WO2014146623A1 - Induktiver wegmesssensor und verfahren zu seinem betrieb - Google Patents
Induktiver wegmesssensor und verfahren zu seinem betrieb Download PDFInfo
- Publication number
- WO2014146623A1 WO2014146623A1 PCT/DE2013/000155 DE2013000155W WO2014146623A1 WO 2014146623 A1 WO2014146623 A1 WO 2014146623A1 DE 2013000155 W DE2013000155 W DE 2013000155W WO 2014146623 A1 WO2014146623 A1 WO 2014146623A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- sensor
- distance
- neural network
- artificial neural
- output
- Prior art date
Links
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01D—MEASURING NOT SPECIALLY ADAPTED FOR A SPECIFIC VARIABLE; ARRANGEMENTS FOR MEASURING TWO OR MORE VARIABLES NOT COVERED IN A SINGLE OTHER SUBCLASS; TARIFF METERING APPARATUS; MEASURING OR TESTING NOT OTHERWISE PROVIDED FOR
- G01D5/00—Mechanical means for transferring the output of a sensing member; Means for converting the output of a sensing member to another variable where the form or nature of the sensing member does not constrain the means for converting; Transducers not specially adapted for a specific variable
- G01D5/12—Mechanical means for transferring the output of a sensing member; Means for converting the output of a sensing member to another variable where the form or nature of the sensing member does not constrain the means for converting; Transducers not specially adapted for a specific variable using electric or magnetic means
- G01D5/14—Mechanical means for transferring the output of a sensing member; Means for converting the output of a sensing member to another variable where the form or nature of the sensing member does not constrain the means for converting; Transducers not specially adapted for a specific variable using electric or magnetic means influencing the magnitude of a current or voltage
- G01D5/20—Mechanical means for transferring the output of a sensing member; Means for converting the output of a sensing member to another variable where the form or nature of the sensing member does not constrain the means for converting; Transducers not specially adapted for a specific variable using electric or magnetic means influencing the magnitude of a current or voltage by varying inductance, e.g. by a movable armature
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01D—MEASURING NOT SPECIALLY ADAPTED FOR A SPECIFIC VARIABLE; ARRANGEMENTS FOR MEASURING TWO OR MORE VARIABLES NOT COVERED IN A SINGLE OTHER SUBCLASS; TARIFF METERING APPARATUS; MEASURING OR TESTING NOT OTHERWISE PROVIDED FOR
- G01D5/00—Mechanical means for transferring the output of a sensing member; Means for converting the output of a sensing member to another variable where the form or nature of the sensing member does not constrain the means for converting; Transducers not specially adapted for a specific variable
- G01D5/12—Mechanical means for transferring the output of a sensing member; Means for converting the output of a sensing member to another variable where the form or nature of the sensing member does not constrain the means for converting; Transducers not specially adapted for a specific variable using electric or magnetic means
- G01D5/14—Mechanical means for transferring the output of a sensing member; Means for converting the output of a sensing member to another variable where the form or nature of the sensing member does not constrain the means for converting; Transducers not specially adapted for a specific variable using electric or magnetic means influencing the magnitude of a current or voltage
- G01D5/20—Mechanical means for transferring the output of a sensing member; Means for converting the output of a sensing member to another variable where the form or nature of the sensing member does not constrain the means for converting; Transducers not specially adapted for a specific variable using electric or magnetic means influencing the magnitude of a current or voltage by varying inductance, e.g. by a movable armature
- G01D5/204—Mechanical means for transferring the output of a sensing member; Means for converting the output of a sensing member to another variable where the form or nature of the sensing member does not constrain the means for converting; Transducers not specially adapted for a specific variable using electric or magnetic means influencing the magnitude of a current or voltage by varying inductance, e.g. by a movable armature by influencing the mutual induction between two or more coils
- G01D5/2053—Mechanical means for transferring the output of a sensing member; Means for converting the output of a sensing member to another variable where the form or nature of the sensing member does not constrain the means for converting; Transducers not specially adapted for a specific variable using electric or magnetic means influencing the magnitude of a current or voltage by varying inductance, e.g. by a movable armature by influencing the mutual induction between two or more coils by a movable non-ferromagnetic conductive element
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
Definitions
- the invention is based on an inductive displacement sensor and a method for its operation according to the preamble of the independent claims.
- Evaluation unit are provided, by means of which the detection, processing and
- the KNN described there comprises an input layer, at least one (hidden) intermediate layer, an output layer, and weights provided at the junctions between two individual layers, respectively. Suitable values for the respective weighting factors are determined in a learning phase in which test measurements are carried out on a number of different target objects of known material and on a known distance from the sensor.
