Radar calibration system
A system includes a computer including a processor and a memory. The memory includes instructions such that the processor is programmed to: receive, from a radar sensor of a vehicle, radar data indicative of a stationary object proximate to the radar sensor; receive, from a non-radar sensor of the vehicle, vehicle state data indicative of a vehicle state, the vehicle state data indicative of at least a longitudinal velocity and a yaw rate of the vehicle; determine an orientation estimate and an offset estimate of the radar sensor based on the radar data and the vehicle state data; and determine whether to actuate a vehicle system based on at least one of the orientation estimate or the offset estimate.
1. A system, comprising a computer including a processor and a memory, the memory including instructions such that the processor is programmed to:
receive, from a radar sensor of a vehicle, radar data indicative of a stationary object proximate to the radar sensor;
receive, from a non-radar sensor of the vehicle, vehicle state data indicative of a vehicle state, the vehicle state data indicative of at least a longitudinal velocity and a yaw rate of the vehicle;
associate the vehicle state data with the radar data based on a respective timestamp of the vehicle state data and the radar data being received within a defined time period;
determine an orientation estimate and an offset estimate of the radar sensor based on the radar data and the vehicle state data; and
determine whether to actuate a vehicle system based on at least one of the orientation estimate or the offset estimate.
2. The system of claim 1 , wherein the processor is further programmed to:
actuate the vehicle system based on the at least one of the orientation estimate or the offset estimate.
3. The system of claim 2 , wherein the actuation of the vehicle system includes disabling an autonomous driving mode of the vehicle.
4. The system of claim 1 , wherein the processor is further programmed to:
determine the orientation estimate and the offset estimate of the radar sensor according to v x cos(θ+ϕ)+v y sin(θ+ϕ)+{dot over (ψ)}(L x sin(θ+ϕ)−L y cos(θ+ϕ))+{dot over (r)}=0, where v x represents the longitudinal velocity, v y represents lateral velocity, {dot over (ψ)} represents the yaw rate, ϕ represents an angle of a radar sensor return relative to a center of the radar sensor, {dot over (r)} represents a range rate of the radar sensor return, θ represents the orientation estimate of the radar sensor, L x represents a longitudinal position of the radar sensor, and L y represents a lateral position of the radar sensor.
5. The system of claim 4 , wherein the processor is further programmed to:
determine the orientation estimate of the radar sensor based on a Kalman filter.
6. The system of claim 5 , wherein a Kalman filter gain of the Kalman filter is defined as K=ΣH T (HΣH T +J Vx Σ Vx J Vx T +J Vy Σ Vy J Vy T +J ϕ Σ ϕ J ϕ T +Σ{dot over (r)}) −1 , where H is a Jacobian of the {dot over (r)} with respect to the orientation estimate, H T represents a transpose of the Jacobian H, J Vx represents a Jacobian of the longitudinal velocity with respect to the {dot over (r)}, J Vy represents a Jacobian of the lateral velocity with respect to the {dot over (r)}, J ϕ represents a Jacobian of an azimuth of the radar sensor with respect to the {dot over (r)}, Σ Vx represents a variance associated with v x , Σ Vy represents a variance associated with v y , Σ ϕ represents a variance associated with ϕ, and Σ{dot over (r)} represent a variance associated with {dot over (r)}.
7. The system of claim 4 , wherein the processor is further programmed to:
determine the offset estimate of the radar sensor based on a Kalman filter.
8. The system of claim 7 , wherein a Kalman filter gain of the Kalman filter is defined as Ko=ΣH T (HΣH T +J {dot over (ψ)} Σ {dot over (ψ)} J {dot over (ψ)} T +J Vx Σ Vx J Vx T +J Vy Σ Vy J Vy T +J ϕ Σ ϕ J ϕ T +J θ Σ θ J θ T +Σ{dot over (r)}) −1 , where H is a Jacobian of the {dot over (r)} with respect to the offset estimate, H T represents a transpose of the Jacobian H, J {dot over (ψ)} represents a Jacobian of the yaw rate with respect to the {dot over (r)}, J Vx represents a Jacobian of the longitudinal velocity with respect to the r , J Vy represents a Jacobian of the lateral velocity with respect to the {dot over (r)}, J ϕ represents a Jacobian of an azimuth of the radar sensor with respect to the {dot over (r)}, Σ{dot over (ψ)} represents a variance associated with {dot over (ψ)}, Σ Vx represents a variance associated with v x , Σ Vy represents a variance associated with v y , Σ ϕ represents a variance associated with ϕ, and Σ{dot over (r)} represent a variance associated with {dot over (r)}.
