DE202012012649U1 - Device for measuring an alternating current - Google Patents

Abstract

Device for measuring an alternating current in a conductor or measured variables derived therefrom by an AC voltage with a mains frequency, comprising a current channel for detecting a current signal i (t) of the conductor comprising a network period, a voltage channel for detecting a voltage signal associated with the current signal i (t) u (t) of the conductor, wherein in the voltage channel at a sampling frequency fa working first analog-to-digital converter for converting the voltage signal u (t) into a discrete-time voltage signal u (n) and in the current channel a Rogowski coil, in which by the Current signal i (t) a Rogowski coil voltage r (t) proportional to the differential of the current signal i (t) is formed, a second analog-to-digital converter connected to the Rogowski coil for converting the Rogowski coil voltage r (t) into a discrete-time Rogowski coil signal r (n) and an integrator unit which, on the basis of the Rogowski coil signal r (n), enters the current sig nal i (t) proportional integrator signal i (n) are determined, are arranged, characterized in that in the voltage channel, a phase locked loop is arranged, which determines the network period and the latter to the integrator unit for use in the determination of the integrator signal i (n) transmits.

Description

The present invention relates to a device for measuring an alternating current in a conductor or measured variables derived therefrom by an AC voltage with a mains frequency, comprising a current channel for detecting a current signal i (t) of the conductor having a mains period, a voltage channel for detecting a current signal i (t) associated voltage signal u (t) of the conductor, wherein in the voltage channel one operating at a sampling frequency f _a first analog-to-digital converter for converting the voltage signal u (t) in a discrete-time voltage signal u (n) and in the current channel a Rogowski Coil in which a Rogowski coil voltage r (t) proportional to the differential of the current signal i (t) is formed by the current signal i (t), a second analog-to-digital converter connected to the Rogowski coil for conversion of the Rogowski coil. Coil voltage r (t) into a time discrete Rogowski coil signal r (n) and an integrator unit, which is based on the Rog owski coil signal r (n) a the current signal i (t) proportional integrator signal i (n) determined, are arranged.

• (a) Erfassen eines eine Netzperiode aufweisenden Spannungssignals u(t) des Leiters über einen Spannungssensor,
• (b) Digitalisieren des Spannungssignals u(t) durch einen mit einer Abtastfrequenz f_a arbeitenden ersten Analog-Digital-Wandler unter Erhalt eines zeitdiskreten Spannungssignals u(n),
• (c) Erfassen eines Stromsignals i(t) des Leiters über eine Rogowski-Spule, wobei in der Rogowski-Spule durch das Stromsignal i(t) eine dem Differential des Stromsignals i(t) proportionale Rogowski-Spulenspannung r(t) gebildet wird,
• (b) Wandeln der Rogowski-Spulenspannung r(t) in ein zeitdiskretes Rogowski-Spulensignal r(n) unter Einsatz eines zweiten Analog-Digital-Wandlers,
• (e) Ermitteln eines dem Stromsignal i(t) proportionalen Integrator-Signals i(n) unter Zugrundelegung des Rogowski-Spulensignals r(n) mittels einer Integrator-Einheit.

Furthermore, the present invention relates to a method for measuring an alternating current in a conductor or measured variables derived therefrom that is characterized by an alternating voltage with a line frequency, comprising the method steps:

• (a) detecting a line period voltage signal u (t) of the conductor via a voltage sensor,
• (b) digitizing the voltage signal u (t) by a first analog-to-digital converter operating at a sampling frequency f _{a to} obtain a time-discrete voltage signal u (n),
• (c) detecting a current signal i (t) of the conductor via a Rogowski coil, wherein in the Rogowski coil by the current signal i (t) a Rogowski coil voltage r (t) proportional to the differential of the current signal i (t) is formed .
• (b) converting the Rogowski coil voltage r (t) into a time discrete Rogowski coil signal r (n) using a second analog-to-digital converter,
• (E) determining an integrator signal i (n) proportional to the current signal i (t) on the basis of the Rogowski coil signal r (n) by means of an integrator unit.