- the sensor arrangement should be suitable for both distances and thicknesses irrespective of the material of the respective target object
- spectral analysis Subjected to spectral analysis.
- the basis for this is the dependence of the measured Spectrum of the spatial distance to the target object.
- numerical calculations are performed on the measured time-varying magnitude of electrical voltage and current, which means considerable computational effort and a compact design and low-cost implementation of the sensor prevents.
- the invention has for its object to provide an inductive displacement sensor of the type mentioned, which eliminates the disadvantages of the prior art.
- the invention is based on the finding that a named spectral analysis is not required for a displacement measuring system or a corresponding displacement measuring sensor that is involved here, since the calculated frequency spectra are measured over time
- variable signals from which they are calculated do not contain any additional information.
- KNN-based analysis of the frequency spectra it is not possible to find any spectral or temporal characteristics which are useful for distance measurement.
- the invention proposes that detected by the measuring coil
- an impulse response caused by the target object to be measured is evaluated by the KNN by means of a non-periodic (transient) signal sent by the transmitter coil.
- the impulse response is in the
- the KNN Essentially generated by in the target induced eddy currents and magnetic polarization. As an output signal, the KNN provides distance data of the target object.
- the distance measuring sensor according to the invention has one of the ambient temperature or operating temperature of the sensor and the respective material to be measured
- Target object independent characteristic In this characteristic, result data (eg distance data) supplied by the distance measuring sensor are plotted over the actual, otherwise determined distance.
- the characteristic curve resulting from the displacement sensor according to the invention is preferably linear or at least represents a strictly monotone curve. In the ideal case of a linear progression, the slope of the characteristic essentially corresponds to the value 1. Due to the direct evaluation of the inductively detected measurement signal by means of an ANN, ie without the need for an intermediate spectral analysis, there is a significantly reduced hardware complexity for the sensor, whereby a comparison with the prior art much more compact design is possible. For example, this makes it possible to integrate the KNN and the additionally required logic in a microcontroller, which can also significantly reduce the manufacturing costs.
- the inductive displacement sensor according to the invention is suitable for determining the
- the KNN can be configured or taught in accordance with the intended use of the sensor so that instead of the material-independent measurement mentioned, a material-specific measurement is also made possible.
- the displacement sensor according to the invention can only be used with inductively operating displacement measuring systems and, in principle, can also be used with non-inductively operating displacement measuring sensors, in which a corresponding impulse response is evaluated, with the advantages described herein.
- non-inductively operating displacement measuring sensors are optical, acoustic (ultrasonic) or electrically capacitive sensors.
- FIG. 1 shows an exemplary embodiment of a distance measuring sensor according to the invention
- FIG. 2 shows an exemplary embodiment of an artificial neural network (KNN);
- Figure 3 illustrates the typical temporal relationship between a
- FIG. 4 shows an exemplary embodiment of the method according to the invention for
- Figure 5 shows an embodiment of the inventive method for
- FIG. 6 shows a characteristic diagram for the comparison of ANNs calculated
- the inductive distance sensor 1 shown in FIG. 1 comprises an (analog) measuring sensor or measuring transducer 10, a time control unit 20, a digitizing unit 30 for digitizing the detected signal, a signal evaluation unit 40 and a
- Output unit 50 The said functional components are arranged in the embodiment in a single housing 5.
- Digitizing unit 30 and the signal evaluation unit 40 are implemented in particular in a microcontroller 4.
- the distance sensor 1 also includes a voltage supply, not shown here.
- the transmitter 10 comprises at least one transmitter coil 11, at least one
- Receiver coil 12 an operational amplifier 13 for the transmitter coil 11, a triggered waveform generator 14 and connected to the receiver coil 12, the respective applied signal processing operational amplifier 15.
- the transmitter coil 11 and the receiver coil 12 represent the primary sensor elements of the distance sensor 1.
- the digitizing unit 30 includes an A / D converter 31 and a memory 32 for storing waveforms.
- the signal evaluation unit 40 comprises a downsampler 41 for reducing the number of samples as well as an artificial neural network (KNN) 42. This arrangement allows the immediate evaluation of a sample rate downsampled, time-dependent
- FIG. 2 shows the structure of an ANN.
- connecting lines arranged between neurons of an input layer 200 and an intermediate layer 210 are shown by dashed lines only for the purpose of illustration.
