Behavior model of an environment sensor
AssigneeAVL List GmbH
InventorsKmeid Saad, Michael Paulweber
Life of this patent
Estimating the term…
Abstract
A computer-aided method for training an artificial neuronal network for the simulation of an environment sensor of a driver assistance system includes the following work steps: reading in traffic scenarios of a test journey; deriving test journey data and the sensor data to be output by the environment sensor from the traffic scenarios; generating a data stream which depicts the traffic scenarios from the perspective of the environment sensor; outputting the data stream such that the environment sensor can generate sensor data on the basis of the data stream and can provide same to a data interface at which the test journey data and the sensor data to be output are also provided; and reading the provided data into the artificial neuronal network.
Claims12
A computer-aided method for training an artificial neuronal network for simulating the behavior of an environment sensor of a driver assistance system in traffic scenarios, comprising:
- importing the traffic scenarios of a road test;
- deriving road test data from the traffic scenarios;
- providing the road test data to a data interface;
- deriving ground truth sensor data for the environment sensor from the traffic scenarios;
- providing the ground truth sensor data to the data interface;
- generating a data stream, which maps the traffic scenarios from a perspective of the environment sensor;
- outputting the data stream to the environment sensor such that the environment sensor can generate sensor data on the basis of the data stream and the environment sensor can provide the generated sensor data to the data interface;
- and
- importing the generated sensor data, the road test data, and the ground truth sensor data into the artificial neuronal network.
A computer-aided method for analyzing an environment sensor of a driver assistance system, comprising:
- importing traffic scenarios from a plurality of road tests;
- deriving road test data from the traffic scenarios;
- deriving ground truth sensor data for the environment sensor from the traffic scenarios;
- generating a data stream, which maps the traffic scenarios from a perspective of the environment sensor;
- outputting the data stream to the environment sensor which generates sensor data on the basis of the data stream;
- performing a curve fitting via an artificial neuronal network on the basis of the road test data, the ground truth sensor data, and the sensor data generated by the environment sensor;
- and
- applying a simulation model based on the curve fitting, wherein the simulation model is configured to simulate and output sensor data on the basis of a data stream from any given road test.
A computer-aided method for simulating the behavior of an environment sensor of a driver assistance system in traffic scenarios, comprising:
- deriving ground truth sensor data for the environment sensor from the traffic scenarios;
- capturing a data stream, which maps a road test from a perspective of the environment sensor;
- determining sensor data using a simulation model of the environment sensor based on the captured data, wherein the simulation model is configured to simulate the sensor data on the basis of a data stream of any given road test, and wherein the simulation model is based on a curve fitting via an artificial neuronal network on the basis of road test data, data streams from a plurality of different road tests, the ground truth sensor data, and the sensor data generated by the environment sensor during said road tests;
- and
- outputting the sensor data determined via the simulation model.
A system for training an artificial neuronal network for simulating an environment sensor of a driver assistance system, comprising:
- an interface for importing traffic scenarios of a road test;
- means for deriving road test data from the traffic scenarios;
- means for deriving ground truth sensor data for the environment sensor from the traffic scenarios;
- means for generating a data stream, which maps the traffic scenarios from a perspective of the environment sensor;
- an interface for outputting the data stream such that the environment sensor can generate sensor data on the basis of the data stream;
- and
- a data interface at which the generated sensor data, the road test data, and the ground truth sensor data can be provided to the artificial neuronal network.
A system for analyzing an environment sensor of a driver assistance system, comprising:
- an interface for importing traffic scenarios from a plurality of road tests;
- means for deriving road test data from the traffic scenarios;
- means for deriving ground truth sensor data for the environment sensor from the traffic scenarios;
- means for generating a data stream, which maps the traffic scenarios from a perspective of the environment sensor;
- an interface for outputting the data stream to the environment sensor which generates sensor data on the basis of the data stream;
- means for performing a curve fitting via an artificial neuronal network on the basis of the road test data, the ground truth sensor data, and the sensor data generated by the environment sensor;
- means for determining sensor data using a simulation model based on the curve fitting, wherein the simulation model is configured to simulate sensor data on the basis of a data stream of any given road test;
- and
- a data interface for outputting the sensor data determined via the simulation model.
A system for simulating an environment sensor of a driver assistance system, comprising:
- means for deriving ground truth sensor data for the environment sensor from traffic scenarios;
- an interface for capturing a data stream which maps a road test from a perspective of the environment sensor;
- means for determining sensor data using a simulation model configured to simulate sensor data on the basis of a data stream of any given road test, wherein the simulation model is based on a curve fitting via an artificial neuronal network on the basis of road test data, data streams from a plurality of different road tests, the ground truth sensor data, and sensor data generated by the environment sensor during said road tests;
- and
- a data interface for outputting the sensor data determined via the simulation model.
Dependencies are inferred from the claim text.