collision point Pc when the detected object position Pt reaches the predicted collision point Pc. If the selected detection object collides with the vehicle 1, it is estimated that the selection detection object collides with the side surface of the vehicle 1. On the other hand, when the selected detected object first arrives at the predicted collision point Pc, it is considered that the detected object position Pt has passed the predicted collision point Pc when the vehicle position Ps reaches the predicted collision point Pc. If the vehicle 1 collides with the selection detection object, it is estimated that the vehicle 1 collides with the selection detection object in the front.

  In step S77, the collision prediction device 13 reads the front collision threshold table. The front collision threshold table is a threshold table that is referred to by the collision prediction device 13 when it is estimated that the selected detection object may collide with the front surface of the vehicle 1. The threshold table is a data table showing the relationship between the detected angle change amount threshold θvth and the expected collision time TTC.

  FIG. 6 is a diagram illustrating an example of a front collision threshold table. As shown in FIG. 6, the threshold value table is a data table in which the predicted collision time TTC and the detected angle change amount threshold value θvth are arranged for each column. In the threshold table, the value of the predicted collision time TTC is shown in the first column, and the value of the detected angle change threshold θvth corresponding to each predicted collision time TTC is shown in the second column. As shown in FIG. 6, in the front collision threshold table, the value of the threshold θvth is set to 0.01 regardless of the value of the corresponding expected collision time TTC. Although details will be described later in step S79, the collision prediction device 13 refers to the threshold value table and identifies the value of the detected angle change amount threshold value θvth corresponding to the value of the predicted collision time TTC. The collision prediction device 13 stores a plurality of threshold tables in advance, and switches the threshold table to be read according to which surface of the vehicle 1 the selected detection object collides with. In the present embodiment, the collision prediction apparatus 13 stores in advance two types of threshold tables, a front collision threshold table and a side collision threshold table. In step S77, the collision prediction device 13 reads the front collision threshold table stored in the storage device in advance in the storage device. When the process of step S77 is completed, the collision prediction apparatus 13 advances the process to step S79.

  On the other hand, in step S78, the collision prediction apparatus 13 reads the side collision threshold table. The side collision threshold table is a threshold table that the collision prediction device 13 refers to when it is estimated that the selected detection object may collide with the front surface of the vehicle 1. FIG. 7 is a diagram illustrating an example of a side collision threshold table. As shown in FIG. 6 and FIG. 7, the threshold value θvth of the detected angle change amount corresponding to the expected collision time TTC is different between the front collision threshold table and the side collision threshold table. As shown in FIG. 7, in the side collision threshold table, the corresponding threshold value θvth increases stepwise as the predicted collision time TTC value decreases. Specifically, in the side collision threshold table, when the value of the predicted collision time TTC is 1.03 or more, the corresponding threshold value θvth is set to 0.01, and the value of the predicted collision time TTC is 0.58. When the value is equal to or greater than 1.02 and equal to or less than 1.02, the corresponding threshold value θvth is set to 0.02. When the value of the predicted collision time TTC is less than 0.57, the corresponding threshold value θvth is 0.0381. Set to When the process of step S78 is completed, the collision prediction apparatus 13 advances the process to step S79.

  In step S79, the collision prediction device 13 sets a threshold value θvth of the detected angle change amount based on the threshold value table and the predicted collision time TTC. Specifically, the collision prediction device 13 searches for the value of the predicted collision time TTC calculated in the process of step S70 in the first column in any threshold table read in step S77 or step S78. Then, the value written in the second column of the same row as the row where the expected collision time TTC is written is set as the threshold value θvth of the detected angle change amount. When the collision prediction apparatus 13 completes the process of step S79, it completes the threshold setting process and advances the process to the process of step S8 in FIG.

  In step S8, the collision prediction apparatus 13 determines whether or not the value of the detection angle θv is larger than the threshold value θvth. If the collision prediction device 13 determines that the value of the detection angle θv is greater than the threshold value θvth, the process proceeds to step S9. On the other hand, when the collision prediction device 13 determines that the value of the detection angle θv is equal to or less than the threshold value θvth, the process proceeds to step S10.

  In step S9, the collision prediction device 13 turns on the collision flag. The collision flag is a flag indicating whether or not the selected detection object is predicted to collide with the vehicle 1 when the flag is on. The collision prediction device 13 sets a collision flag for each detected object detected by the radar device 10, and holds the state of the flag in the storage device. In step S9, the collision prediction device 13 sets the collision flag of the selected detection object to the on state, and stores the state. When the collision prediction apparatus 13 completes the process of step S9, the collision prediction apparatus 13 advances the process to step S11.