- the KNN includes an input layer 200, at least one hidden layer 210, and an output layer 220.
- Input node or input neuron 201-205 of input layer 200 is physically (electrically, optically, etc.) or logically connected to each hidden neuron 211-216 disposed in intermediate layer 210 via predetermined weighting factors 207.
- Each hidden neuron 211-216 disposed in the intermediate layer 210 overlaps with each output neuron 221 disposed in the output layer 220
- predetermined weighting factors 207 connected. Should the ANN have more than one hidden interlayer 210, then all of the input neurons 201-205 are connected to each neuron located in the first interlayer 210 via predetermined weighting factors, with each neuron of a previous hidden interlayer connected to each neuron of the subsequent hidden interlayer via predetermined weighting factors and wherein all neurons of the last hidden intermediate layer are connected to each output neuron of the output layer 220.
- Each neuron carries out, in a manner known per se, a summation of the values provided by the respective preceding layer and each acted upon with given weighting factors and evaluates the resulting sum by means of a neural function.
- the result of this evaluation represents the initial value of the respective neuron.
- known functions such as e.g. the linear function, the sigma function, the hyperbolic tangent ("hyperbolic tangent") or the sign function.
- the single output neuron 221 disposed in the output layer 220 in the present embodiment provides the output values of the entire KNN 42.
- the input layer 200 and the output layer 220 are connected to the surroundings of the KNN 42, in the present embodiment with the downsampler 41 and the output unit 50, whereas the said concealed layers or intermediate layers 210 are not directly accessible from the outside.
- the KNN 42 even with the use of only a few neurons of the input layer 200, in the present embodiment, 5 neurons, and only 6 arranged in the hidden layer 210 neurons, independent of the material of the target object distance data with a maximum error of only 2.65% and a mean error of 0.79%. This roughly corresponds to the accuracy obtained with an KNN topology with 24 input nodes or neurons and 20 hidden neurons.
- the reduced topology can therefore be implemented in a common microcontroller 4, which delivers the results of the aforementioned calculation in a relatively short time. So will be at one with 24 MHz and a 32 bit wide
- Data path powered microcontroller 36 multiplies at 7 functions less than 200 [is required computing time.
- the described distance sensor 1 is suitable for determining the distance, the spatial orientation, the thickness and / or the material properties of a metallic target object 2 to be measured, which is in spatial proximity to the said primary
- Sensor element 1 1, 12 is arranged, and provides at the sensor output 60, a corresponding signal or result data.
- the transmitter coil 1 1 by one of a
- Waveform generator 14 supplied, non-periodic (transient) current signal, wherein the excitation signal is amplified by the operational amplifier 13.
- Waveform generator 14 is triggered by a trigger signal 6 provided by timing unit 20.
- the transmitter coil 1 1 generates a time-varying (i.e., transient) magnetic field in the vicinity of the sensor 1. Due to the temporal change behavior and the inhomogeneity of this magnetic field, a voltage is induced in the at least one receiver coil 12. In the case where a target metallic object 2 is placed near the primary sensor element, the changing magnetic field generates
- Transmitter coil 1 1 in the target object 2 eddy currents and causes there a magnetic
- the target object 2 is magnetically polarized, which in turn reacts on the magnetic field generated by the transmitter coil 1 1 and this modified, whereby the time course of the induced voltage in the receiver coil 12 changes accordingly.
- the induced voltage in the receiver coil 12 is determined by means of
- the AD converter 31 Parallel to the delivery of the trigger signal 6, the AD converter 31 performs a periodic
- the waveform 9 is obtained by means of the step down counter 41 reduced in the sampling rate and fed to the KNN 42 as input data 70.
- the KNN 42 generates output data 80 from the input data 70.
- Sensor output data 60 are generated from the output data 80 by means of an output unit 50.
- Data acquisition based on equal sampling intervals and subsequent downsampling can also be detected on the basis of unequal sampling intervals.
- Temperature changes cause characteristic changes in the waveform and thus also corresponding changes in the measured waveform.
- temperature-dependent waveforms are used as input data as well
- the KNN 42 delivers temperature-independent under normal operating conditions
- the temperature-dependent waveform in particular at medium or larger target object distances, can be regarded as an additive component of the digitized waveform 9. Therefore, with known temperature dependence of the sensor system, training data for different temperatures and different target materials, target distances, etc., can be generated by extrapolation of actually measured data, even if these data are among the
- Temperature conditions are detected in the manufacture of the sensors.