Devices in this sense are, for example, digital multimeters which are suitable for determining electrical measured variables, preferably for measuring electrical power or work in alternating current networks. Typical mains frequencies are 33.3 Hz, 50 Hz or 60 Hz.

State of the art

In known AC meters, a Rogowski coil is often provided for measuring AC electrical currents. A Rogowski coil is a toroidal coil with an air core. This Rogowski toroid consists of a helical coil of wire with two coil ends and can be opened for ambulatory measurement purposes, when the lead of one coil end is returned through the center of the coil to the other coil end, so that the two coil ends are arranged on the same side of the Rogowski coil are. Due to the air core, Rogowski coils have a low inductance and are therefore particularly suitable for detecting rapidly changing currents.

To measure a current in a conductor, the Rogowski coil is positioned to surround the conductor. It provides a voltage that is proportional to the differential of the phase current. In order to obtain a size proportional to the conductor current, it is necessary to integrate this voltage. The integration can be done by means of an analog or a digital integrator unit. Digital integrator units have the advantage that on the one hand they have high measuring accuracy and on the other hand good long-term stability. The use of a digital integrator unit first requires an analog-to-digital converter, which converts the analog measuring signal of the Rogowski coil, ie the Rogowski coil voltage, into a time-discrete, digital Rogowski coil signal. Subsequently, the digitized signal is supplied to the digital integrator unit, which generates therefrom a measurement signal proportional to the current signal.

However, this measuring signal can be impressed by the circuit arrangement, a DC component, which distorts the measurement signal. Such a DC component can be composed of several components of different origin. For example, the analog part of the transmission path, that is to say the signal path from the Rogowski coil to the analog-to-digital converter, imparts to the measurement signal a circuit-related DC component. This DC component is referred to below as a path error. It may also cause the digital integrator unit to detect in addition to the phase current over time a path error current that accumulates in the integrator unit.

A digital power meter, which has a reduced circuit-related impressed DC component and thus a reduced path error is, for example, from US 7,548,054 B2 known. The known power meter has a current and a voltage channel, wherein in the current channel a Rogowski coil, an analog-to-digital converter and an integrator unit are arranged. In addition, the power meter has a voltage channel with an analog-to-digital converter arranged therein.

To reduce the path error, a digital filter in the form of a high-pass filter is arranged in the current channel of the power meter. As a high-pass filter in electronics called filters that allow frequencies above a cutoff frequency pass almost unattenuated. Frequencies below the cutoff frequency are attenuated by the high pass. An ideal high-pass filter has a cutoff frequency that defines the transition between a passband and a stopband, with all vibrations having a frequency above the cutoff frequency being transmitted unchanged and all frequencies below the cutoff frequency being eliminated.

However, an ideal high-pass filter is not feasible in practice. Available real high pass filters have a transition region between passband and stopband. A real high-pass changes in particular the signal form of the signal components with frequencies in the transition region, resulting in both a change in the signal amplitude (amplitude error) and a phase shift (phase error) of the signal components of the corresponding frequencies. The amplitude error as well as the phase error affect the accuracy of the meter. To eliminate the phase error in the power determination is in the power meter from the US 7,548,054 B2 also arranged in the voltage channel a high-pass filter.

Although this eliminates the phase error, it does not eliminate the amplitude error. In addition, the high-pass filter arranged in the voltage channel causes an additional amplitude error in the measurement of the voltage.

Technical task

The invention is therefore based on the object of specifying a device for measuring an alternating current in a conductor, which allows a high measuring accuracy in the determination of AC measured variables and which is also easy and inexpensive to manufacture.

Furthermore, the invention has for its object to provide a method for measuring a mains period having alternating current in a conductor, which allows the most accurate measurement of AC measured variables.

General description of the invention

With regard to the device, this object is achieved on the basis of a device of the type mentioned in the present invention, that in the voltage channel, a phase locked loop is arranged, which determines the network period and this to the integrator unit for use in the determination of the integrator signal i (n ) transmits.

