Method for measuring impedance of electronic component

Abstract

The invention relates to a method for determining an electrical signal characterizing the magnitude and / or variation of an impedance, which varies under the influence of a magnetic field, an electric field and / or an electromagnetic field, which in turn varies as a result of an object approaching the impedance, at a plurality of intervals and thus repeatedly. The intervals include at least one set of measurement intervals and at least one compensation interval temporally prior to the at least one set of measurement intervals or each set of measurement intervals, the set of measurement intervals including one interval or a plurality of consecutive measurement intervals. For each measurement interval, a sinusoidal excitation voltage is applied to the impedance (ZX), so that the impedance (ZX) induces a measurement current (IX) as a result of the excitation voltage. The measurement current (IX) of the compensation interval is used as a compensation current of at least one subsequent measurement interval in such a way that the compensation current in the at least one subsequent measurement interval is subtracted from the measurement current (IX) of the measurement interval in order to form a simulated evaluation current signal. The evaluation current signal is a signal characterizing the magnitude and / or variation of the impedance (ZX).

Description

Technical Field

[0001] The present invention relates to a method for measuring the impedance of an electrical component, and in particular to a method for determining an electrical signal characterizing the magnitude and / or change of the impedance at multiple intervals and thus repeatedly, wherein the impedance can change, for example, under the influence of a magnetic field, an electric field and / or an electromagnetic field, which can change due to the proximity of an object to the impedance. Background Art

[0002] Depending on the application, it is very important to repeatedly determine the impedance of electrical components installed in the vehicle. Application examples include detecting whether the driver's hands are holding the steering wheel or detecting seat occupancy.

[0003] The measurement signal for impedance measurement is not offset-free due to various influencing factors or compensation measures. Externally connected EMC filters can cause such an offset, for example.

[0004] The offset signal component is sometimes significantly larger than the useful signal component. Therefore, when digitally processing the measurement signal, an analog-to-digital converter (ADC) with relatively high resolution and number of bits and a high permissible input voltage must usually be used, which, however, means too high costs and is therefore disadvantageous. If the maximum resolution of the ADC used is limited, the resolution of the measurement signal will also be severely limited, which is particularly disadvantageous for relatively large signals.

[0005] A method for measuring the impedance of an electrical component is known from US-A-2021 / 0081073. According to this method, a transimpedance amplifier is used to amplify the measurement signal, which means that relatively large measurement signals can be processed. However, this carries the risk that when the measurement signal is digitally processed, the analog-to-digital converter required for this should have a relatively high resolution and number of bits and a high permissible input voltage, which requires correspondingly high costs as described above.

[0006] US-B-9575105 describes a method and a device for measuring complex impedance.

[0007] Finally, DE-A-10 2013 227 225 discloses a current-based charge compensation in a capacitive touch sensor. Summary of the invention

[0008] The object of the invention is to provide a method of the type mentioned above with which the signal representing the impedance can be kept low without compromising the accuracy of the measurement.

[0009] To achieve this object, the present invention proposes in a first variant embodiment a method for determining at multiple intervals and thus repeatedly an electrical signal characterizing the magnitude and / or variation of an impedance, which varies under the influence of a magnetic, electric and / or electromagnetic field, which in turn varies due to the proximity of an object to the impedance, wherein in the method,

[0010] - the intervals comprise at least one group of measurement intervals and at least one compensation interval temporally preceding the at least one group of measurement intervals or each group of measurement intervals, the at least one group of measurement intervals comprising one measurement interval or a plurality of consecutive measurement intervals;

[0011] -For each measurement interval,

[0012] - apply a sinusoidal excitation voltage to the impedance, and

[0013] - the measurement current induced in the impedance due to the excitation voltage,

[0014] - wherein the measured current (I X ) as a compensation current for at least one subsequent measurement interval: the compensation current in the at least one subsequent measurement interval is converted from the measurement current (I X ) to form an analog evaluation current signal, and

[0015] - The evaluation current signal is a characteristic of the impedance (Z X )’s size and / or changes.

[0016] According to the invention, the determination is carried out in an interval comprising at least one group of measuring intervals and at least one compensation interval temporally preceding the at least one group of measuring intervals or each group of measuring intervals, wherein the group comprises one measuring interval or a plurality of consecutive measuring intervals. In all these intervals (measuring intervals and compensation intervals), the measurement current induced by the impedance under the action of an excitation voltage, which excitation voltage is excited in the form of a sine wave, is recorded. The measurement current of the compensation interval is used as the compensation current in at least one subsequent measuring interval in the following manner: the measurement current in the at least one subsequent compensation interval is subtracted from the measurement current obtained in the measuring interval to form an analog evaluation current signal. The evaluation current signal now characterizes the magnitude and / or change of the impedance.