- the measurement signal 8 is not only the result of the currents induced in the target object 2 and the magnetic polarization occurring there, but also of structural features of the distance sensor 1 and the environment 3 of the sensor 1.
- the metal influences the measurement results , Therefore, the time profile of the measurement signal 8 is influenced or determined as a whole by the electrical conductivity as well as the magnetic permeability of the target object, the structural features of the sensor 1 and its surroundings 3, as well as their size, shape, position and orientation. Therefore, the input data 70 is in any case sufficient to generate the said output data.
- the abovementioned downsampling is therefore carried out in order to generate the smallest possible subset of the digitized waveform 9, which is sufficient to be able to calculate the result data or result signals present at the output 60 of the sensor 1 with predetermined accuracy Decreasing the sample rate thus provides a beneficial reduction in the hardware resources required to implement the KNN 42, by reducing the
- excitation pulse It is therefore advantageous to sample the measurement signal in an odd-numbered or low-sampling rate after rapid transient detection or recording has already taken place.
- the input data 70 may also be composite data consisting of a subset of the digitized converted waveform 9 and the digitized values of measured disturbance parameters, e.g. the temperature, are formed. This ensures trouble-free sensor operation. Physical sizes like that
- Target object distance and physical disturbances such as the temperature, such as existing external magnetic fields, etc. influence the measuring signal 8 in a certain way. Therefore, at the input of the KNN 42 present information on these physical quantities can be used to decouple the measurement signal from the said disturbing effects.
- the output unit 50 converts the output data 80 into sensor output data.
- each output date of the sensor 1 is the result of a simple mathematical calculation or evaluation of a given threshold using logic operations, the respective output signal applied to the output 60 being physically generated by a voltage or current source, and an output 60 attached signal additionally by means of a
- the Power amplifier can be amplified.
- the sensor output signal is also a data format or conventional signal type common in industrial distance sensors and proximity switches.
- the sensor output may e.g. one
- the KNN 42 has at least one output layer 220 through which the the
- the KNN is able to provide correct output data 80 to the measured input data 70.
- the named weightings of KNN 42 are iteratively adapted to available input data 70 and the expected output data 80 in a manner known per se.
- the resulting weighting values are stored in the system memory of the microcontroller 4.
- the KNN-based signal evaluation unit 41 carries out the following work steps for training purposes.
- the declaration of the structure and the dimensions of the KNN 42 takes place in step 405.
- the following steps 410 and 415 represent the actual training phase 420.
- the collection of data records is carried out to obtain a knowledge-based database generate, by means of which the KNN 42 is to be trained.
- the KNN 42 is to be trained.
- a suitable training method is e.g. the Levenberg-Marquardt algorithm, which is known per se, in which the weighting factors are changed both between the input layer and the associated neurons as well as between neurons and associated output layer until correct output data (ie distance values or the like) is obtained for each input vector (ie measurement signal). result. Since the values of such an input vector are dependent on the material of the target object, the KNN 42 can learn to calculate the correct distance or the like for any material.
- step 425 a test of the sensor functions is then performed with the appropriately trained KNN 42. If necessary, after checking the accuracy (step 430) of the resulting sensor data, a repetition of said steps occurs (conditional return 432 to step 415) the structure and the topology of the KNN 42 can be optimized in order to achieve the required accuracy In addition, a dashed line, optional return path 434 is provided, which is used, for example, when the required accuracy is not reached several times and therefore the conclusion is drawn may that the training data itself is faulty or are insufficient. If the accuracy resulting from test 430 is sufficient, the routine is ended at step 435.
- the actual training of KNN 42 begins with the compilation of a training record, in the form of a field (arrays) of training-dedicated input data sets 70 for KNN 42.
- These input data sets 70 comprise sampled-rate digitized signal waveforms for the transmitter 10 and one on the temperature-related size.
- the temperature-related variable can be measured in a conventional manner by means of a heat-dependent resistor (thermistor).
- the aforementioned training data include a plurality of target object distances, target material, target object shapes, and target orientations of the target
- Target as well as data measured for different temperatures.
- the training data are generally generated under fixed temperature conditions prevailing in the manufacture of the sensors, so that the corresponding signal waveforms for different temperatures still have to be obtained by extrapolation. Extrapolation can be based on empirically obtained temperature behavior.
- the measuring conditions such. Material properties, geometry and distance of the target object 2 as well as optionally the
- Material properties and the geometry of the environment 3 considered numerical data. These measurements are performed repeatedly, but avoiding the over-training of CNN 42.
