Automotive driver assistance systems

Abstract

To provide a driver assistance system with improved usability. [Solution] A driver assistance system for an automobile, comprising a processor system (14) having an input of a sensor signal (S), an output of an actuator command (A), and a memory device (18) storing a set of rules that determine which actuator command should be output in a particular constellation of sensor signals, wherein the driver assistance system is further characterized by a machine learning module (20) configured to learn at least one of the rules based on observed driver behavior.

Claims

1. A driver assistance system for an automobile, comprising a processor system (14) having an input of a sensor signal (S), an output of an actuator command (A), and a memory device (18) storing a set of rules that determine which actuator command should be output in a particular constellation of sensor signals, wherein the driver assistance system is further characterized by a machine learning module (20) configured to learn at least one of the rules based on observed driver behavior.

2. The driver assistance system according to claim 1, wherein a rule-based learning algorithm is implemented in the machine learning module (20).

3. The driver assistance system according to claim 2, wherein, for any type of driving operation in which the learning of the aforementioned rules is expected, a plurality of mutually interchangeable and distinct rules are stored, and the machine learning module (20) selects the rule having the greatest similarity to the observed driver behavior for the execution of the assistance function.

4. The driver assistance system according to claim 3, wherein the driving operation is to adapt the speed of the vehicle to the speed of a preceding vehicle that is slower, and each of the selectable rules is characterized by a different deceleration profile (22, 24, 26).

5. The driver assistance system according to claim 3, wherein the driving operation is following a preceding vehicle at a certain target distance, and each of the selectable rules is characterized by a different dependence of the target distance on the speed.

6. The driver assistance system according to any one of claims 3 to 5, wherein the machine learning module (20) is configured to learn different preferences for rules selected according to different constellations of influencing factors.

7. The driver assistance system according to claim 6, wherein each of the aforementioned influence factors is characterized by a finite number of attributes.

Images