Method and apparatus for target vehicle identification in automatic cruise control and collision avoidance systems

• US 6,826,479 B2
• Filed: 06/03/2002
• Issued: 11/30/2004
• Est. Priority Date: 06/03/2002
• Status: Active Grant

• Current Assignee: Visteon Global Technologies Incorporated (Visteon Corporation)
• Original Assignee: Visteon Global Technologies Incorporated (Visteon Corporation)
• Inventors: Miyahara, Shunji
• Primary Examiner(s): LOUIS JACQUES, JACQUES H
• Application Number:
  US10/161,112
• Publication Number:
  US 20030225519A1
• Time in Patent Office: 911 Days
• Field of Search: 701/1, 701/23-26, 701/93, 701/96, 701/300-301, 180/167-170, 180/271, 180/282, 340/901-905, 340/435-436
• US Class Current: 701/301
• CPC Class Codes: B60K 31/0008 including means for detecting potential obstacles in vehicle path
• CPC Class Codes: B60K 31/0066 responsive to vehicle path curvature
• CPC Class Codes: B60W 2552/20 Road profile, i.e. the change in elevation or curvature of a plurality of continuous road segments
• CPC Class Codes: B60W 2552/30 Road curve radius
• CPC Class Codes: B60W 2554/4041 Position
• CPC Class Codes: B60W 30/16 Control of distance between vehicles, e.g. keeping a distance to preceding vehicle
• CPC Class Codes: G08G 1/161 Decentralised systems, e.g. inter-vehicle communication

Abstract

An apparatus and method are provided for identifying a preceding vehicle as a target vehicle in transient and non-transient situations. The method generally includes, estimating a position of the target vehicle relative to a position of an ACC vehicle from the position of the ACC vehicle to create a target vehicle estimated position; creating a time-dependent position trace of the target vehicle; estimating a radius of the position trace of the target vehicle by applying circular curve fitting to the position trace of the target vehicle; and determining whether the target vehicle is in an ACC vehicle lane using the radius of the position trace, the target vehicle estimated position and the position of the ACC vehicle.

Claims

1. A method for identifying a target vehicle in transient and non-transient situations, comprising:

determining a position of an ACC vehicle;

estimating a position of a preceding vehicle relative to the position of the ACC vehicle from the position of the ACC vehicle;

creating a time-dependent position trace of the preceding vehicle;

estimating a radius “

a”

of the position trace of the preceding vehicle by applying circular curve fitting including calculating an equation of a curve estimating the position trace of the preceding vehicle; and

determining whether the preceding vehicle is in a path of the ACC vehicle by examining a curve fitting error of the curve.

2. A method for identifying a target vehicle in transient and non-transient situations as claimed in claim 1, wherein determining the position of the ACC vehicle comprises:

assigning a value to the position of the ACC vehicle at time t=0; and

calculating the position of the ACC vehicle at time t>

0 by using a speed and an angular velocity of the ACC vehicle.

3. A method for identifying a target vehicle in transient and non-transient situations as claimed in claim 2, wherein the position of the ACC vehicle comprises an ACC location and an ACC bearing and assigning a value to the position of the ACC vehicle at time t=0 comprises:

assuming the ACC location of (0,0) in an x-y coordinate system; and

assuming the ACC bearing along a y-axis of the x-y coordinate system.

4. A method for identifying a target vehicle in transient and non-transient situations as claimed in claim 2, wherein the angular velocity is determined from a lateral velocity and a longitudinal velocity of the ACC vehicle at time t>

0 wherein the longitudinal and lateral velocities are calculated from a yaw-rate and a speed of an ACC vehicle, respectively.

5. A method for identifying a target vehicle in transient and non-transient situations as claimed in claim 1, wherein estimating the position of the preceding vehicle comprises estimating the position of the preceding vehicle relative to that of the ACC vehicle, wherein the position relative to the position of the ACC vehicle is determined by the measured distance and azimuth angle, and the position of the ACC vehicle.

6. A method for identifying a target vehicle in transient and non-transient situations as claimed in claim 5, wherein estimating the position of the preceding vehicle further comprises expressing the position of the preceding vehicle as a position in an x-y coordinate system (x(t), y(t)) as a function of the position of the ACC vehicle in the x-y coordinate system (xa(t), ya(t)), an ACC vehicle bearing φ

_t, an azimuth angle θ

_t, and a distance between the ACC vehicle and the preceding vehicle R_t over a portion of time t=m,m+1, . . . ,n−

1, n by an equation (x(t), y(t))=(xa(t)+R_t cos(φ

_t+θ

_t), ya(t)+R_t sin(φ

_t+θ

_t)).