- the learning ability of KNN 42 can be increased by providing some of the data available for training
- said characteristic curve of the sensor preferably represents a linear or at least strictly monotonically variable curve as a function of the target distance, which enables a very effective and relatively simple evaluation of the result data.
- the characteristic curve is independent of the ambient or operating temperature and the respective material of the target object.
- the output signal of the sensor represents a conventional voltage signal already mentioned above for distance sensors used in industry.
- the output signal may also be formed by an analog current or a digital signal.
- the operation of the sensor 1 can be preset very varied by means of training, so that very different applications are possible.
- the sensor can be trained to provide an output signal specific to the particular material of the target object and / or to discriminate between ferromagnetic and non-ferromagnetic material of the target object.
- the sensor 1 may be trained as a metal detector or coin sorter, or as a sensor for measuring the thickness of the target object or the like.
- inventive sensor 1 is only preferred and the sensor can also be realized with the function of a position sensor or proximity switch. in case of a
- the senor can also be trained so that it works either material-independently or material-selectively, either material-dependent detected or the detection material can be denied dependent.
- the input data 70 of the KNN 42 can be evaluated in at least two neurons of the output layer 220 either by a numerical hysteresis comparator or by encoding a step function as an evaluation function.
- the topology shown in FIG. 2 is to be modified for this purpose and the only one output neuron 221 is to be modified by at least two
- An example of a suitable topology having three output neurons is the KNN 42 shown in FIG. 1.
- the KNN 42 is trained to switch to one of the outputs of the KNN 42 when the measured distance between the sensor and the target object reaches or exceeds the upper threshold on a given hysteresis curve and that is switched to another output when the measured distance between the sensor and the target reaches or falls below a lower threshold on the hysteresis curve.
- the sensor outputs 60 are in each case updated on the basis of the output data 80 of the output unit 50 ("updated").
- step 500 in FIG. 5 the downstroke 41 initially provides input data to the KNN 42.
- step 505 the evaluation takes place by means of the hysteresis comparator or the mentioned step function or threshold.
- step 510 it is checked whether the value of the measured distance provided by the KNN 42 is the said upper threshold
- step 520 Checks whether the value of the measured distance supplied by the KNN 42 falls below the said lower threshold. If this condition is met, the sensor output is switched to the 'OFF' state in step 525. If neither of these two conditions is met, the sensor output is not changed or switched over.
- the evaluation of a hysteresis curve can be performed by means of a reduced KNN topology in which only a single output binary neuron (i.e., connected to the input layer) is provided.
- FIG. 3 shows the results in the form of a characteristic curve in which distance data supplied by KNN 42 are plotted using reference data measured in a separate way. In this case, distance data obtained at 2, 3, 4 and 5 mm were evaluated.
- the different materials of the target object are copper (Cu), aluminum (AI), V2A steel (V2A) and EC80 mild steel or tempered steel with the steel key EC80 (EC80).
- the KNN 42 was "fed” with the corresponding sensor data at its input layer 200 and recorded at the output 220 of the KNN 42 resulting output data.
- the data shown in Figure 6 illustrates the high quality of the results of KNN 42, i. the resulting distance data are very close to or on a linear curve also drawn there (dashed line). Accordingly, the distance values calculated by the ANN are correct and, in particular, independent of the material of the target object, since all curves (with the sole exception of V2A) are very precisely superimposed.
