Inductive displacement measuring sensor and method for operating the latter

• US 10,209,097 B2
• Filed: 03/19/2013
• Issued: 02/19/2019
• Est. Priority Date: 03/19/2013
• Status: Active Grant

• Current Assignee: Balluff Verwaltungsges mbH
• Original Assignee: Balluff Verwaltungsges mbH
• Inventors: Polik, Zoltan, Kantor, Zoltan
• Primary Examiner(s): Nguyen, Tan T.
• Application Number:
  US14/778,219
• Publication Number:
  US 20160076912A1
• Time in Patent Office: 2,163 Days
• Field of Search: 702 66, 702170
• CPC Class Codes: G01D 5/20 by varying inductance, e.g. by a movable armature
• CPC Class Codes: G01D 5/2053 by a movable non-ferromagnetic conductive element
• CPC Class Codes: G06N 3/04 Architecture, e.g. interconnection topology

Abstract

In an inductive displacement measuring sensor for measuring the distance, the spatial orientation, the thickness, the material properties or the like of a target object, which sensor has a transmitter element which emits a pulsed signal and a receiver element for detecting a pulse response caused by the emitted pulsed signal in the target object, provision is made, in particular, for the detected pulse response to be immediately evaluated using an artificial neural network.

Claims

1. Inductive displacement measuring sensor for measuring distance, spatial orientation, thickness, or material properties of a target object, the sensor comprising:

a waveform generator configured to emit a non-periodic, current signal,

a transmitter element configured to emit a non-periodic, pulsed signal and configured to be stimulated by the non-periodic, current signal emitted by the waveform generator,

a receiver element for detecting a pulse response in the target object caused by the emitted non-periodic, pulsed signal,

a timing unit configured to supply trigger signals,

a signal digitalization unit configured to;

digitize the detected pulse response such that a digitized waveform is formed and

supply the digitized waveform to a microcontroller, and

the microcontroller, the microcontroller comprising a signal evaluation unit, the signal evaluation unit comprising an artificial neural network, the artificial neural network being configured to evaluate the digitized waveform without intermediate spectral analysis,

wherein the waveform generator and the signal digitization unit are triggered by the trigger signals supplied by the timing unit.

2. Displacement measuring sensor according to claim 1, wherein the signal evaluation unit further comprises a downsampler to reduce the number of samples.

3. Displacement measuring sensor according to claim 2, wherein the downsampler is configured to generate input data for the artificial neural network with unequal sampling intervals.

4. Displacement measuring sensor according to claim 1, wherein the artificial neural network comprises a hidden layer and fewer than ten nodes and fewer than twelve neurons arranged in the hidden layer.

5. Displacement measuring sensor according to claim 4, wherein the artificial neural network has five nodes and six neurons arranged in the hidden layer.

6. Displacement measuring sensor according to claim 1, wherein the artificial neural network supplies results which are independent of the material of the target object as well as of the ambient temperature and/or the temperature of the displacement measuring sensor.

7. Displacement measuring sensor according to claim 1, wherein the artificial neural network supplies results which are dependent on the material of the target object.

8. Displacement measuring sensor according to claim 1, wherein the output of the sensor is formed by a linear function of the target object distance.

9. Method for operating a displacement measuring sensor, the method comprising steps of:

emitting a non-periodic, current signal via a waveform generator of the sensor,

stimulating a transmitter element of the sensor via the non-periodic, current signal such that the transmitter signal emits a non-periodic, pulsed signal,

detecting a pulse response in a target object via a receiver element of the sensor, the pulse response being caused by the emitted non-periodic pulsed signal,

digitizing via a signal digitalization unit the detected pulse response such that a digitized waveform is formed,

supplying the digitized waveform to an artificial neural network of a signal evaluation unit of a microcontroller of the sensor,

evaluating the digitized waveform without intermediate spectral analysis and via the artificial neural network, and

if a result supplied by the artificial neural network exceeds an upper threshold value, switching an output of the sensor into a first state or

if a result supplied by the artificial neural network falls below a lower threshold value, switching the output of the sensor into a second state,

wherein the waveform generator and the signal digitization unit are triggered by trigger signals supplied by a timing unit.

10. Method according to claim 9, wherein the evaluation of the input data occurs via a hysteresis comparator or a step function in the artificial neural network.