  In step S10, the collision prediction device 13 turns off the collision flag. The collision prediction device 13 sets the collision flag of the selected detection object to the off state and stores the state. When the collision prediction apparatus 13 completes the process of step S10, the process proceeds to step S11.

  In step S11, the collision prediction apparatus 13 determines whether all the detected objects have been selected. Specifically, the collision prediction device 13 determines whether or not the processing from step S3 to step S10 has been executed for all detected objects detected by the radar device 10. If the collision prediction device 13 determines that all the detected objects have not been selected, the process returns to step S3, and the processes from step S3 to step S10 are performed on the unselected detected objects. On the other hand, if the collision prediction apparatus 13 determines that all detected objects have been selected, the process proceeds to step S12.

  In step S12, the collision prediction apparatus 13 determines whether the collision flag of any detected object is on. Specifically, the collision prediction device 13 reads the state of the collision flag of each detected object stored in the storage device, and determines whether or not the collision flag of any detected object is on. When the collision flag of any detected object is in the on state, the collision predicting apparatus 13 advances the process to step S13 and performs a process of issuing an alarm. On the other hand, when the collision flags of all the detected objects are in the off state, the collision prediction device 13 proceeds to the process at step S14 and performs a process that does not issue an alarm.

  In step S13, the collision prediction apparatus 13 issues an alarm. Specifically, the collision prediction device 13 continuously outputs an instruction signal (hereinafter referred to as a warning signal) for issuing a warning to the warning device 14. The alarm device 14 outputs an alarm sound according to the instruction signal from the collision prediction device 13. The collision prediction apparatus 13 returns the process to step S1 when the process of step S13 is completed.

  On the other hand, in step S14, the collision prediction device 13 does not issue a warning. Specifically, the collision prediction device 13 sets a state in which an alarm signal is not output to the alarm device 14. When the warning signal is continuously output to the warning device 14, the collision prediction device 13 stops the output of the warning signal and maintains the stopped state. The collision prediction apparatus 13 returns the process to step S1 when the process of step S14 is completed.

  According to the processing from step S12 to step S14, when even one detected object detected by the radar device 10 is predicted to collide with the vehicle 1, an alarm sound is emitted, so that the driver detects the detection. It is possible to detect a collision with an object and perform an operation to avoid the collision.

  FIG. 8 shows how the detected angle change amount θv of the vehicle 2 calculated based on the processing of the collision prediction device 13 described above changes when the vehicle 2 approaches or collides with the vehicle 1. This will be described with reference to FIG.

  First, with reference to FIGS. 8 to 11, the transition of the detected angle change amount θv when the vehicle 2 passes the vehicle 1 and when the vehicle 2 collides with the front surface of the vehicle 1 will be described. FIG. 8 is a plan view showing a track W_p where the vehicle 2 traveling opposite the vehicle 1 passes the right side of the vehicle 1. FIG. 9 is a plan view showing a track W_fL in which the vehicle 2 approaches from the diagonally right front of the vehicle 1 and collides with the front left side of the vehicle 1. FIG. 10 is a plan view showing a track W_fR in which the vehicle 2 approaches from the diagonally right front side of the vehicle 1 and collides with the front right side of the vehicle 1. FIG. 11 shows changes in the detected angle change amount θv of the vehicle 2 in each of the tracks indicated by the tracks W_p, W_fL, and the track W_fR. FIG. 11 is a graph showing the relationship between the detected angle change amount θv and the expected collision time TTC when the vehicle 2 approaches the vehicle 1 in each of the tracks W_p, W_fL, and W_fR.

  As described in the processing from step S72 to step S75 of the collision prediction apparatus 13, when the traveling path of the detected object and the traveling path of the vehicle 1 are parallel, and when the detected object collides with the front surface of the vehicle 1, The value of the threshold θvth is determined based on the front collision threshold table. That is, when the vehicle 2 approaches or collides with the vehicle 1 in the above-described trajectory W_p, trajectory W_fL, or trajectory W_fR, the collision prediction device 13 determines the value of the threshold θvth based on the front collision threshold table in any case. To decide. The graph shown in FIG. 11 further shows the value of the threshold θvth calculated by the collision prediction device 13. As shown in FIG. 11, the value of the threshold θvth determined based on the front collision threshold table is 0.1 regardless of the value of the predicted collision time TTC.