- the KNN 42 already provides sufficiently precise data even when only a few network nodes or neurons are used (eg 5 as in FIG. 2) and a correspondingly reduced data set to be evaluated, and the CNN 42 is therefore in a conventional microcontroller 4,
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Computational Linguistics (AREA)
- General Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Measurement Of Length, Angles, Or The Like Using Electric Or Magnetic Means (AREA)
Abstract
Description
Claims
Priority Applications (4)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
DE112013006849.4T DE112013006849A5 (de) | 2013-03-19 | 2013-03-19 | Induktiver Wegmesssensor und Verfahren zu seinem Betrieb |
US14/778,219 US10209097B2 (en) | 2013-03-19 | 2013-03-19 | Inductive displacement measuring sensor and method for operating the latter |
CN201380075585.7A CN105122009B (zh) | 2013-03-19 | 2013-03-19 | 电感式位移测量传感器和用于操作所述传感器的方法 |
PCT/DE2013/000155 WO2014146623A1 (de) | 2013-03-19 | 2013-03-19 | Induktiver wegmesssensor und verfahren zu seinem betrieb |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
PCT/DE2013/000155 WO2014146623A1 (de) | 2013-03-19 | 2013-03-19 | Induktiver wegmesssensor und verfahren zu seinem betrieb |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2014146623A1 true WO2014146623A1 (de) | 2014-09-25 |
Family
ID=48170382
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/DE2013/000155 WO2014146623A1 (de) | 2013-03-19 | 2013-03-19 | Induktiver wegmesssensor und verfahren zu seinem betrieb |
Country Status (4)
Country | Link |
---|---|
US (1) | US10209097B2 (de) |
CN (1) | CN105122009B (de) |
DE (1) | DE112013006849A5 (de) |
WO (1) | WO2014146623A1 (de) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2016095882A1 (de) | 2014-12-16 | 2016-06-23 | Balluff Gmbh | Berührungsloser positions-/abstandssensor mit einem künstlichen neuronalen netzwerk und verfahren zu seinem betrieb |
EP3829064A1 (de) * | 2019-11-28 | 2021-06-02 | Sick Ag | Sensoren zum ermitteln eines ausgabewerts, verfahren zum auswerten eines sensorsignals und verfahren zum trainieren einer ausgabeeinheit zum auswerten eines sensorsignals |
EP3968520A1 (de) | 2020-09-10 | 2022-03-16 | Sick Ag | Sensor zur erfassung eines objekts und verfahren zur auswertung eines sensorsignals |
DE102016115015C5 (de) | 2016-08-12 | 2023-01-26 | Sick Ag | Induktiver Näherungssensor |
Families Citing this family (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPWO2019048985A1 (ja) * | 2017-09-06 | 2020-10-29 | 株式会社半導体エネルギー研究所 | 蓄電システム、車両、電子機器及び半導体装置 |
DE102020107229A1 (de) * | 2020-03-17 | 2021-09-23 | Balluff Gmbh | Verfahren zum Betrieb eines absolut messenden Positionserfassungssystems mit einem einspurigen Magnetcodeobjekt |
DE102021000157A1 (de) | 2021-01-15 | 2022-07-21 | Pepperl+Fuchs Se | lnduktive Annäherungssensoreinheit und Verfahren zur Störungsüberprüfung bei einer induktiven Annäherungssensoreinheit |
DE102021000156A1 (de) | 2021-01-15 | 2022-07-21 | Pepperl+Fuchs Se | lnduktive Annäherungssensoreinheit und Verfahren zur Bestimmung einer Objekteigenschaft eines metallischen Erfassungskörpers |
EP4306910A1 (de) * | 2022-07-11 | 2024-01-17 | Melexis Technologies SA | Magnetisches positionssensorsystem, vorrichtung und verfahren |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
DE4201502A1 (de) * | 1992-01-21 | 1993-07-22 | Siemens Ag | Verfahren und anordnung zur elektrischen wirbelstrompruefung |
US5898304A (en) | 1994-09-22 | 1999-04-27 | Micro-Epsilon Messtechnik Gmbh & Co. Kg | Sensor arrangement including a neural network and detection method using same |
US20020059022A1 (en) * | 1997-02-06 | 2002-05-16 | Breed David S. | System for determining the occupancy state of a seat in a vehicle and controlling a component based thereon |
EP2124044A1 (de) * | 2008-05-20 | 2009-11-25 | Siemens Aktiengesellschaft | Verfahren zum Bestimmen und Bewerten von Wirbelstromanzeigen, insbesondere von Rissen, in einem Prüfgegenstand aus einem elektrisch leitfähigen Material |
Family Cites Families (19)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
