DriveWorks SDK Reference 3.0.4260 Release For Test and Development only
Note
SW Release Applicability: This tutorial is applicable to modules in both NVIDIA DriveWorks and NVIDIA DRIVE Software releases.

## Radar Calibration - Operating Principle

Radar calibration estimates the current sensor orientation (yaw angle) with respect to the vehicle's coordinate frame. This yaw angle is estimated by first using the Doppler signal of radar detections to estimate the radar's motion. Subsequently, the radar motion is matched with egomotion estimates to estimate the yaw angle. Calibration measurements require a sufficiently large speed of driving.

The Doppler motion observed by a radar can also be used to calibrate odometry speed factors, i.e., a factor which maps measured longitudinal speed to the actual driven speed. In addition the method can be used to calibration individual wheel radii by looking for the radius which maps wheel's rotational velocity to the longitudinal velocity measured by radar. If requested, this calibration will be performed during straight driving maneuvers.

Radar yaw (top) and wheel radii / velocity speed factor estimation histograms (bottom), collected over a period of time

## Requirements

### Initialization Requirements

• Nominal values on radar calibration
• Orientation(roll/pitch/yaw): less than 10 degree error
• Position(x/y/z): x and y are not used for now, z is less than 10 cm error

### Runtime Calibration Dependencies

• Yaw calibration: IMU-based egomotion needs to be based on accurate IMU calibration to provide accurate yaw-rates

### Input Requirements

• Sensors: in order to achieve good performance, radar calibration requires data from IMU sensors.
• Assumption: Vehicle performs normal driving maneuvers until calibration convergence.
• Egomotion: to perform calibration of wheel radii egomotion used in dwCalibrationEngine has to be fed with dwVehicleIOState

### Output Requirements

• Corrected yaw value: less than 0.16 degrees
• Time to correction: less than 2 minutes for radar sensor with 15HZ spinning frequency

## Cross-validation KPI

Several hours of data are used to produce a reference calibration value for cross-validation. Then, short periods of data are evaluated for whether they can recover the same values. For example, the graph below shows precision/recall curves of radar self-calibration. Precision indicates that an accepted calibration is within a fixed precision threshold from the reference calibration, and recall indicates the ratio of accepted calibrations in the given amount of time.

## Workflow

The following code snippet shows the general structure of a program that performs Radar self-calibration

dwCalibrationEngine_initialize(...); // depends on sensor from rig configuration module
dwCalibrationEngine_initializeRadar(...); // depends on nominal calibration from rig configuration
dwCalibrationEngine_startCalibration(...); // runtime calibration dependencies need to be met
while (true) // main loop
{
// code to get IMU measurement
// code to get CAN measurement
// code to feed IMU and CAN measurements to vehicle egomotion
// code to get radar motion
// feed lidar sweep into self-calibration
// retrieve calibration status
// retrieve self-calibrated result
{