[0017] The method according to the invention only records the change in impedance between the intervals, so the current signal remains "small". Even when, for example, an ADC with limited resolution (i.e. number of bits) and limited allowed input voltage range is used for digital signal processing purposes, the high resolution can still be used for further processing.

[0018] According to the invention, it is advantageous if the analog evaluation current signal is amplified and the evaluation current signal amplified in this way is a signal that characterizes the magnitude and / or change of the impedance. Here, too, there are no significant limitations when converting the amplified analog evaluation current signal into a digital signal if the amplified analog evaluation current signal does not (yet) cause the ADC to be out of range. If necessary, the gain factor of the amplifier can be automatically and dynamically adjusted accordingly so that the amplified signal always has the maximum permissible value required to avoid overloading the ADC, regardless of the magnitude of the analog evaluation current signal at the amplifier input.

[0019] In a further advantageous embodiment of the invention, the compensation current can be generated by the signal generation unit after it has been determined, in order to subtract it from the measurement current of the at least one measurement interval.

[0020] Compensation intervals can be placed between measurement interval groups and within measurement interval groups in a variety of ways. For example, at the beginning of the process, a (single) compensation interval can be provided, which provides a compensation current for subsequent measurement intervals, which compensation current is always subtracted from the corresponding measurement current measured in each subsequent measurement interval. However, if there is a concern that the compensation current may no longer be adjusted well enough during intermittent impedance measurements to produce a sufficiently small analog evaluation current signal after subtracting the compensation current from the corresponding measurement current, it is advantageous to perform another compensation interval as soon as the overload or overdrive limit of the ADC is exceeded in order to obtain a relatively small analog evaluation current signal in the next measurement interval. If the evaluation current signal is amplified before the analog-to-digital conversion (which is the usual case), the condition for redetermining the compensation current can be: exceeding a specified maximum allowed value of the evaluation current signal at the input of the amplifier or, when using an amplifier whose amplification factor can automatically adapt to the input signal size, falling below a specified minimum allowed amplification factor.

[0021] As described above, the object of the present invention is to be able to continue to use relatively small analog evaluation current signals in order to be able to continue to use ADCs which are limited in terms of resolution and maximum permissible analog input voltage but which generally still have a sufficiently high resolution after amplification, provided that further signal processing takes place digitally, as will be discussed further below.

[0022] In a further advantageous embodiment of the invention, it is advantageous if the possibly amplified analog evaluation current signal can be subjected to an IQ demodulation after the analog-to-digital conversion, wherein the I signal component and / or the Q signal component of the IQ demodulation are signals characterizing the magnitude and / or change of the impedance. In this embodiment of the invention, digital IQ demodulation is used, whose I and Q amplitudes provide values ​​characterizing the magnitude and / or change of the impedance. In a refinement of the invention, the compensation current can now be generated based on the parameters of the digital IQ demodulation of the last compensation interval. In this case, the compensation current is also generated digitally, and then the generated compensation current is converted into an analog compensation current by a digital-to-analog converter (DAC), which is then subtracted from the measured current of the corresponding measurement interval.

[0023] Digital signal processing in the form of digital IQ demodulation has many advantages. However, because the ADC input voltage is limited, the maximum resolution that can be achieved by digital signal processing is limited. For example, if the offset increases the measured value, the maximum resolution that can be achieved will be reduced accordingly. Therefore, the current connected in parallel with the measured variable impedance should be compensated. According to the present invention, this is achieved by subtracting the compensation current from the actual measured current, so that the analog evaluation current signal is as small as possible. Then, after possible signal amplification, high-resolution measurement values ​​for digital signal processing can be obtained despite the limited resolution of the ADC. The amplitude of the analog evaluation current can even be amplified before processing in the ADC. For the sinusoidal excitation signal advantageously used within the scope of the present invention, the offset signal can be generated as the sum of the scaled IQ demodulated signals, and can be fed back, for example, to perform subtraction using a current DAC.

[0024] When using analog demodulation, the achievable signal resolution can be increased by subtracting the DC offset after demodulation and then amplifying the signal before feeding it into the ADC. This is not possible in the case of digital demodulation, as provided in an advantageous development of the invention, because an alternating signal is present in the ADC and this signal cannot be compensated by the DC offset.

[0025] The compensation method according to the invention, which advantageously includes digital demodulation, can amplify the evaluation current signal to a higher level without exceeding the (limited) input voltage range of the ADC. As with analog demodulation, this can improve the resolution.