7. A method for determining a target vehicle in transient and non-transient situations as claimed in claim 6, wherein creating a time-dependent position trace of the preceding vehicle further comprises estimating the position of the preceding vehicle at a plurality of times.

8. A method for determining a target vehicle in transient and non-transient situations as claimed in claim 6, wherein creating a time-dependent position trace of the preceding vehicle further comprises transforming the time-dependent position trace from an x-y coordinate system into a local coordinate system (ξ

,η

).

9. A method for identifying a target vehicle in transient and non-transient situations as claimed in claim 8, transforming the time-dependent position trace from the x-y coordinate system into the local coordinate system (ξ

,η

) to define the circle comprises, transforming a plurality of points (x_n,y_n),(x<sub>n−

1</sub>,y<sub>n−

1</sub>),(x<sub>n−

2</sub>,y<sub>n−

2</sub>),(x<sub>n−

3</sub>,y<sub>n−

3</sub>), . . . in the x-y coordinate system into a second plurality of points (ξ

_n,η

_n),(0,0),(ξ

<sub>n−

2</sub>,η

<sub>n−

2</sub>),(ξ

<sub>n−

3</sub>,η

<sub>n−

3</sub>), . . . in the local coordinate system.

10. A method for identifying a target vehicle in transient and non-transient situations as claimed in claim 9, wherein transforming the x-y coordinate to the local coordinate system (ξ

,η

) to apply the curve fitting further comprises, using a relationship among the position vector of an i-th point of the preceding vehicle and unit vectors of the local coordinate system defined by (({circumflex over (r)}_i−

{circumflex over (r)}<sub>n−

1</sub>)•

{circumflex over (ξ

)},({circumflex over (r)}_i−

{circumflex over (r)}<sub>n−

1</sub>)•

{circumflex over (η

)}), where i=n, n−

1, n2, . . . .

11. A method for identifying a target vehicle in transient and non-transient situations as claimed in claim 1, wherein estimating the radius of the position trace of the preceding vehicle by applying circular curve fitting to the position trace of the preceding vehicle in a local coordinate comprises:

defining the curve in a local (ξ

,η

) coordinate system;

defining an error in the local (ξ

,η

) coordinate system;

approximating the curve by a quadratic equation; and

using quadratic regression to calculate the radius “

a”

.

12. A method for identifying a target vehicle in transient and non-transient situations as claimed in claim 11, wherein defining the curve in the local coordinate system comprises assuming the position trace is on the circle, wherein the circle has a radius “

a” and

is defined by;

(ξ

−

ξ

_c)²+(η

−

η

_c)²=a².

13. A method for identifying a target vehicle in transient and non-transient situations as claimed in claim 11, wherein defining the error in the local coordinate system comprises defining an error “

e”

between the position trace and the curve, by;

e=Σ

((ξ

_i−

ξ

_c)²+(η

_i−

η

_c)²−

a²)².

14. A method for identifying a target vehicle in transient and non-transient situations as claimed in claim 11, wherein approximating the curve by quadratic equation comprises defining the curve by:

$\mathrm{\eta <br><br>}=\frac{1}{2<br><br>\text{\hspace{1em}<br><br>}<br><br>{\mathrm{\eta <br><br>}}_c}<br><br>{\mathrm{\xi <br><br>}}^2-\frac{{\mathrm{\xi <br><br>}}_c}{{\mathrm{\eta <br><br>}}_c}<br><br>\mathrm{\xi <br><br>}+\frac{\left({\mathrm{\eta <br><br>}}_c^2-a^2\right)}{2<br><br>\text{\hspace{1em}<br><br>}<br><br>{\mathrm{\eta <br><br>}}_c}<br><br>\text{\hspace{1em}<br><br>}.$

15. A method for identifying a target vehicle in transient and non-transient situations as claimed in claim 14, wherein using quadratic regression to calculate the radius “

a”

comprises, applying quadratic regression to the quadratic equation which defines a as a function of regression coefficients c0 and c2, by $a=\sqrt{\frac{1}{4·<br><br>{\mathrm{c2}}^2}-\frac{\mathrm{c0}}{\mathrm{c2}}}<br><br>\text{\hspace{1em}<br><br>}.$

16. A method for identifying a target vehicle in transient and non-transient situations as claimed in claim 1, wherein determining whether the preceding vehicle is in the path of the ACC vehicle by examining the curve fitting error comprises examining an accuracy of the radius “

a”

.