  According to the graph shown in FIG. 11, when the vehicle 1 and the vehicle 2 pass each other without colliding as in the track W_p, the change amount θv of the detected angle increases exponentially as the predicted collision time TTC decreases, and is always a threshold value. It becomes a value larger than θvth. Therefore, the collision prediction device 13 predicts that the vehicle 1 and the vehicle 2 do not collide when the vehicle 2 passes without colliding with the vehicle 1 like the track W_p, and sets the collision flag of the vehicle 2 to OFF. Can do. On the other hand, when the vehicle 2 collides with the front surface of the vehicle 1 like the track W_fL or the track W_fR, the change amount θv of the detected angle decreases exponentially as the predicted collision time TTC decreases in any track. And always below the threshold value θvth. Therefore, the collision prediction device 13 can set the collision flag of the vehicle 2 to ON when the vehicle 1 and the vehicle 2 collide like the track W_fL or the track W_fR.

  Next, transition of the detected angle change amount θv when the vehicle 2 collides with the side surface of the vehicle 1 will be described with reference to FIGS. FIG. 12 is a plan view showing a track W_sF in which the vehicle 2 approaches from the diagonally right front side of the vehicle 1 and collides with the front side of the right side surface of the vehicle 1. FIG. 13 is a plan view showing a track W_sB in which the vehicle 2 approaches from the diagonally right front side of the vehicle 1 and collides with the rear side of the right side surface of the vehicle 1. FIG. 14 shows the transition of the detected angle change amount θv of the vehicle 2 in each track indicated by the track W_sF and the track W_sB. FIG. 14 is a graph showing the relationship between the detected angle change amount θv and the expected collision time TTC in the trajectory W_sF and the trajectory W_sB.

  As described in the processing from step S72 to step S75 of the collision prediction device 13, when the detected object collides with the side surface of the vehicle 1, the value of the threshold value θvth is determined based on the side surface collision threshold value table. That is, when the vehicle 2 collides with the vehicle 1 on the track W_sF or the track W_sB, the collision prediction device 13 determines the value of the threshold θvth based on the side collision threshold table. The graph shown in FIG. 14 further shows the value of the threshold θvth determined based on the side collision threshold table. As shown in FIG. 14, the value of the threshold θvth determined based on the side collision threshold table changes in a stepwise manner as the value of the predicted collision time TTC decreases.

  According to the graph shown in FIG. 14, when the vehicle 2 collides with the side surface of the vehicle 1 like the track W_sB, the value of the detected angle change amount θv increases as the predicted collision time TTC decreases, and the change amount θv. The value of may exceed 0.1. That is, if the threshold value θvth is set based on the front collision threshold table, the collision prediction device 13 predicts that the collision will not occur even though the vehicle 2 and the vehicle 1 collide. However, as described above, when the vehicle 2 collides with the side surface of the vehicle 1, the threshold value θvth is set based on the side surface collision threshold value table. Therefore, the value of the threshold value θvth increases as the predicted collision time TTC decreases. Therefore, even when the vehicle 2 collides with the side surface of the vehicle 1 on the track W_sB, the change amount θv of the detection angle is always below the threshold θvth. As shown in FIG. 14, similarly, when the vehicle 2 collides with the side surface of the vehicle 1 on the track W_sF, the change amount θv of the detected angle is always below the threshold value θvth. As described above, the collision prediction device 13 can set the collision flag of the vehicle 2 to ON when the vehicle 2 collides with the side surface of the vehicle 1 like the track W_sF or the track W_sB.

  As described above, according to the collision prediction apparatus 13 according to the present embodiment, since the observation point Po is set at the rear end of the vehicle, the case where the vehicle and the detected object pass each other and the vehicle and the detected object are different. It is calculated that the transition state of the change amount θv of the detection angle differs depending on the collision. Therefore, according to the collision prediction device 13, it is possible to accurately predict whether or not the vehicle and the detected object collide based on the change amount θv and the threshold value θvth of the detection angle. Further, as described above, since the threshold table is switched according to which position of the vehicle 1 the detected object collides with, the collision prediction device 13 is used when the vehicle 2 collides with either the front surface or the side surface of the vehicle 1. Even if it exists, it can be determined more accurately whether the detected object collides with the vehicle 1 or not.

  In the above embodiment, the collision prediction device 13 has described an example in which the position of the observation point Po is the center of the rear end of the vehicle. However, the position of the observation point Po is not limited to the center of the rear end of the vehicle. It only has to be set on the rear side.