[0026] According to an advantageous embodiment of the invention, the compensation current can be calculated as follows:

[0027] - a preferably computer-implemented neural network model of any type and any structure as is known in principle from the prior art, and / or

[0028] - preferably a computer implemented Hidden Markov Model, and / or

[0029] - Petri nets and / or any type of automatic learning that generates knowledge from past and previous experience, such as preferably computer-implemented machine learning, preferably computer-implemented deep learning and / or preferably computer-implemented processing of predictors, i.e. predictor variables based on past registered events,

[0030] - wherein the execution code for executing the corresponding method is preferably stored in a memory and is preferably executed by a processor (computer implementation).

[0031] According to a variant of the above object, a method is proposed for determining at multiple intervals and thus repeatedly an electrical signal characterizing the magnitude and / or variation of an impedance, which varies under the influence of a magnetic, electric and / or electromagnetic field, which in turn varies due to the proximity of an object to the impedance, wherein in the method,

[0032] -For each interval,

[0033] - apply a sinusoidal excitation voltage to the impedance, and

[0034] - the measurement current induced in the impedance due to the excitation voltage,

[0035] - the difference between the measured current and the compensation current is fed to the input of an analog or digital integrator, the output of which provides the compensation current after possible digital-to-analog conversion, and

[0036] - The output of the integrator (42) forms a signal representative of the magnitude and / or change of the impedance.

[0037] In a second variant of the invention, a feedback system is used to form a compensation current, which is subtracted from the measured current. In this respect, each interval is therefore a measurement interval in the above sense of the first variant of the invention. The difference between the compensation current and the measured current is integrated by an error or delta integrator that reacts to changes in the measured signal. In the steady-state case, the delta integrator provides a constant signal at its output, which indicates that the impedance does not change. If the impedance changes, the output signal of the integrator will also change and is then fed back as a changing compensation current, thereby also indicating a change in the impedance.

[0038] The second variant of the invention is also suitable for the smallest possible signals, which in turn brings about the advantages already mentioned: for the conversion of the analog measurement signals into digital signals, a conventional ADC with limited resolution and a limited permissible input voltage range can be used (so that these small analog measurement signals can be amplified), without having to give up the high resolution required for evaluating the measurement current.

[0039] In another advantageous embodiment of the invention, according to its second variant, a difference signal according to the difference between the measurement current and the compensation current is subjected to an analog-to-digital conversion after possible amplification and is then fed to a digital IQ demodulator, whose I and Q signal components are fed to the input of a digital integrator and, in order to form the compensation current, the time-integrated I and Q signal components are modulated in a digital IQ modulator, whose output is connected to a digital-to-analog converter which outputs the compensation current.

[0040] According to a third variant of the invention, the above object is achieved by a method for determining, at multiple intervals and thus repeatedly, an electrical signal representative of the magnitude and / or variation of an impedance that varies under the influence of a magnetic, electric and / or electromagnetic field that in turn varies due to the proximity of an object to the impedance, wherein in the method,

[0041] -For each interval,

[0042] - Apply a sinusoidal excitation voltage to the impedance,

[0043] - the measurement current induced in the impedance due to the excitation voltage,

[0044] - applying a compensation voltage 180° phase-shifted relative to the excitation voltage to a reference impedance of known magnitude, and

[0045] - a compensation current is induced in the reference impedance due to the compensation voltage,

[0046] the difference between the measured current and the compensation current is fed, possibly after amplification, to the input of an analog or digital integrator, the output of which provides the magnitude of the compensation voltage, possibly after digital-to-analog conversion,

[0047] - Generate compensation voltage,

[0048] - The output of the integrator (42) is a signal representative of the magnitude and / or change of the impedance.

[0049] In a third variant of the invention, a reference impedance is used to determine the size of the impedance. The reference impedance is excited with a compensation voltage and then induces a compensation current, which is subtracted from the measurement current induced by the impedance to be measured under the action of the excitation voltage. The difference signal is fed to a digital delta integrator, which integrates the difference and determines a compensation voltage at its output, which in turn is used to excite the reference impedance. The size of the compensation voltage characterizes the size of the impedance and / or its changes. In addition to selecting a reference impedance, several reference impedances of different sizes can also be selected, depending on the size of the impedance to be measured. In this way, the difference signal is minimal, which brings the above-mentioned advantages: despite the limited number of bits and input voltage of the ADC, the digital part still has high resolution.

[0050] Different reference impedances can be selectively controlled via a multiplexer. Alternatively, the reference impedance can also be adjusted in magnitude. Here, it is also proposed to use a superposition control loop in order to normalize the measurement result to 1.

