Real-time and dynamic localization using active doppler sensing systems for vehicles

1. A system on a moving vehicle comprising:

an active Doppler sensor;

one or more processors; and

a memory storing instructions that, when executed by the one or more processors, causes the system to perform:

obtaining a Doppler velocity from each of one or more entities using the active Doppler sensor, wherein the Doppler velocity is with respect to the moving vehicle;

determining one or more calibration parameters of the active Doppler sensor based on respective Doppler velocities from at least a portion of the entities, wherein the determining of the one or more calibration parameters comprises:

determining one or more stationary entities, within the one or more entities;

forming one or more clusters, each of the clusters comprising at least a portion of the one or more stationary entities and defining a relationship between a location of one of the entities with respect to the active Doppler sensor and the Doppler velocity of the respective one of the entities;

fitting, to each of the one or more clusters, a signal model defining the Doppler velocities of the each of the entities in relation to a mounting offset angle or an azimuth offset angle of the active Doppler sensor and a mounting location of the active Doppler sensor;

determining a degree of fit of the signal model to each of the one or more clusters;

removing any of the clusters having a degree of fit lower than a threshold while retaining remaining clusters; and determining the one or more calibration parameters based on the retained remaining clusters;

determining a location of the vehicle based on one or more of the calibration parameters; and

causing the vehicle to perform a driving action based on the determined location.

2. The system of claim 1 , wherein the determining the one or more calibration parameters further comprises:

determining the mounting location, the mounting offset angle, a linear velocity component, and a rotational velocity component of the active Doppler sensor.

3. The system of claim 1 , wherein the determining the location of the vehicle comprises determining the location of the vehicle by integrating the determined linear velocity component over time or the determined rotational velocity component over time.

4. The system of claim 3 , wherein the instructions further cause the system to perform:

comparing the determined location with a location obtained from a GPS or an IMU.

5. The system of claim 1 , wherein:

the signal model is defined by V D =V L cos (Θ+β)+V R sin (Θ+β−γ), and wherein:

the calibration parameters comprise V L , V R , β, and γ;

V D indicates the Doppler velocity of the each of the entities;

V L indicates a linear velocity component of the each of the entities in a forward direction of the vehicle;

V R indicates a rotational velocity component of the each of the entities;

Θ indicates an angle defining the location of the each of the entities with respect to the active Doppler sensor;

β is the mounting offset angle of the active Doppler sensor; and

γ is defined as γ=β−arctan (y/x), wherein y and x define the mounting location of the active Doppler sensor with respect to an ego coordinate system of the vehicle.

6. The system of claim 5 , wherein the determining the one or more calibration parameters further comprises:

determining, for each of the retained clusters, a first difference between the linear velocity component and a linear velocity obtained from a GPS or an IMU and a second difference between the rotational velocity component and a rotational velocity obtained from the GPS or the IMU;

removing any of the retained clusters in which the first difference is lower than a first threshold or the second difference is lower than a second threshold;

combining the unremoved clusters;

refitting the signal model to the combined clusters; and

determining V L , V R , β, and γ.

7. The system of claim 6 , wherein the determining the one or more calibration parameters further comprises:

obtaining the Doppler velocity, for a plurality of frames, from the each of the entities while the vehicle is driving; and

determining the one or more calibration parameters based on the Doppler velocity from at least the portion of the entities, for the plurality of frames.

8. The system of claim 7 , wherein the determining the one or more calibration parameters further comprises:

obtaining values of V L , V R , β, and γ for each of the plurality of frames;

comparing the values of V L , V R , β, and γ, for each of the respective entities, between each of the plurality of frames;

determining whether any of the values of V L , V R , β, and γ of any of the plurality of frames differs from a mean value or a median value of any of V L , V R , β, and γ across the plurality of frames by more than a third threshold;

removing any of the frames for which any of the values differ from the mean value or the median value by more than the third threshold; and

determining the one or more calibration parameters based on the unremoved frames.