  For example, the collision prediction device 13 may set the position of the observation point Po at the left corner of the rear end of the vehicle as shown in FIG. FIG. 15 is a plan view showing the vehicle 1 in which the observation point Po is set at the rear left corner. When the position of the observation point Po is set to the left corner of the rear end of the vehicle, the detected object approaching from the right side of the vehicle 1 and the vehicle 1 pass each other compared to the case where the position of the observation point Po is the center of the rear end of the vehicle ( The difference between the change amount θv of the detection angle of the track W_p and the change amount θv of the detection angle when the detection object approaching from the right side of the vehicle 1 collides with the vehicle (the track W_fL, W_fR, and W_sF). Becomes larger. Therefore, when the position of the observation point Po is set at the rear left corner of the vehicle, the collision prediction apparatus 13 is more accurate in an environment where detected objects frequently appear on the right side of the vehicle 1, for example, on a left-handed road. A collision between the vehicle and the detected object can be predicted.

  Further, the collision prediction device 13 may set the position of the observation point Po at the right end corner of the vehicle. When the position of the observation point Po is set to the right corner of the rear end of the vehicle, the detected object approaching from the left side of the vehicle 1 and the vehicle 1 pass each other compared to the case where the position of the observation point Po is the center of the rear end of the vehicle. The change amount θv of the detection angle of the (trajectory W_p) and the change amount θv of the detection angle when the detection object and the vehicle approaching from the right side of the vehicle 1 collide (the trajectories W_fL, W_fR, and W_sF). The difference is greater. Therefore, when the position of the observation point Po is set at the right corner of the rear end of the vehicle, the collision prediction apparatus 13 is configured to use the observation point in an environment where detected objects frequently appear on the left side of the vehicle 1, for example, on a right-hand traffic road. Similar to the case where the position of Po is set to the left corner of the rear end of the vehicle, the collision between the vehicle and the detected object can be predicted with higher accuracy.

  As described above, the position of the observation point Po is preferably set to an appropriate position such as the rear end of the vehicle 1 in accordance with the traveling environment of the vehicle, but at least the position of the observation point Po is behind the vehicle from the front end of the vehicle. If it is set to the side, the value of the detected angle change amount θv is smaller than that when the position of the observation point Po is set at the front end of the vehicle. Can be predicted.

  In the above embodiment, the example in which the collision prediction device 13 switches the threshold table according to which surface of the vehicle 1 the detected object collides with has been described. However, the detected object and the vehicle 1 can be switched without switching the threshold table. Can be prepared, a threshold value table that can set a threshold value θvth that can be accurately discriminated whether or not to collide with each other can be prepared. That is, the collision prediction apparatus 13 may always set the threshold value θvth based on one threshold value table.

  For example, the processing from step S72 to step S73 may be omitted, and the threshold value θvth may always be set based on the side collision threshold table regardless of which surface of the vehicle 1 the detected object collides with. FIG. 14 also shows the value of the detected angle change amount θv when the vehicle 2 passes by the track W_p without colliding with the vehicle 1. As shown in FIG. 14, even if the vehicle 1 and the vehicle 2 pass each other without colliding like the track W_p, even if the value of the threshold θvth is determined based on the side collision threshold table, the detection angle of the vehicle 2 The change amount θv is always larger than the threshold value θvth. Even if the value of the threshold θvth is determined based on the side collision threshold table when the vehicle 2 collides with the front surface of the vehicle 1 as in the track W_fL and the track W_fR, the change amount θv of the detection angle of the vehicle 2 is determined. Is always below the threshold value θvth. In this way, if the threshold value θvth is always set based on the side collision threshold table, the threshold value θvth that can accurately determine whether the detected object and the vehicle 1 collide without switching the threshold table is set. Can do.

  In the above embodiment, the value of the threshold θvth is set to be a constant value regardless of the value of the predicted collision time TTC in the front collision threshold table, and the value of the threshold θvth is the value of the predicted collision time TTC in the side collision threshold table. However, the setting of the threshold value θvth in each threshold value table is not limited to the above, and the optimal threshold value θvth corresponding to the value of each predicted collision time TTC has been described. The value may be obtained experimentally and set.