[0051] In a further advantageous embodiment of the invention, a difference signal according to the difference between the measurement current and the compensation current can be subjected to an analog-to-digital conversion after possible amplification and then fed to a digital IQ demodulator, whose I and Q signal components are fed to the input of a digital integrator and, in order to form the compensation voltage, the time-integrated I and Q signal components are modulated in a digital IQ modulator, whose output is connected to a digital-to-analog converter which outputs the compensation voltage.

[0052] The impedance to be measured is in particular a seat occupancy sensor operated capacitively, the electrodes of which are brought close to each other under the influence of weight, which results in a change in capacitance (see, for example, US Pat. No. 8,896,326 B2, which describes an evaluation principle for determining the magnitude of the impedance using IQ demodulation, which can also be advantageously used according to the invention). Another field of application of the method according to the invention is the detection of whether the driver is holding the steering wheel. For this purpose, separate electrode surfaces are arranged on the steering wheel, which together with the vehicle chassis form a capacitor, the dielectric of which is influenced by the hand, which results in a change in capacitance. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] The present invention is explained in more detail below by means of various embodiments and with reference to the accompanying drawings.

[0054] Figure 1 A block diagram showing a first embodiment of an impedance measurement circuit,

[0055] Figure 2 A block diagram showing a second embodiment of an impedance measurement circuit,

[0056] Figure 3 A block diagram showing a third embodiment of an impedance measurement circuit,

[0057] Figure 4 A block diagram showing a fourth embodiment of an impedance measurement circuit,

[0058] Figure 5 A block diagram showing a fifth embodiment of an impedance measurement circuit, and

[0059] Figure 6 A block diagram showing a sixth embodiment of an impedance measurement circuit is shown. DETAILED DESCRIPTION

[0060] Figure 1 The block diagram of the circuit 10 of the first embodiment of the present invention is shown. The impedance Z to be measured is excited by a sinusoidal voltage signal generated by the digital part 12 of the circuit 10. X The digital sinusoidal signal is converted into an analog sinusoidal signal by the DAC 14 and then fed to the impedance Z to be measured through the driver 18 after being low-pass filtered in the low-pass filter 16. X In response to this, the impedance Z X The induced current I X , which is mirrored in the current mirror 20. After current-to-voltage conversion in the IV converter 22 (eg, a shunt resistor) and bandpass filtering in the bandpass filter 24, the mirror current I X is fed to an amplifier 26, the output signal of which is converted to a digital signal by an ADC 28 for processing in the digital section 12. In the digital section 12, IQ demodulation is performed in an IQ demodulator 30 to determine the complex induced current I X Then, after further filtering in the digital filter 32, the impedance Z is obtained. X The two signals of the amplitude can be further processed, for example, outside the circuit 10.

[0061] Due to interference suppression measures, e.g. for EMC protection, the impedance to be measured Z X The input 34 of the typical integrated circuit 10 is connected to a resistor relative to the impedance to be measured Z X This results in an offset in the measured signal which should be compensated in order to enable high-resolution measurements using a relatively simple method. XThe compensation current is subtracted from the offset, thereby compensating for the offset. Due to this compensation, the remaining signal is very small, which can still be converted into a digital signal with sufficiently high resolution after amplification by the ADC 28 with a limited number of bits and a limited input voltage and further processed in the digital part 12. The compensation current is generated in the digital part 12, IQ modulated in the IQ modulator 36 and then converted into an analog compensation current by the DAC 38.

[0062] The circuit 10 operates intermittently or in a plurality of intervals. The measured value of the measurement current determined in the digital part 12 in one of these intervals can be used to operate the IQ modulator 36 for generating a digital compensation current using the parameters of the IQ demodulator 30.

[0063] Figure 2 shows the method for measuring the impedance Z X The second variant embodiment of the circuit 40. As long as the components of the circuit 40 are Figure 1 The components of the circuit 10 correspond to or are similar to those in Figure 2 Use with Figure 1 The same reference numerals are used to represent the same figures in the drawings.

[0064] according to Figure 2 The circuit 40 is based on Figure 1 The difference between the circuit of FIG. 1 and FIG. 2 is that the (Δ) integrator 42 integrates the I and Q signal components at the output of the IQ demodulator 30. The integrator 42 integrates both the I signal component and the Q signal component. The output signal of the integrator 42 is fed to the IQ modulator 36 to generate a digital compensation current, which is converted into an analog compensation current by the DAC 38 and then converted from the measured current I X The analog compensation current is subtracted from the IQ modulator. The function of the I output of the IQ modulator is as follows:

[0065] .

