Back to results
US 2025/0028041 A1Published application
Systems and methods for online calibration in distributed aperture radar
AssigneeGM Cruise Holdings LLC
InventorsMichael MEYER, Jessica Bartholdy SANSON, Marc UNZUETA CANALS, Kalin Hristov KABAKCHIEV
Life of this patent
Estimating the term…
Abstract
Technologies described herein relate to a distributed aperture radar (DAR) system that includes multiple radar sensors. A graph neural network (GNN) is employed to cause radar data output by the multiple radar sensors to correspond to a same coordinate system.
Claims20
A method performed by a radar system, the method comprising:
- receiving a primary reflected radar signal at a primary radar sensor of a distributed aperture radar system, the primary radar sensor having a primary field of view;
- generating primary radar data with the primary radar sensor, the primary radar data comprising a first graph of detections corresponding to a target object that is at least partially within the primary field of view;
- receiving a secondary reflected radar signal at a secondary radar sensor of the distributed aperture radar system, the secondary radar sensor having a secondary field of view;
- generating secondary radar data with the secondary radar sensor, the secondary radar data comprising a second graph of detections corresponding to the target object that is at least partially within the secondary field of view;
- calibrating the secondary radar data using a graph neural network to compensate for movement of the secondary radar sensor relative to the primary radar sensor;
- and
- associating the first graph of detections with the second graph of detections based upon the calibrating of the secondary radar data.
4 more elements
A distributed aperture radar (DAR) system comprising:
- a radar array comprising:
- a primary radar sensor having a primary field of view;
- a secondary radar sensor having a secondary field of view, the secondary field of view overlapping at least a portion of the primary field of view;
- wherein the primary radar sensor generates primary radar data comprising a graph of detections and raw radar data corresponding to the primary field of view and including a target object;
- and
- wherein the secondary radar sensor generates secondary radar data comprising a graph of detections corresponding to the secondary field of view and including the target object;
- a computer comprising a processor and memory, where the computer processor is configured to:
- store the primary and secondary radar data in the memory;
- provide the primary radar data and the secondary radar data as input to a graph neural network, wherein the graph neural network is stored in the memory and is executed by the processor, and further where the graph neural network outputs a transformation matrix based upon the radar data;
- and
- apply the transformation matrix to the secondary radar data such that the secondary radar data is transformed to share a coordinate system with the primary radar data.
8 more elements
An autonomous vehicle (AV) comprising:
- a propulsion system;
- a steering system;
- a braking system;
- a distributed aperture radar (DAR) system that includes a first radar sensor and a second radar sensor, where the first radar sensor generates first radar data and the second radar sensor generates second radar data;
- and
- a computing system in communication with the propulsion system, the steering system, the braking system, and the DAR system, where the computing system is programmed to perform acts, the acts comprising:
- receiving the first radar data and the second radar data;
- using a graph neural network (GNN), computing a transformation matrix based upon the first radar data and the second radar data;
- using the transformation matrix, transforming the second radar data such that the second radar data is in a coordinate system of the first radar data;
- subsequent to transforming the second radar data, computing at least one of location of an object relative to the first radar sensor or velocity of the object relative to the first radar sensor based upon the transformed second radar data;
- and
- controlling at least one of the propulsion system, the steering system, or the braking system based upon the at least one of the location of the object or the velocity of the object.
9 more elements
Dependencies are inferred from the claim text.