The device according to the invention is suitable for measuring an alternating current in a conductor and for determining measured variables derived therefrom, for example an alternating current power or an alternating current work.

Since alternating current is an electric current that changes its size and direction regularly, it has a periodic course. It can correspond to a sinusoidal course or deviate from it. A current signal i (t), which is determined by an alternating voltage u (t) with a relatively constant line frequency f and an arbitrary load resistance R via the equation i (t) = u (t) / R, has equidistant signal or mains periods.

In order to detect both the voltage and the associated current signal of a conductor, in the device according to the invention in addition to a current channel for detecting the current signal of the conductor and a voltage channel for detecting the voltage signal of the conductor is provided. The digitization of current and voltage signal is carried out via two analog-digital converters, which sample both signals at the same time. A sampling frequency fa, which corresponds to at least 100 times the analyzed network frequency f, ensures a high measured value resolution.

In order to achieve a high measurement accuracy in the determination of the alternating current signal and the measured variables derived therefrom, it is provided in the device according to the invention that the alternating current to be measured is determined by the network period, that is, based on one or more network periods. A network period is understood to mean the time-smallest repetitive interval of the alternating current profile, which corresponds to the reciprocal of the network frequency.

The current signal i (t) is initially detected via a Rogowski coil arranged in the current channel, with the Rogowski coil voltage r (t), which is subsequently converted via an analog-to-digital converter into a time-discrete Rogowski coil signal r (n). To get out of this Rogowski coil signal r (n) To obtain a magnitude proportional to the current signal i (t), it is necessary to integrate the Rogowski coil signal r (n). For this purpose, an integrator unit is provided in the current channel of the device according to the invention, which is synchronized with control signals from the phase locked loop to the network periods. The integrator unit supplies an integrator signal i (n) proportional to the current signal i (t).

The already mentioned phase-locked loop for detecting the individual network periods is arranged in the voltage channel of the device according to the invention. In an alternative embodiment, the phase locked loop could also be in the current channel.

A phase-locked loop - also referred to below as "PLL" by Engl. phase-locked loop - is an electronic circuit arrangement comprising a phase detector, a loop filter, an integrator and a controllable oscillator. At the phase detector, the voltage signal is present as an external reference signal. The phase detector compares the phase position of the voltage signal with that of the controllable oscillator and regulates the phase position and frequency of the variable oscillator such that they are tracked to the voltage signal. The oscillator signal generated by the phase-locked loop is synchronized in phase with the external reference signal after a transient process. The phase-locked loop finally determines successive network periods that underlie the voltage signal and the current signal.

To dimension a digital phase-locked loop, the components loop filter and integrator are defined for both the analog and the discrete range. For example, the integrator is specified as a backward integrator in the z range and the loop filter as a delay element of the first order in the s range. H _filter (s) = K / (1 + s) ↔ H _filter (z ^-1 ) = b (1 + z ^-1 ) / (1 + cz ^-1 ) H _integrator (s) = (L - s) / (2s) ↔ H _integrator (z ^-1 ) = z ^-1 / (1 - z ^-1 )

Together with the phase detector, this results in a second-order phase locked loop which can be dimensioned in the s range.

The relationship between the s-plane and the z-plane is shown by the following substitutions: s = L (z-1) / (z + 1) ↔ z = (L + s) / (L-s)

The use of z = exp (jw), sampling frequency w _a and natural angular frequency w _n leads to the transformation parameter L = w / w _n / tan (πw _n / w _a ).