[0066] The equation for the Q output of the IQ modulator results in the following:

[0067] .

[0068] here, , is the output of the delta integrator.

[0069] If the impedance Z XIf the impedance value does not change with the interval, the demodulator 36 no longer outputs any signal (signal zero), which means that the impedance value last calculated by the integrator 42 continues to apply. If the impedance value changes, the output signal of the demodulator 36 deviates from zero. This changes the output of the integrator 42 and thus also changes the compensation current, which is reduced from the measured current I in the next measurement interval. X If the impedance Z X If there is no change during this period, a zero signal is again generated at the output of the demodulator 36 .

[0070] Figure 3 shows a method for measuring the impedance Z which may change during operation. X Here, again, those of the circuit 50 are similar to those of the circuit 50. Figure 2 The components of the circuit 10 are the same as or correspond to the components in Figure 3 Also used in Figure 2 The same reference numerals are used to represent the same components.

[0071] exist Figure 3 In the example shown, the excitation reference impedance Z REF To generate a compensation current, and use its inductive current I REF To this end, a digital compensation voltage is generated in the digital part 12 by means of an IQ modulator 36, which is DAC-converted in a DAC 38 and, after being filtered in a low-pass filter 52 if necessary, is applied to the reference impedance Z via a driver 54. REF .

[0072] Inductive current I REF The current I is mirrored by the current mirror 56 and is measured from the mirror in a known manner. X Minus.

[0073] In the digital part 12, the I and Q signal components of the IQ demodulator 30 are integrated again, wherein the output signal of the integrator 42 is IQ-converted in the circuit 50. The reason for this is that for the I and Q signal components of the IQ demodulator 30, the outputs of the delta integrators may have to be swapped, depending on the reference impedance Z REF Is it composed of only capacitors or only resistors. If Z is realized only by capacitors REF , then the I output of delta integrator 42 must be connected to the inverting Q input of IQ modulator 36, and the Q output of delta integrator must be connected to the I input of IQ modulator 36. REF If it is composed of resistors only, Figure 2Similarly, the I output of delta integrator 42 must be connected to the I input of IQ modulator 36, and the Q output of delta integrator 42 must be connected to the Q input of IQ modulator 36. REF has an imaginary part (e.g., a capacitance component) and a real part (e.g., a resistance component), then a transformation is required so that the impedance to be measured Z X The change of capacitance and resistance of the IQ demodulator 30 only changes the imaginary part (Q component) or the real part (I component) in the measurement result. The compensator must be phase neutral. To this end, the IQ conversion circuit 57 is used to rotate the Z REF The phase shift caused by this is such that the sum is -180º. The transfer function for the I output of the IQ modulator is as follows:

[0074] .

[0075] The equation for the Q output of the IQ modulator results in the following:

[0076] .

[0077] here, , is the output of the Δ integrator. The phase φ is generated due to Z REF The phase shift required to achieve phase neutrality is:

[0078] ).

[0079] The result for amplitude A is:

[0080] .

[0081] Using R NORM The parameter performs IQ normalization. This has the following effect: if the impedance to be measured Z X Capacitance C X Equal to the normalization parameter R NORM =1 / ωC X (Or, the impedance to be measured Z X The resistance R X Equal to parameter R NORM =R X ), the measurement result of the imaginary or real part becomes 1. This amplitude normalization can be performed using an IQ transform, or can be performed outside the control loop at the output of the delta integrator 42, such as in Figure 3 58 executions shown.

[0082] Figure 4 Again, the impedance Z is measured. X Here, it also applies that the individual components of the circuit 60 are Figure 4 Also used in Figure 3 The same reference numerals are used as long as they correspond to or are identical to components of the circuit 50 .

[0083] Circuit 60 with Figure 3 The difference between the circuits shown is that the reference impedance can be switched to a different reference impedance Z REF1 and Z REF2 It is also possible to select between a plurality of discrete reference impedances. The switching is performed by a multiplexer 62 which receives a control signal from the digital section 12 and switches the compensation voltage to a plurality of reference impedances Z REF1 , Z REF2 , .... The induced compensation current I REF is fed to the current mirror via the same multiplexer 62 or another multiplexer (not shown). Thereafter, the sequence behaves in the same way as Figure 3 Exactly the same as shown.

[0084] Figure 5 shows the method for measuring the impedance Z X Here, it also applies that the individual components of the circuit 70 are Figure 5 Also used in Figure 4 The same reference numerals are used as long as they correspond to or are identical to components of the circuit 60 .