  Moreover, although the collision prediction apparatus 13 explained the example which maps the position of the vehicle 1 and a detected object based on the surrounding map information obtained from the own vehicle position detection apparatus 12 in the said embodiment, the collision prediction apparatus 13 is the following. A virtual map (hereinafter referred to as a virtual map) is created without using actual map information such as a surrounding map, and the positions of the vehicle 1 and the detected object are mapped on the map. It doesn't matter. That is, the vehicle 1 does not necessarily need to include the own vehicle position detection device 12. When performing the mapping process on such a virtual map, the collision prediction device 13 in step S4, the movement information such as the traveling speed of the vehicle 1, the distance from the vehicle position to the radar device 10 that has detected the selected detection object, Based on the detection direction and the detection distance of the selected detection object viewed from the radar device 10, absolute position information of the selection detection object on the virtual map is calculated.

  Moreover, although the collision prediction apparatus 13 calculated | required the own vehicle advancing route line Bs based on the displacement of the own vehicle position Ps in the said embodiment, the collision prediction apparatus 13 not only detected object position Pt, The detected object travel route line may be calculated in consideration of road information included in the surrounding map information. For example, when the vehicle 1 is traveling on a curved road, the collision prediction device 13 estimates that the vehicle 1 travels along a route along the curve, and calculates the own vehicle travel route line Bs along the curve of the road. You may do it. By calculating the own vehicle travel route line Bs in consideration of the road shape in this way, the predicted collision point Pc can be estimated more accurately, and the collision between the vehicle 1 and the detected object can be estimated more accurately.

  Moreover, although the collision prediction apparatus 13 demonstrated the example which estimates whether the site | part of the vehicle which a detection object may collide with is the front or side of a vehicle in the said embodiment, the collision prediction apparatus 13 You may determine in more detail not only a front surface and a side surface, but the site | part to which a detected object collides. For example, when it is estimated that the detected object collides with the front surface of the vehicle 1, the collision prediction device 13 may further estimate whether the detected object collides with the left side or the right side of the front surface of the vehicle. In addition, in the case where two or more parts are estimated and determined as parts of the vehicle 1 with which the detected object may collide as described above, the collision prediction device 13 uses a threshold value table corresponding to each part that is a candidate. May be stored, and the threshold value table may be switched.

  In the above-described embodiment, the collision prediction device 13 has described an example in which a part of a vehicle where a detected object may collide is estimated based on the own vehicle traveling route line, the detected object traveling route line, and the like. The collision prediction device 13 may estimate the part of the vehicle where the detected object may collide based on another conventionally known method.

  In the above-described embodiment, an example in which the radar device 10 is used as a means for detecting an object existing around the vehicle has been described. However, any other device can be used as long as it can detect an object present around the vehicle. It may be used instead of the device 10. For example, you may use the image processing apparatus etc. which can detect the position of the own vehicle and the object around a vehicle based on the image around a vehicle.

  Moreover, in the said embodiment, although the collision prediction apparatus 13 demonstrated the example which estimates the collision with the vehicle 2 which is a moving body, and the vehicle 1, the collision prediction apparatus 13 is not restricted to a moving body like the vehicle 2, It is possible to accurately predict a collision between a stationary object and the vehicle 1 based on the processing described above.

  Moreover, in the said embodiment, when it was estimated that the vehicle 1 collides with a detected object, the example which alert | reports an alarm by the alarm device 14 was shown, However, The result of the said prediction is not restricted to operation | movement of an alarm device, Other It may be used to control the operation of the device. For example, a safety device such as a seat belt device mounted on the vehicle 1 or a braking device such as a brake device may be automatically controlled.

  The collision prediction apparatus according to the present invention is useful as a collision prediction apparatus that can accurately predict the risk of collision between an object existing ahead of the host vehicle and the host vehicle.

