Encoder device and method of determining a kinematic value

1. An encoder device for determining a kinematic value of the movement of a first object relative to a second object, wherein the encoder device comprises:

a standard associated with the first object;

at least one scanning unit associated with the second object for producing at least one scanning signal by detection of the standard; and

a control and evaluation unit that is configured to determine the kinematic value from the at least one scanning signal,

wherein the control and evaluation unit is further configured to determine the kinematic value by an evaluation of the at least one scanning signal using a method of machine learning, with the method of machine learning being pre-trained with a plurality of scanning signals and associated kinematic values.

2. The encoder device in accordance with claim 1 ,

wherein the control and evaluation unit has a deep neural network.

3. The encoder device in accordance with claim 2 ,

wherein an architecture of the deep neural network has an at least partially predefined architecture.

4. The encoder device in accordance with claim 3 ,

wherein the architecture of the deep neural network has a predefined number of layers and/or neurons per layer.

5. The encoder device in accordance with claim 1 ,

wherein the at least one scanning unit comprises a plurality of scanning units using different sensor principles.

6. The encoder device in accordance with claim 1 ,

wherein the at least one scanning unit is configured for a sensor principle that directly produces a piece of speed and/or acceleration information.

7. The encoder device in accordance with claim 1 ,

wherein the standard is configured for the generation of a non-periodic pattern.

8. The encoder device in accordance with claim 1 ,

wherein the kinematic value comprises one of a rotary position and a translatory offset of the first and second objects from one another.

9. The encoder device in accordance with claim 1 ,

wherein the kinematic value comprises at least one of a speed and an acceleration.

10. The encoder device in accordance with claim 1 ,

wherein the control and evaluation unit is configured for an advance determination of a rough estimate of the kinematic value.

11. The encoder device in accordance with claim 1 ,

wherein the control and evaluation unit is configured for an advance determination of a rough estimate of the kinematic value using a method without machine learning.

12. The encoder device in accordance with claim 1 ,

wherein the control and evaluation unit is configured for a training phase in which the encoder device is exposed to different known movement scenarios with known location, speed, and/or acceleration profiles.

13. The encoder device in accordance with claim 1 ,

wherein the evaluation is trained while varying environmental conditions and/or mechanical influences.

14. The encoder device in accordance with claim 1 ,

wherein the evaluation is trained while varying at least one of temperature, humidity, shock load, and vibration.

15. The encoder device in accordance with claim 1 ,

wherein the movement is a rotary movement and wherein the evaluation is trained while varying eccentricity, radial runout, and/or different rotary supports.

16. The encoder device in accordance with claim 1 ,

wherein a partially trained evaluation that is trained for a class of encoder devices is specified for the control and evaluation unit.

17. The encoder device in accordance with claim 16 ,

wherein the control and evaluation unit is configured to subsequently train the partially trained evaluation individually.

18. A method of determining a kinematic value of the movement of a first object relative to a second object, wherein a standard is associated with the first object and at least one scanning unit is associated with the second object, the method comprising:

producing at least one scanning signal by detection of the standard using the at least one scanning unit; and

evaluating the at least one scanning signal to determine the kinematic value therefrom, wherein the kinematic value is determined by an evaluation of the at least one scanning signal using a method of machine learning, with the method of machine learning being pre-trained with a plurality of scanning signals and associated kinematic values.