[0085] According to Figure 4 The circuit 60 is compared according to Figure 5 The circuit 70 does not switch to a different discrete impedance reference, but uses the reference impedance Z REF , the reference impedance Z REF Capacitance C REF and resistor R REF In fact, it is continuously adjustable. The control signal comes from the digital part 12. Then, the digital controller (72) outputs the signal corresponding to R X The value of I is the I component and corresponds to Z X The value of is taken as the Q component. This means that R X =R REF And C X =C REF However, if C X Can be greater than C REF The maximum possible value is set, and R X Can be greater than R REF The maximum possible value set, the specification of the digital controller (72) must be less than 1, for example 0.1, which means that the output of the digital controller (72) will be output as C REF or R REFThe I and Q components are 1 / 10 of the set value, so the impedance to be measured Z X R X and C X will be 10 times greater than the output of the digital controller (72). REF (For example, implemented as a capacitance decade and a resistance decade) is set by means of a higher-level control loop so that the I and Q outputs of the delta integrator 42 take the value of 1 in the steady state. Then, the unknown impedance Z X The measured value of Z REF The setting is expressed in decimal. Figures 2 to 4 An advantage of the sensor architecture shown is that non-linearities in the signal chain (IV converter 22, bandpass filter 24, amplifier 26, ADC 28) have no impact on the measurement accuracy. Figure 5 The sensor shown is also insensitive to nonlinearities in the signal and reference paths (DAC 14, DAC 38, current mirror 20, current mirror 56). The measurement accuracy is entirely determined by Z REF Decimal decision.

[0086] Finally, reference Figure 6 A sixth embodiment of the present invention is described. This embodiment differs from the other embodiments in that a computer-implemented artificial intelligence element is preferably used to generate the offset signal (compensation current) over both hardware and software artificial intelligence elements.

[0087] as Figure 2 In an embodiment of the invention, the signal at the output of the delta integrator is fed to the IQ modulator 36', but intermediately processed by a feature extraction section 74, to which the IQ data at the output of the demodulator 30 and the signal after processing in the delta integrator 42 are fed. The resulting feature vector 76 is passed to an importance enhancement stage 78, resulting in a modified feature vector 80, which is fed to a preferably computer-implemented neural network 82, the structure of which is arbitrary and therefore corresponds to a generally known neural network. The importance enhancement stage 78 can be formed, for example, by a processor of the device, which processor in particular also executes a computer-implemented method for increasing the importance of the feature vector 76, so as to generate, for example, a modified feature vector 80. The I and Q data of the IQ modulator 36' are then output at the output of the neural network 82, the output signal of which specifies the compensation current after conversion into an analog signal in the DAC 38.

[0088] Therefore, the feature extraction section provided in the present embodiment detects the I and Q data of the demodulator and the output signal after being processed in the delta integrator 42, and generates a feature vector MV therefrom (see reference numeral 76).

[0089] Information about eigenvectors can be found, for example, at https: / / de.wikipedia.orq / wiki / Merkmalsvektor.

[0090] For each dimension of the feature vector there is an allowed value range of a certain size. The significance enhancement step SST (see reference numeral 78) now distorts the feature vector so that in each dimension 50% of the training values ​​are above a threshold of half the value range and 50% are below a threshold of half the value range, but preferably still within the allowed value range of the corresponding dimension (these are exemplary illustrations). The significance enhancement stage 78 generates a mapping that corresponds to a matrix polynomial. In order to form the nth addend of the polynomial, by means of a matrix-vector multiplication X*MV n A distortion matrix Xn is formed with n instances of the eigenvector MV and added to all the summands of the polynomial.

[0091] Other clustering methods are also possible.

[0092] For example, the importance enhancement stage may increase the importance of a feature vector based on the training data in one of the following different ways:

[0093] Feature Engineering

[0094] By extracting the value of new features (feature engineering), the importance enhancement stage can extract relevant information from the existing features of the feature vector MV and add it to the feature vector MV in the form of new dimensions. This can increase the importance of the feature vector.

[0095] Feature Selection

[0096] Instead of using all available features of the feature vector, the dimensionality of the feature vector can be reduced again by feature selection dependent on the feature vector in order to select those features that are most relevant to the current pattern recognition task.

[0097] For example, importance enhancement can use computer-implemented machine learning methods that apply regularization techniques (e.g., LI and L2 regularization) to influence the weights of features and control the importance of specific features.

[0098] Ensemble-Method

[0099] Importance boosting can use ensemble methods such as random forest or gradient boosting to increase the importance of feature vectors by combining multiple models and selecting the best features.