Block diagram showing the functional configuration of the collision prediction device An example of a flowchart showing processing executed by the collision prediction device 13 The top view which shows the position of the vehicle 1, the position of the vehicle 2 which is a detected object, and the detection angle of the vehicle 2 in a surrounding map coordinate system An example of a flowchart showing threshold setting processing The top view which shows the vehicle 1, the vehicle 2, the own vehicle travel route line Bs, and the detected object travel route line Bt in the surrounding map coordinate system The figure which showed an example of the front collision threshold value table The figure which showed an example of the side collision threshold value table The top view which shows track | orbit W_p where the vehicle 2 which advances facing the vehicle 1 passes the right side of the vehicle 1 The top view which shows track | orbit W_fL which the vehicle 2 approaches from the diagonally right front of the vehicle 1, and collides with the front left side of the vehicle 1 The top view which shows track | orbit W_fR which the vehicle 2 approaches from the diagonal right front of the vehicle 1, and collides with the front right side of the vehicle 1 The graph which shows the relationship between detection angle variation | change_quantity (theta) v in the track | orbit W_p, the track | orbit W_fL, and the track | orbit W_fR, and the collision estimated time TTC. The top view which shows track | orbit W_sF which the vehicle 2 approaches from the diagonally right front of the vehicle 1, and collides with the right side front surface of the vehicle 1 The top view which shows track | orbit W_sB which the vehicle 2 approaches from the diagonal right front of the vehicle 1, and collides with the right side rear surface of the vehicle 1 The graph which shows the relationship between detection angle variation | change_quantity (theta) v in track | orbit W_sF and track | orbit W_sB, and the collision estimated time TTC. The top view which shows the vehicle 1 by which the observation point Po is set to the rear end left corner A graph showing the relationship between the estimated time until the vehicle collides with the detected object and the change rate of the angular position when the observation point is set at the front end of the vehicle

Explanation of symbols

DESCRIPTION OF SYMBOLS 1, 2 Vehicle 10 Radar apparatus 11 Speedometer 12 Own vehicle position detection apparatus 13 Collision prediction apparatus 14 Alarm apparatus

Claims (11)

Object detection means for detecting the position of an object in front of the vehicle;
Observation point setting means for determining an observation point for observing the object in the vehicle or on the outer peripheral surface of the vehicle and behind the front end of the vehicle;
Detection angle calculation means for calculating a detection angle indicating a detection direction of the object viewed from the observation point;
A change amount calculating means for calculating a change amount per unit time of the detection angle;
A collision prediction device comprising: a collision determination unit that predicts a risk of collision between the vehicle and the object based on the value of the change amount. Collision part estimation means for estimating an outer peripheral surface of the vehicle, which is likely to collide with the object, as a predicted collision part;
Threshold setting means for setting a threshold used for predicting the risk of collision between the vehicle and the object according to the predicted collision site estimated by the collision site estimation means,
The collision prediction device according to claim 1, wherein the collision determination unit predicts a collision between the vehicle and the object according to whether or not the change amount satisfies the threshold value. Own vehicle travel route calculating means for calculating the travel route of the vehicle;
A detected object travel path calculating means for calculating a travel path of the object that moves relative to the vehicle;
The collision prediction apparatus according to claim 2, wherein the collision site estimation unit estimates the predicted collision site based on a travel path of the vehicle and a travel path of the object.   The collision prediction apparatus according to claim 3, wherein the collision site estimation unit sets the front surface of the vehicle as the predicted collision site when the traveling path of the vehicle and the traveling path of the object are parallel to each other. A collision point estimation means for estimating an intersection between the traveling path of the vehicle and the traveling path of the object moving relative to the vehicle as a predicted collision point;
The collision according to claim 3, wherein the collision site estimation unit determines which of the vehicle or the object first reaches the predicted collision point and estimates the predicted collision site according to the determination result. Prediction device.   When the collision part estimation unit determines that the object reaches the predicted collision point before the vehicle, the front surface of the vehicle is set as the predicted collision part, and the vehicle moves to the predicted collision point before the object. The collision prediction device according to claim 5, wherein, when it is determined that the vehicle reaches the vehicle, the side surface of the vehicle is set as the collision predicted portion. A collision time calculating means for calculating an estimated collision time required for the collision when the vehicle and the object collide,
The collision according to claim 2, wherein the threshold setting unit sets the threshold according to the predicted collision portion estimated by the collision portion estimation unit and the predicted collision time calculated by the collision time calculation unit. Prediction device. When the side surface of the vehicle is the predicted collision part, the threshold setting unit sets the threshold to a larger value stepwise as the predicted collision time becomes shorter.
The collision determination unit predicts that the object and the vehicle collide when the change amount is larger than the threshold value, and predicts that the object and the vehicle do not collide when the change amount is equal to or less than the threshold value. The collision prediction apparatus according to claim 7.   The collision prediction apparatus according to claim 1, wherein the observation point setting unit sets the observation point at a center of a rear end of the vehicle.   The collision prediction apparatus according to claim 1, wherein the observation point setting unit sets the observation point at a left corner of a rear end of the vehicle.   The collision prediction apparatus according to claim 1, wherein the observation point setting unit sets the observation point at a right corner of a rear end of the vehicle.

Images