9. A method implemented by a system on a moving vehicle including an active Doppler sensor, one or more processors, and a storage media storing machine-readable instructions, the method comprising:

obtaining a Doppler velocity from each of one or more entities using the active Doppler sensor, wherein the Doppler velocity is with respect to the moving vehicle;

determining one or more calibration parameters of the active Doppler sensor based on respective Doppler velocities from at least a portion of the entities, wherein the determining of the one or more calibration parameters comprises:

determining one or more stationary entities, within the one or more entities;

forming one or more clusters, each of the clusters comprising at least a portion of the one or more stationary entities and defining a relationship between a location of one of the entities with respect to the active Doppler sensor and the Doppler velocity of the respective one of the entities;

fitting, to each of the one or more clusters, a signal model defining the Doppler velocities of the each of the entities in relation to a mounting offset angle or an azimuth offset angle of the active Doppler sensor and a mounting location of the active Doppler sensor;

determining a degree of fit of the signal model to each of the one or more clusters;

removing any of the clusters having a degree of fit lower than a threshold while retaining remaining clusters; and determining the one or more calibration parameters based on the retained remaining clusters;

determining a location of the vehicle based on one or more of the calibration parameters; and

causing the vehicle to perform a driving action based on the determined location.

10. The method of claim 9 , wherein the determining the one or more calibration parameters further comprises:

determining the mounting location, the mounting offset angle, a linear velocity component, and a rotational velocity component of the active Doppler sensor.

11. The method of claim 9 , wherein the determining the location of the vehicle comprises determining the location of the vehicle by integrating the determined linear velocity component over time or the determined rotational velocity component over time.

12. The method of claim 9 , further comprising:

comparing the determined location with a location obtained from a GPS or an IMU.

13. The method of claim 9 , wherein:

the signal model is defined by V D =V L cos (Θ+β)+V R sin (Θ−β−γ), and wherein:

the calibration parameters comprise V L , V R , β, and γ;

V D indicates the Doppler velocity of the each of the entities;

V L indicates a linear velocity component of the each of the entities in a forward direction of the vehicle;

V R indicates a rotational velocity component of the each of the entities;

Θ indicates an angle defining the location of the each of the entities with respect to the active Doppler sensor;

β is the mounting offset angle of the active Doppler sensor; and

γ is defined as γ=β−arctan (y/x), wherein y and x define the mounting location of the active Doppler sensor with respect to an ego coordinate system of the vehicle.

14. The method of claim 13 , wherein the determining the one or more calibration parameters further comprises:

determining, for each of the retained clusters, a first difference between the linear velocity component and a linear velocity obtained from a GPS or an IMU and a second difference between the rotational velocity component and a rotational velocity obtained from the GPS or the IMU;

removing any of the retained clusters in which the first difference is lower than a first threshold or the second difference is lower than a second threshold;

combining the unremoved clusters;

refitting the signal model to the combined clusters; and

determining V L , V R , β, and γ.

15. A non-transitory computer readable medium comprising instructions that, when executed by one or more processors on a moving vehicle, cause the one or more processors to perform:

obtaining a Doppler velocity from each of one or more entities using an active Doppler sensor on the moving vehicle, wherein the Doppler velocity is with respect to the moving vehicle;

determining one or more calibration parameters of the active Doppler sensor based on respective Doppler velocities from at least a portion of the entities, wherein the determining of the one or more calibration parameters comprises:

determining one or more stationary entities, within the one or more entities;

forming one or more clusters, each of the clusters comprising at least a portion of the one or more stationary entities and defining a relationship between a location of one of the entities with respect to the active Doppler sensor and the Doppler velocity of the respective one of the entities;

fitting, to each of the one or more clusters, a signal model defining the Doppler velocities of the each of the entities in relation to a mounting offset angle or an azimuth offset angle of the active Doppler sensor and a mounting location of the active Doppler sensor;

determining a degree of fit of the signal model to each of the one or more clusters;

removing any of the clusters having a degree of fit lower than a threshold while retaining remaining clusters; and determining the one or more calibration parameters based on the retained remaining clusters;

determining a location of the vehicle based on one or more of the calibration parameters; and

causing the vehicle to perform a driving action based on the determined location.