By coupling the individual components of the PLL into a control loop, a quadratic function of s is formed in the denominator of the transfer function H _pll (s). H _pll (s) = (H _filter (s) * H _integrator (s)) / (1 + H _filter (s) * H _integrator (s)) H _pll (s) = (-K / 2 * s + K * L / 2) / (s * s + (1-K / 2) * s + K * L / 2)

With the parameter resonant circuit damping D, the control behavior of the phase-locked loop can be determined and the normal form of the quadratic equation in the denominator polynomial s · s + 2 · s · D · w _n + w _n · w _n can be solved. X = 4 · L · D · D - 2 K = X - √ (X · X - 4)

The application of the substitution equation in the transfer function H _filter (s) and the result from the quadratic equation finally lead to the coefficients b and c of the loop _filter . H _filter (z ^-1 ) = K / (L + 1) (1 + z ^-1 ) / (1 + (1-L) / (1 + L) z ^-1 ).

In a preferred embodiment of the device according to the invention, it is provided that the integrator unit comprises a bilinear integrator supplying an integrator raw signal and a DC equalizing unit downstream of the bilinear integrator with a second mean value filter for determining a DC signal component of the integrator raw signal, the DC signal correction Unit determines the DC component of the integrator raw signal for a first network period interval and subtracts this DC component of an integrator raw signal of a subsequent network period interval.

An integrator unit with a bilinear integrator is inexpensive to manufacture. At the bilinear integrator is present as an input signal to the Rogowski coil signal r (n) or an input signal derived therefrom. By integrating the input signal, the bilinear integrator provides a raw integrator signal proportional to the current signal i (t). However, this integrator raw signal is often impressed by a circuit-related DC component, which can falsify the measurement result.

In order to reduce the DC component of the integrator raw signal, a DC signal correction unit is provided after the bilinear integrator. This first determines the DC component in the integrator raw signal for a first network period interval and then subtracts it from the integrator raw signal of a subsequent network period interval. Through this DC signal correction Unit is further avoided accumulating the DC components in the integrator unit, so that the required by the use of the Rogowski coil integrator unit is protected from overflow without thereby distorting the signal to be measured.

A network period interval is understood to be a period of time after which the course of the current signal repeats itself. In the simplest case, this period of time is a network period. Preferably, the mean value filter is a tracking filter that can be adapted to the network period. A tracking filter ensures an exact determination of the measured quantities.

It has proven to be advantageous if the bilinear integrator has a transfer function which obeys the following mathematical expression: (π / M) · ((1 + z ^-1 ) / (1-z ^-1 )) where M is the subsampling factor and z is a delay line, and if the second average filter has a transfer function satisfying the following mathematical expression: (1-z ^-M ) / (Mx (1-z- ¹ )).

The bilinear integrator is provided for integrating the Rogowski coil signal r (n) or signals derived therefrom, which is proportional to the differential of the measurement current i (t). A bilinear integrator with a transfer function according to the above-mentioned mathematical expression ensures an exact integration of the discrete-time Rogowski coil signal. The integration is based on a grid period interval.

Characterized in that after the bilinear integrator a DC signal correction unit is arranged with a mean value filter with the above transfer function, the output of the bilinear integrator, so the determined for a network period interval integrator raw signal, or a signal derived therefrom, as an input to the average filter at. The mean value filter allows any DC component present in this signal to pass unfiltered, so that it is available for a subtraction from an integrator raw signal of a subsequent network period interval.

A mean value filter having the above-mentioned transfer function is a keyed, signal-adapted mean value filter whose integration interval is continuously adjusted to the signal period 1 / f via the sub-sampling factor M. The fact that the mean value filter can be adapted exactly to the network period, this allows an accurate determination of the DC component of the Rogowski coil signal r (n); In addition, it can also be understood as a rectangular window of a discrete Fourier transformation (DFT) with the window length M and the sampling frequency fa. It is assumed that the windowed signal continues periodically.

It has proven useful when the second analog-to-digital converter operates at the sampling frequency fa of the first analog-to-digital converter, the sampling frequency fa being at least 100 times the mains frequency.

Characterized in that the first and the second analog-to-digital converter operate at the sampling frequency fa, a simple synchronization of the measurement signals is made possible. A sampling frequency that is at least 100 times the mains frequency ensures an exact measured value determination.

An advantageous embodiment of the device according to the invention provides that the phase locked loop for determining a grid period interval sets a sub-sampling time p and a sub-sampling factor M, and transmits them to the integrator unit for use in the determination of the integrator signal i (n).