[0100] Domain knowledge

[0101] The incorporation of domain knowledge of device application scenarios in importance enhancement and the implementation of corresponding computer-implemented methods can improve the importance of feature vector MV' (see reference numeral 80) by incorporating relevant information into the model. This can be achieved by manual adjustment or by using a pre-trained model with domain knowledge.

[0102] The neural network NN (see reference numeral 82 ) evaluates the corresponding feature vectors with increased selectivity.

[0103] The above explanations about feature extraction and importance enhancement are not limited to Figure 6 Embodiment of the invention.

[0104] Reference numerals list

[0105] 10 Circuit

[0106] 12 Digital part

[0107] 14 DAC

[0108] 16 Low pass filter

[0109] 18 Driver

[0110] 20 Current mirror

[0111] 22 IV Converter

[0112] 24 Bandpass filter

[0113] 26 Amplifier

[0114] 28 ADC

[0115] 30 IQ Demodulator

[0116] 32 Digital Filter

[0117] 34 Input

[0118] 36 IQ Modulator

[0119] 36' IQ Modulator

[0120] 38 Digital to Analog Converter

[0121] 40 Circuit

[0122] 42 Δ Integrator

[0123] 50 Circuit

[0124] 52 Low pass filter

[0125] 54 Driver

[0126] 56 Current Mirror

[0127] 57 IQ conversion circuit

[0128] 58 IQ Normalization Circuit

[0129] 60 Circuit

[0130] 62 Multiplexer

[0131] 70 Circuit

[0132] 72 Digital Controller

[0133] 74 Feature Extraction Unit

[0134] 76 Eigenvectors

[0135] 78 Importance Enhancement Level

[0136] 80 Eigenvectors with increased importance

[0137] 82 Neural Networks

[0138] Z REF1 Reference impedance

[0139] Z REF2 Reference impedance

[0140] Z X Impedance to be measured

[0141] I X Measuring current

[0142] I REF Reference current.

Claims

1. A method for determining at multiple intervals and thus repeatedly an electrical signal characterizing the magnitude and / or change of an impedance, said impedance changing under the influence of a magnetic, electric and / or electromagnetic field, which in turn changes due to the proximity of an object to said impedance, wherein: In the method, - the intervals comprise at least one group of measurement intervals and at least one compensation interval temporally preceding the at least one group of measurement intervals or each group of measurement intervals, the group of measurement intervals comprising one measurement interval or a plurality of consecutive measurement intervals, -For each measurement interval, - to the impedance (Z X ) applies a sinusoidal excitation voltage, and - due to the excitation voltage in the impedance (Z X ) is induced into the measuring current (I X ), - wherein the measured current (I X ) as a compensation current for at least one subsequent measurement interval: the compensation current in the at least one subsequent measurement interval is converted from the measurement current (I X ) to form an analog evaluation current signal, and - The evaluation current signal is a characteristic of the impedance (Z X )’s size and / or changes.

2. The method according to claim 1, characterized in that The compensating current is generated by a signal generating unit after it has been determined in order to separate it from the measuring current (I X ) is subtracted.

3. The method according to claim 1 or 2, characterized in that: The analog evaluation current signal is amplified, and the evaluation current signal amplified in this way is a signal that characterizes the magnitude and / or the change in the impedance.

4. The method according to any one of claims 1 to 3, characterized in that The interval includes a first measurement interval group that is located before the first compensation interval in time, and the compensation current is provided for the first measurement interval group, and / or in an initial state before the first change, the impedance (Z X ) is known in size.

5. The method according to any one of claims 1 to 4, characterized in that The compensation current is calculated as follows: - neural networks of any type and structure, and / or -HMM, and / or - Petri Nets and / or any type of automatic learning that generates knowledge from past and previous experiences, such as machine learning, deep learning and / or processing of predictors, i.e. predictor variables based on past registered events.

6. The method according to any one of claims 1 to 5, characterized in that The analog evaluation current signal, which may have been amplified, is subjected to IQ demodulation after analog-to-digital conversion, wherein the I signal component and / or the Q signal component of the IQ demodulation are representative of the impedance (Z X )’s size and / or changes.

7. The method according to claim 6, characterized in that The compensation current is digitally generated in the signal generation unit using parameters of the digital IQ demodulation in a temporally previous compensation interval.

8. The method according to any one of claims 1 to 7, characterized in that The impedance (Z X ), wherein the capacitive impedance is designed as a capacitor with a dielectric or as an electrode surface.