GB9023909D0 (en) * | 1990-11-02 | 1990-12-12 | Secr Defence | Radar apparatus |
US5835613A (en) * | 1992-05-05 | 1998-11-10 | Automotive Technologies International, Inc. | Optical identification and monitoring system using pattern recognition for use with vehicles |
US7660437B2 (en) * | 1992-05-05 | 2010-02-09 | Automotive Technologies International, Inc. | Neural network systems for vehicles |
US6397136B1 (en) | 1997-02-06 | 2002-05-28 | Automotive Technologies International Inc. | System for determining the occupancy state of a seat in a vehicle |
US5485908A (en) | 1994-07-12 | 1996-01-23 | Coin Acceptors, Inc. | Pattern recognition using artificial neural network for coin validation |
US5742741A (en) * | 1996-07-18 | 1998-04-21 | Industrial Technology Research Institute | Reconfigurable neural network |
GB2366651A (en) * | 2000-09-08 | 2002-03-13 | Ncr Int Inc | Evaluation system |
DE10117218A1 (de) | 2001-04-06 | 2002-10-17 | Daimler Chrysler Ag | Verfahren zur Positions-oder Geschwindigkeitsbestimmung eines Ankers |
US6819790B2 (en) * | 2002-04-12 | 2004-11-16 | The University Of Chicago | Massive training artificial neural network (MTANN) for detecting abnormalities in medical images |
CN2809640Y (zh) | 2005-07-15 | 2006-08-23 | 华南理工大学 | 一种结构缺陷超声在线智能识别系统 |
CN1743839A (zh) | 2005-07-15 | 2006-03-08 | 华南理工大学 | 一种结构缺陷超声在线智能识别系统及识别方法 |
IL172480A (en) * | 2005-12-08 | 2011-11-30 | Amir Zahavi | Method for automatic detection and classification of objects and patterns in low resolution environments |
EP1992939A1 (de) * | 2007-05-16 | 2008-11-19 | National University of Ireland, Galway | Kornbasiertes Verfahren und Vorrichtung für die Klassifizierung von Materialien oder Chemikalien sowie zur Quantifizierung der Eigenschaften von Materialien oder Chemikalien in Mischungen mittels spektroskopischer Daten |
KR100936892B1 (ko) * | 2007-09-13 | 2010-01-14 | 주식회사 엘지화학 | 배터리의 장기 특성 예측 시스템 및 방법 |
CN102050366B (zh) | 2009-11-05 | 2013-02-13 | 上海三菱电梯有限公司 | 人数检测装置及方法 |
CN101706882B (zh) * | 2009-11-23 | 2013-04-03 | 浙江大学 | 基于嵌入式平台的神经网络模型在线训练方法 |
CN102269972B (zh) | 2011-03-29 | 2012-12-19 | 东北大学 | 基于遗传神经网络的管道压力缺失数据补偿方法及装置 |
CN102402835A (zh) | 2011-11-27 | 2012-04-04 | 哈尔滨功成科技创业投资有限公司 | 一种智能家居安全监测系统 |
US9599576B1 (en) * | 2013-03-06 | 2017-03-21 | Nokomis, Inc. | Acoustic—RF multi-sensor material characterization system |
-
2013
- 2013-03-19 DE DE112013006849.4T patent/DE112013006849A5/de active Pending
- 2013-03-19 WO PCT/DE2013/000155 patent/WO2014146623A1/de active Application Filing
- 2013-03-19 CN CN201380075585.7A patent/CN105122009B/zh active Active
- 2013-03-19 US US14/778,219 patent/US10209097B2/en active Active
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
DE4201502A1 (de) * | 1992-01-21 | 1993-07-22 | Siemens Ag | Verfahren und anordnung zur elektrischen wirbelstrompruefung |
US5898304A (en) | 1994-09-22 | 1999-04-27 | Micro-Epsilon Messtechnik Gmbh & Co. Kg | Sensor arrangement including a neural network and detection method using same |
US20020059022A1 (en) * | 1997-02-06 | 2002-05-16 | Breed David S. | System for determining the occupancy state of a seat in a vehicle and controlling a component based thereon |
EP2124044A1 (de) * | 2008-05-20 | 2009-11-25 | Siemens Aktiengesellschaft | Verfahren zum Bestimmen und Bewerten von Wirbelstromanzeigen, insbesondere von Rissen, in einem Prüfgegenstand aus einem elektrisch leitfähigen Material |
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2016095882A1 (de) | 2014-12-16 | 2016-06-23 | Balluff Gmbh | Berührungsloser positions-/abstandssensor mit einem künstlichen neuronalen netzwerk und verfahren zu seinem betrieb |
US10796222B2 (en) | 2014-12-16 | 2020-10-06 | Balluff Gmbh | Contactless position/distance sensor having an artificial neural network and method for operating the same |
DE102016115015C5 (de) | 2016-08-12 | 2023-01-26 | Sick Ag | Induktiver Näherungssensor |
EP3829064A1 (de) * | 2019-11-28 | 2021-06-02 | Sick Ag | Sensoren zum ermitteln eines ausgabewerts, verfahren zum auswerten eines sensorsignals und verfahren zum trainieren einer ausgabeeinheit zum auswerten eines sensorsignals |