The device is based on a signal processing algorithm which operates as a multirate system with a number M of sampling times n and a sub-sampling time p per network period 1 / f.

Since the phase-locked loop follows the voltage signal u (n), the individual sampling times n are defined by the sampling frequency of the associated digital-to-analog converter.

The sub-sampling time p corresponds to the last sampling time n of a detected signal period of the reference signal. Preferably, a sub-sampling time p is the time of the positive zero crossing, for example a sinusoidal signal period. In this case, the period duration is the time that elapses until the next positive zero crossing, which sets the sub-sampling time p + 1 of the subsequent period. A period duration includes a plurality of sampling times n.

The subsampling factor M denotes the number of sampling times n which are between sub-sampling time p and sub-sampling time p + 1. The number of sampling times M may vary over time from network period to network period.

The control parameters p and M generated by the phase locked loop enable a synchronous evaluation of all measured signals detected by the device according to the invention, such as, for example, the grid period-based integration of the Rogowski coil signal. The network period-synchronous evaluation ensures in particular an exact determination of composite measured variables, such as AC work or AC power.

Since a Rogowski coil can only detect alternating currents, the Rogowski coil signal r (n) recorded with the Rogowski coil and processed in the measuring channel should ideally not contain a DC component. In practice, however, this is not achievable. Thus, due to the circuit, the Rogowski coil signal is often impressed with a DC component, which distorts the measurement result.

In order nevertheless to ensure an exact determination of the current signal of the conductor and the measured variables derived therefrom, it is provided in the method according to the invention that interfering DC components of the Rogowski coil signal are detected and eliminated for each individual network period in the integrator unit.

Sub-sampling time p and sub-sampling factor M are detected by the phase locked loop. Any existing direct current or direct current components, for example in the Rogowski coil signal r (n) or in the output signal of the integrator unit, are determined for a first network period interval starting from a starting time p of the first network period interval. They are then available for a subtraction of these DC components from the Rogowski coil signal r (n) or the output signal of the integrator unit in a subsequent network period interval, for example the network period interval with the start time p + 1.

The integrator necessary for simulating the current signal i (t) is the transformation equation s = 2 / T (1-z ^-1 ) / (1 + z ^-1 ) of the bilinear z-transformation for both the analog and the discrete range Are defined. H _INT (s) = 1 / s ↔ H _INT (z ^-1 ) = T / 2 (1 + z ^-1 ) / (1 - z ^-1 )

T = 2π / M results in the transfer function for the bilinear integrator H _INT (z ^-1 ) = π / M (1 + z ^-1 ) / (1-z ^-1 ) where M is the sub-sampling factor and z ^{-1 is} a delay element with the fundamental delay 1 / f _a .

In a preferred embodiment of the device according to the invention, discrete average filters are used as tracking or tracking filters for determining unwanted DC components in the measurement signals, as well as for summarizing measurement results whose filter length is continuously adapted to the period 1 / f of the analyzed signal. Such tracking filters enable a high measured value resolution (f / f _a ) and can be realized by the following transfer function: H _MEAN (z ^-1 ) = 1 / M (1-z ^-M ) / (1-z ^-1 ).

A filter which has this transfer function and whose filter length (integration interval) is adapted to the period 1 / f of the input signal via the subsampling factor M respectively indicates the DC component of a single signal period of the input signal at sub-sampling time p. The function of the mean value filter can thus be understood as a single channel of a discrete Fourier transformation (DFT) with a rectangular window (window length M, sampling frequency f _a ). Again, it is assumed that the windowed signal continues periodically.

To eliminate disturbing DC components in the measurement signals for regenerating the current signal i (t), the integrator unit has both a DC component correction module for the Rogowski coil signal r (n) and a start value correction module for the integrator signal i (n) ,

The DC component correction module first determines the DC component r _mean [p] of the Rogowski coil signal r (n) _p via a first tracked mean value filter and subtracts this DC component r _mean [p] from the individual sample values of the Rogowski coil signal r (n) _{p + 1 of} the subsequent network period p + 1, so that a signal signal without Rogowski coil r ° (n) _{p + 1} = r (n) _{p + 1} - r _mean [p] is formed.