17. A method for identifying a target vehicle in transient and non-transient situations as claimed in claim 16, wherein examining the accuracy of the radius “

a”

comprises;

defining a first relationship between a coefficient c2 and a standard deviation of the coefficient c2 “

σ

_c,”

wherein the relationship between c2 and σ

_c is defined by;

$<br><br>\frac{{\mathrm{\sigma <br><br>}}_{\mathrm{c2}}}{\mathrm{c2}}<br><br><<br><br>0.25\sim <br><br>1;<br><br>\mathrm{or}$

defining a second relationship between the radius “

a” and

a standard deviation of the radius “

σ

_a”

wherein the relationship between “

a” and

“

σ

_a,”

is defined by;

|σ

_a/a|<

0.25;

wherein either of the first or second relationships define a regression criteria.

18. A method for identifying a target vehicle in transient and non-transient situations as claimed in claim 17, wherein examining the accuracy of the radius “

a”

further comprises defining the preceding vehicle as the target vehicle if the regression criteria are satisfied.

19. A method for identifying a target vehicle in transient and non-transient situations as claimed in claim 1, wherein determining whether the preceding vehicle is in the path of the ACC vehicle by examining the curve fitting error comprises:

determining a lateral distance between the position of the ACC vehicle and the curve; and

comparing a magnitude of the lateral distance to a lane width and a vehicle width, using a lateral relationship;

|lateral distance|≦

(lane width+vehicle width)/2, wherein if the lateral relationship is satisfied, the preceding vehicle is the target vehicle.

20. A method for identifying a target vehicle in transient and non-transient situations as claimed in claim 19, wherein determining the lateral distance comprises:

determining a location of a centerline forward point of the ACC vehicle in an x-y coordinate system;

transforming the location of the centerline forward point to a local coordinate system;

determining a point on the curve (ξ

_k, η

_k), wherein the point on the curve defines a line tangent to the circle and parallel to an ACC vehicle centerline; and

defining the lateral distance as a distance between the centerline forward point and the point (ξ

_k, η

_k).

21. The method according to claim 1, wherein estimating a radius “

a”

of the position trace includes estimating a radius “

a”

of a position trace curve defined by positions along the position trace of the preceding vehicle by calculating an equation of a curve estimating the position trace curve.

22. The method according to claim 1, wherein determining whether the preceding vehicle is in the path of the ACC vehicle includes determining whether the preceding vehicle is in the path of the ACC vehicle by examining an error between the curve and the position trace.

23. The method according to claim 22, wherein estimating a radius “

a”

of the position trace includes estimating a radius “

a”

of a position trace curve defined by positions along the position trace of the preceding vehicle by calculating an equation of a curve estimating the position trace curve.

24. A method for determining the lane position of a preceding vehicle in transient and non-transient situations, comprising:

determining a position of an ACC vehicle;

estimating a position of a preceding vehicle relative to the position of the ACC vehicle from the position of the ACC vehicle to create the preceding vehicle estimated position;

creating a time-dependent position trace of the preceding vehicle;

estimating a radius “

a”

of the position trace of the preceding vehicle by applying circular curve fitting including calculating an equation of a curve estimating the position trace of the preceding vehicle; and

determining whether the preceding vehicle is in an ACC vehicle lane by examining a curve fitting error of the curve.

25. A method for determining the lane position of a preceding vehicle in transient and non-transient situations as claimed in claim 24, wherein estimating the position of the preceding vehicle relative to the position of the ACC vehicle from the position of the ACC vehicle to create the preceding vehicle estimated position comprises, determining a distance between the ACC vehicle and the preceding vehicle and determining an azimuth angle between the ACC vehicle and the preceding vehicle.

26. A method for determining the lane position of a preceding vehicle in transient and non-transient situations as claimed in claim 24, wherein creating the time-dependent position trace of the preceding vehicle comprises, estimating the time-dependent position trace from a plurality of positions of the preceding vehicle.