9. A method for determining at multiple intervals and thus repeatedly an electrical signal characterizing the magnitude and / or variation of an impedance, said impedance varying under the influence of a magnetic, electric and / or electromagnetic field, which in turn varies due to the proximity of an object to said impedance, wherein: In the method, - for each of said intervals, - to the impedance (Z X ) applies a sinusoidal excitation voltage, and - due to the excitation voltage in the impedance (Z X ) is induced into the measuring current (I X ), - The measured current (I X ) and the compensation current is fed to the input of an analog or digital integrator (42), the output of which provides the compensation current after possible digital-to-analog conversion, and The output of the integrator (42) forms a characteristic of the impedance (Z X )’s size and / or changes.

10. The method according to claim 9, characterized in that The difference signal based on the difference between the measurement current and the compensation current is subjected to analog-to-digital conversion after possible amplification and is then fed to a digital IQ demodulator (30), whose I signal component and Q signal component are fed to the input of a digital integrator (42), and in order to form the compensation current, the I signal component and the Q signal component integrated over time are modulated in a digital IQ modulator (36), whose output is connected to a digital-to-analog converter (38) which outputs the analog compensation current.

11. A method for determining at multiple intervals and thus repeatedly an electrical signal characterizing the magnitude and / or variation of an impedance, said impedance varying under the influence of a magnetic, electric and / or electromagnetic field, which in turn varies due to the proximity of an object to said impedance, wherein: In the method, - for each of said intervals, - to the impedance (Z X ) applies a sinusoidal excitation voltage, - due to the excitation voltage in the impedance (Z X ) is induced into the measuring current (I X ), - To a reference impedance of known magnitude (Z REF ) applies a compensation voltage that is 180° phase-shifted relative to the excitation voltage, and - Due to the compensation voltage, the reference impedance (Z REF ) to induce a compensation current (I REF ). - The measured current (I X ) and the compensation current (I REF ) is fed to the input of an analog or digital integrator (42) after possibly being amplified, and the output of the integrator provides the magnitude of the compensation voltage after possibly being converted from digital to analog, - generating said compensation voltage, and - The output of the integrator (42) is a signal representative of the magnitude and / or variation of the impedance.

12. The method according to claim 11, characterized in that Multiple reference impedances (Z REF ), wherein, according to the magnitude of the compensation voltage, the reference impedance (Z REF ) is applied with the compensation voltage, and thus provides a compensation current (I REF ).

13. The method according to claim 11, characterized in that The reference impedance (Z REF ) is variably adjustable and is varied by a controller (72) having an input for receiving a signal at the output of the integrator (42) and an output, wherein the output of the controller (72) provides a signal representative of the magnitude and / or variation of the impedance.

14. The method according to any one of claims 11 to 12, characterized in that Based on the measured current (I X ) and the compensation current (I REF ) is subjected to analog-to-digital conversion after possible amplification and is then fed to a digital IQ demodulator (30), whose I signal component and Q signal component are fed to the input of a digital integrator (42), and in order to form the compensation voltage, the I signal component and the Q signal component integrated over time are modulated in a digital IQ modulator (36), whose output is connected to a digital-to-analog converter (38) which outputs the analog compensation current.

15. The method according to any one of claims 11 to 13, characterized in that Based on the measured current (I X ) and the compensation current (I REF ) is subjected to analog-to-digital conversion after possible amplification and then fed to a digital IQ demodulator (30), the I signal component and the Q signal component of the digital IQ demodulator are fed to the input of a digital integrator (42), and in order to form the compensation voltage, the I signal component and the Q signal component integrated over time are modulated in a digital IQ modulator (36), the output of the digital IQ modulator is connected to a digital-to-analog converter (38) outputting the analog compensation current, and in addition, in a superimposed control loop with the digital controller (72), the I component and the Q component of the reference impedance are modulated according to the control index for the I component and the Q component according to the IQ output of the digital integrator (42).

16. A method for determining at multiple intervals and thus repeatedly an electrical signal characterizing the magnitude and / or variation of an impedance, said impedance varying under the influence of a magnetic, electric and / or electromagnetic field, which in turn varies due to the proximity of an object to said impedance, wherein: In the method, - for each of said intervals, - to the impedance (Z X ) applies a sinusoidal excitation voltage, and - due to the excitation voltage in the impedance (Z X ) is induced into the measuring current (I X ), - According to the measured current (I X ) or calculate the compensation current from a value derived from the measured current: - neural networks of any type and structure, and / or -HMM, and / or - Petri nets and / or any type of automatic learning that generates knowledge from past and previous experience, such as machine learning, deep learning and / or processing of predictors, i.e. predictor variables based on past registered events, and - The measured current (I X ) and the difference between the compensation current forms a characteristic of the impedance (Z X )’s size and / or changes.

Images