US11573100B2 (en) | 2019-11-28 | 2023-02-07 | Sick Ag | Sensors for determining an output value, method for evaluating a sensor signal, and method for training an output unit to evaluate a sensor signal |
EP3968520A1 (de) | 2020-09-10 | 2022-03-16 | Sick Ag | Sensor zur erfassung eines objekts und verfahren zur auswertung eines sensorsignals |
US12007229B2 (en) | 2020-09-10 | 2024-06-11 | Sick Ag | Sensor for detecting an object and method of evaluating a sensor signal |
Also Published As
Publication number | Publication date |
---|---|
CN105122009A (zh) | 2015-12-02 |
US20160076912A1 (en) | 2016-03-17 |
CN105122009B (zh) | 2017-09-15 |
US10209097B2 (en) | 2019-02-19 |
DE112013006849A5 (de) | 2015-12-03 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
WO2014146623A1 (de) | Induktiver wegmesssensor und verfahren zu seinem betrieb | |
EP2389564B1 (de) | Verfahren zum induktiven erzeugen eines elektrischen messsignals sowie zugehörige sensorvorrichtung | |
DE10014348B4 (de) | Vorrichtung zur zerstörungsfreien Messung der Dicke dünner Schichten | |
EP0782692B1 (de) | Sensoranordnung und verfahren zur messwerterfassung mit der sensoranordnung | |
EP2511736B1 (de) | Vorrichtung und Verfahren zur Detektion von elektrisch leitfähigen Gegenständen | |
EP2312338A1 (de) | Vorrichtung und Verfahren zur Detektion von elektrisch leitfähigen Gegenständen | |
EP3282586B1 (de) | Induktiver näherungssensor | |
EP3293499B1 (de) | Verfahren zum betreiben eines magnetisch-induktiven durchflussmessgeräts und magnetisch-induktives durchflussmessgerät | |
WO2016095882A1 (de) | Berührungsloser positions-/abstandssensor mit einem künstlichen neuronalen netzwerk und verfahren zu seinem betrieb | |
DE202018006650U1 (de) | Elektronisches Gerät mit induktivem Sensor | |
WO2007048395A1 (de) | Ausleseverfahren für sensorfelder, insbesondere für fingerabdruck-sensoren | |
DE102010039528A1 (de) | Kapazitive und/oder induktive Abstandsmessung | |
EP2318850B1 (de) | Anordnung zur messung wenigstens eines wertes einer an einer elektronischen komponente anliegenden spannung | |
EP2981851B1 (de) | Objektsuchgerät und verfahren zum orten eines metallischen und/oder magnetisierbaren objekts | |
EP3829064B1 (de) | Sensoren zum ermitteln eines ausgabewerts, verfahren zum auswerten eines sensorsignals und verfahren zum trainieren einer ausgabeeinheit zum auswerten eines sensorsignals | |
DE102017128472A1 (de) | Induktiver Näherungsschalter und Verfahren zum Betreiben eines induktiven Näherungsschalters | |
EP2108963A2 (de) | Anordnung zum Erfassen einer Drehung eines Drehelements | |
EP2550540B1 (de) | Verfahren zum erkennen von magnetisch gekennzeichneten objekten sowie entsprechende vorrichtung | |
DE19948892A1 (de) | Impulsdetektor und Verfahren zur Detektion von sinusförmigen Impulsen | |
EP3968520B1 (de) | Sensor zur erfassung eines objekts und verfahren zur auswertung eines sensorsignals | |
DE102017128471A1 (de) | Induktiver Näherungsschalter und Verfahren zum Betreiben eines induktiven Näherungsschalters | |
DE102008064544A1 (de) | Positions-/Wegmesssystem und Verfahren zur Bestimmung der Position eines Gebers | |
AT404408B (de) | Verfahren und vorrichtung zur untersuchung eines mediums | |
DE102019131873A1 (de) | Induktive Sensoranordnung und Betätigungsanordnung | |
WO2018220079A1 (de) | Induktiver touch-sensor und verfahren zum betreiben eines solchen |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 13718090 Country of ref document: EP Kind code of ref document: A1 |
|
WWE | Wipo information: entry into national phase |
Ref document number: 14778219 Country of ref document: US |
|
WWE | Wipo information: entry into national phase |
Ref document number: 1120130068494 Country of ref document: DE Ref document number: 112013006849 Country of ref document: DE |
|
REG | Reference to national code |
Ref country code: DE Ref legal event code: R225 Ref document number: 112013006849 Country of ref document: DE |
|
122 | Ep: pct application non-entry in european phase |
Ref document number: 13718090 Country of ref document: EP Kind code of ref document: A1 |