9. The system of claim 1 , wherein the non-radar sensor includes at least one of a wheel odometry sensor, an inertial measurement unit, a visual odometry sensor, or a Global Positioning System sensor.
10. The system of claim 1 , wherein the processor is further programmed to:
filter radar data and vehicle state data indicating that at least one of a lateral velocity is greater than or equal to a predetermined lateral velocity threshold or a yaw rate is greater than or equal to a predetermined yaw rate threshold.
11. The system of claim 1 , wherein the processor is further programmed to:
select an initial orientation based on a most-occurring calculated orientation value within a plurality of calculated orientation values.
12. A system, comprising a computer including a processor and a memory, the memory including instructions such that the processor is programmed to:
receive, from a radar sensor of a vehicle, radar data indicative of a stationary object proximate to the radar sensor;
receive, from a non-radar sensor of the vehicle, vehicle state data indicative of a vehicle state, the vehicle state data indicative of at least a longitudinal velocity and a yaw rate of the vehicle;
filter radar data and vehicle state data indicating that at least one of a lateral velocity is greater than or equal to a predetermined lateral velocity threshold or a yaw rate is greater than or equal to a predetermined yaw rate threshold;
determine an orientation estimate and an offset estimate of the radar sensor based on the radar data and the vehicle state data; and
determine whether to actuate a vehicle system based on at least one of the orientation estimate or the offset estimate.
13. The system of claim 12 , wherein the processor is further programmed to:
associate the vehicle state data with the radar data based on a respective timestamp of the vehicle state data and the radar data being received within a defined time period.
14. The system of claim 12 , wherein the processor is further programmed to:
actuate the vehicle system based on the at least one of the orientation estimate or the offset estimate.
15. The system of claim 14 , wherein the actuation of the vehicle system includes disabling an autonomous driving mode of the vehicle.
16. The system of claim 12 , wherein the processor is further programmed to: determine the orientation estimate and the offset estimate of the radar sensor according to to v x cos(θ+ϕ)+v y sin(θ+ϕ)+{dot over (ψ)}(L x sin(θ+ϕ)−L y cos(θ+ϕ))+{dot over (r)}=0, where v x represents the longitudinal velocity, v y represents lateral velocity, {dot over (ψ)} represents the yaw rate, ϕ represents an angle of a radar sensor return relative to a center of the radar sensor, {dot over (r)} represents a range rate of the radar sensor return, θ represents the orientation estimate of the radar sensor, L x represents a longitudinal position of the radar sensor, and L y represents a lateral position of the radar sensor.
17. The system of claim 12 , wherein the non-radar sensor includes at least one of a wheel odometry sensor, an inertial measurement unit, a visual odometry sensor, or a Global Positioning System sensor.
18. The system of claim 12 , wherein the processor is further programmed to:
select an initial orientation based on a most-occurring calculated orientation value within a plurality of calculated orientation values.
19. A system, comprising a computer including a processor and a memory, the memory including instructions such that the processor is programmed to:
receive, from a radar sensor of a vehicle, radar data indicative of a stationary object proximate to the radar sensor;
receive, from a non-radar sensor of the vehicle, vehicle state data indicative of a vehicle state, the vehicle state data indicative of at least a longitudinal velocity and a yaw rate of the vehicle;
select an initial orientation based on a most-occurring calculated orientation value within a plurality of calculated orientation values;
determine an orientation estimate and an offset estimate of the radar sensor based on the radar data and the vehicle state data; and
determine whether to actuate a vehicle system based on at least one of the orientation estimate or the offset estimate.
20. The system of claim 18 , wherein the processor is further programmed to:
actuate the vehicle system based on the at least one of the orientation estimate or the offset estimate.
21. The system of claim 18 , wherein the processor is further programmed to:
determine the orientation estimate and the offset estimate of the radar sensor according to to v x cos(θ+ϕ)+v y sin(θ+ϕ)+{dot over (ψ)}(L x sin(θ+ϕ)−L y cos(θ+ϕ))+{dot over (r)}=0, where v x represents the longitudinal velocity, v y represents lateral velocity, {dot over (ψ)} represents the yaw rate, ϕ represents an angle of a radar sensor return relative to a center of the radar sensor, {dot over (r)} represents a range rate of the radar sensor return, θ represents the orientation estimate of the radar sensor, L x represents a longitudinal position of the radar sensor, and L y represents a lateral position of the radar sensor.