Misalignment processing for a vehicle radar sensor
View Patent ↗The present invention relates to a vehicle radar system ( 2 ) arranged to detect objects outside a vehicle ( 1 ). The radar system ( 2 ) a radar detector ( 3 ) and a processing unit ( 4 ). The processing unit ( 4 ) is arranged to obtain values for detected target angle (θ err ) and detected target Doppler velocity (v d ) relative the radar detector ( 3 ) for each detected object ( 10 a ′, 10 b ′, 10 c ′, 10 d ′, 10 e ′) during a certain time interval. If there is a zero crossing ( 14 ) for a derivative ( 13 ) of a function ( 12 ) describing the progression of detected target Doppler velocity (v d ) as a function of detected target angle (θ err ), the processing unit ( 4 ) is arranged to detect the zero crossing ( 14 ). This zero crossing ( 14 ) is indicative of a radar system misalignment (θ m ). The present invention also relates to a corresponding method.
1. A vehicle radar system arranged to detect objects outside a vehicle, the radar system comprising a radar detector and a processing unit, the processing unit being arranged to obtain values for detected target angle (θ err ) and detected target Doppler velocity (v d ) relative to the radar detector for each of the objects during a certain time interval, the processing unit configured such that if there is a zero crossing for a function describing the progression of the detected target Doppler velocity (v d ) as a function of the detected target angle (θ err ), the processing unit is further configured to calculate the zero crossing, the zero crossing being indicative of a radar system misalignment angle (θ m ).
2. A vehicle radar system according to claim 1 , further comprising in that the function describing the progression of detected target Doppler velocity as a function of detected target angle is numerically defined such that the function describing the progression has a progression like a parabola.
3. A vehicle radar system according to claim 2 , further comprising in that the function describing the progression of detected target Doppler velocity (v d ) as a function of detected target angle (θ err ) is defined as:
v
d
=
v
h
·
(
1
-
Θ
err
2
2
)
,
where v h represents vehicle velocity.
4. A vehicle radar system according to claim 1 , further comprising the zero crossing is determined as a derivative of the function describing the progression.
5. A vehicle radar system according to claim 1 , further comprising the radar system misalignment angle (θ m ) represents the difference between the detected target angle (θ err ) and a true target angle (θ ref ), and the processing unit using the system misalignment angle (θ m ) to correct further detected values of the detected target angle (θ err ).
6. A vehicle radar system according to claim 5 , further comprising the difference between the detected target angle (θ err ) and the true target angle (θ err ), is relative to a longitudinal line extending from the vehicle.
7. A method for estimating a vehicle radar system misalignment angle (θ m ), the vehicle radar system being used for detecting objects outside a vehicle, the method comprising the steps of:
detecting a target angle (θ err ) and a target Doppler velocity (v d ) for each of the objects during a certain time interval;
calculating a function describing the progression of the detected target Doppler velocity (v d ) as a function of the detected target angle (θ err );
if there is a zero crossing for the function describing the progression, finding the zero crossing; and
using the zero crossing as an estimation of the vehicle radar system misalignment angle (θ m ).
8. A method according to claim 7 , further comprising in that the function describing the progression of the detected target Doppler velocity (v d ) as a function of the detected target angle (θ err ) is numerically defined as a progression like a parabola.
9. A method according to claim 7 , further comprising in that the function describing the progression of the detected target Doppler velocity (v d ) as a function of detected target angle (θ err ) is defined as:
v
d
=
v
h
·
(
1
-
Θ
err
2
2
)
,
where v h represents velocity of the vehicle.
10. A method according to claim 8 further comprising in that the step of finding the zero crossing comprises the step of:
calculating data Y representing a gradient of the parabola in a center between two data points according to:
Y
(
i
=
1
…
n
-
1
)
=
Vd
(
i
+
1
)
-
Vd
(
i
)
(
i
+
1
)
-
(
i
)
,
where n is the number of data points, and θ is an angle formed between a line extending from the vehicle and a line to the object detected,
modifying the ordinate X according to:
X
(
i
=
1
…
n
-
1
)
=
1
2
*
(
(
i
)
+
(
i
+
1
)
)
,
performing linear regression to estimate the parameters m and b of a straight line according to:
Y=m*X+b , and
calculating the misalignment error (θ m ) as:
m
=
-
b
m
.
11. A method according to claim 7 , further comprising the step of deriving the function describing the progression.
12. A method according to claim 7 , further comprising the radar system misalignment angle (θ m ) represents the difference between the detected target angle (θ err ) and a true target angle (θ ref ), and using the system misalignment angle (θ m ) to correct further detected values of the detected target angle (θ err ).