27. A method for determining the lane position of a preceding vehicle in transient and non-transient situations as claimed in claim 24, wherein estimating the radius of the position trace by applying circular curve fitting to the position trace of the preceding vehicle comprises, obtaining a curve in a local coordinate system with the radius so that an error between the curve and the position trace is minimized.

28. A method for determining the lane position of a preceding vehicle in transient and non-transient situations as claimed in claim 27, wherein obtaining a curve in the local coordinate system comprises selecting an origin approximately near or approximately at one of the plurality of positions of the preceding vehicle.

29. A method for determining the lane position of a preceding vehicle in transient and non-transient situations as claimed in claim 27, obtaining the curve in the local coordinate system wherein a first axis of the local coordinate system is approximately tangent to the curve and a second axis of the coordinate system is normal to the first axis.

30. A method for determining the lane position of a preceding vehicle in transient and non-transient situations as claimed in claim 24, wherein determining whether the preceding vehicle is in an ACC vehicle lane by examining a curve fit error, comprises:

defining regression criteria and determining whether the regression criteria are met.

31. A method for determining the lane position of a preceding vehicle in transient and non-transient situations as claimed in claim 24, wherein determining whether the preceding vehicle is in an ACC vehicle lane by examining a curve fit error, comprises:

determining a lateral distance between the position of the ACC vehicle and the curve.

32. The method according to claim 24, wherein determining whether the preceding vehicle is in the path of the ACC vehicle includes determining whether the preceding vehicle is in the path of the ACC vehicle by examining an error between the curve and the position trace.

33. A target identification system, comprising:

a memory device that stores position information of an ACC vehicle and a distance and azimuth angle between the ACC vehicle and a preceding vehicle;

a processor and computer software that

computes a position of a target vehicle using the position information of the ACC vehicle and the distance and azimuth angle;

creates a position trace;

transforms the position trace into a local coordinate system (ξ

,η

); and

estimates a radius of the position trace using curve fitting, and quadratic regression.

34. An ACC system, comprising:

a target vehicle sensor;

at least one ACC vehicle sensor;

the target identification system as claimed in claim 33; and

an ACC vehicle control mechanism.

35. A method for identifying a target vehicle in transient and non-transient situations, comprising:

determining a position of an ACC vehicle;

estimating a position of a preceding vehicle relative to the position of the ACC vehicle from the position of the ACC vehicle;

creating a time-dependent position trace of the preceding vehicle;

estimating a radius “

a”

of the position trace of the preceding vehicle by applying circular curve fitting to the position trace of the preceding vehicle to create a curve;

determining whether the preceding vehicle is in a path of the ACC vehicle by examining a curve fitting error;

wherein estimating the position of the preceding vehicle comprises estimating the position of the preceding vehicle relative to that of the ACC vehicle, wherein the position relative to the position of the ACC vehicle is determined by the measured distance and azimuth angle, and the position of the ACC vehicle; and

wherein estimating the position of the preceding vehicle further comprises expressing the position of the preceding vehicle as a position in an x-y coordinate system (x(t), y(t)) as a function of the position of the ACC vehicle in the x-y coordinate system (xa(t), ya(t)), an ACC vehicle bearing φ

_t, an azimuth angle θ

_t, and a distance between the ACC vehicle and the preceding vehicle R_t over a portion of time t=m,m+1, . . . , n−

1, n by an equation (x(t), y(t))=(xa(t)+R_t cos(φ

_t+θ

_t), ya(t)+R_t sin(φ

_t+θ

_t)).

36. A method for identifying a target vehicle in transient and non-transient situations, comprising:

determining a position of an ACC vehicle;

estimating a position of a preceding vehicle relative to the position of the ACC vehicle from the position of the ACC vehicle;

creating a time-dependent position trace of the preceding vehicle;

estimating a radius “

a”

of the position trace of the preceding vehicle by applying circular curve fitting to the position trace of the preceding vehicle to create a curve;

determining whether the preceding vehicle is in a path of the ACC vehicle by examining a curve fitting error;

wherein estimating the radius of the position trace of the preceding vehicle by applying circular curve fitting to the position trace of the preceding vehicle in a local coordinate comprises;

defining the curve in a local (ξ

,η

) coordinate system;

defining an error in the local (ξ

,η

) coordinate system;

approximating the curve by a quadratic equation; and

using quadratic regression to calculate the radius “

a”

.