It is provided that the correction value used in the DC component correction module is determined from in each case one network period of the Rogowski coil signal r (n). In addition, a mean DC component, for example in the form of an arithmetic mean, over several network periods could also be used as a correction value.

In a further advantageous embodiment of the device according to the invention it is provided that the integrator unit comprises a correction unit with a first average filter for determining a DC component of the Rogowski coil signal r (n), wherein the correction unit the DC component of the Rogowski coil signal r (n) for a first network period interval determined and this DC component of a determined for a subsequent network period interval Rogowski coil signal r (n) below Output of a low-signal Rogowski coil signal r ⁰ (n) subtracted.

Since even the Rogowski coil signal r (n) is often impressed with a circuit-related DC component which can corrupt the integrator signal when integrating the Rogowski coil voltage r (n), it is provided that the integrator unit is a correction unit for determining the DC component of the Rogowski coil signal r (n). As a result, a low-signal Rogowski coil signal r ⁰ (n) is obtained. The determination of the DC component of the Rogowski coil signal r (n) is also network-period-based. For this purpose, the network period determined by the phase locked loop is transmitted to the mean value filter of the correction unit for use in the determination of the DC component.

The fact that the phase-locked loop transmits the grid period to the mean value filter enables the determination of the DC signal component of the integrated Rogowski coil voltage for a grid period interval. According to the invention, the DC component correction unit determines the DC component of a first network period interval, which is then subtracted from the Rogowski coil voltage of a subsequent network period interval. In addition, as a correction unit, a middle DC component, for example in the form of an arithmetic mean, could be used over several network periods.

The DC-free Rogowski coil signal r ° (n) is located at the input of the bilinear integrator, which generates therefrom an integrator signal i (n) proportional to the current signal i (t). However, this integrator signal may still have a signal-related DC component.

Therefore, a further corrective action is provided for each start time of the individual network periods. A tracked mean value filter in the starting value correction module first determines, for a first network period p, the DC component i _mean [p] of the integrator signal i (n) _p , which at the beginning of the subsequent network period p + 1 in the difference equation of the bilinear integrator as a correction value i _mean [p] is used. That is, at the start time of the grid period p + 1, the difference equation i (n) _{p + 1} = π / M {r ° (n) _{p + 1} + r ° (n-1) _{p + 1} } + i (n-1 ) _{p + 1 is} extended by the subtrahend i _mean [p].

The integrator signal i (n) thus lies at the input of a tracked average filter, which can pass unfiltered any direct component i _mean [p] present in the signal so that it at the beginning of the subsequent network period p + 1 for a start value correction of the integrator signal i (n) _{p + 1} is available. This corrective action prevents the accumulation of DC components and protects the integrator from overflowing without distorting the measurement signal.

An accurate measurement is ensured by the network period-based evaluation of the current signal as well as by a synchronized evaluation of all measured signals detected by the device according to the invention. This applies in particular to the determination of composite measured variables, for example a service or a work.

In another advantageous embodiment of the device according to the invention it is provided that the integrator unit comprises a correction unit with a first average filter for determining a DC component of the Rogowski coil signal r (n), wherein the correction unit the DC component of the Rogowski coil signal r (n) is determined for a first network period interval and subtracts this DC component from a determined for the first network period interval Rogowski coil signal r (n) with the output of a low-signal Rogowski coil signal r ⁰ (n) subtracted.

Characterized in that the subtracted for a first network period DC component of the Rogowski coil signal from the determined for this network period Rogowski coil signal, a high measurement accuracy is ensured. In order to achieve this, for example, an additional delay element is provided in the DC component correction module.

It has proved to be advantageous if the second mean value filter has a transfer function which satisfies the following mathematical expression: (1 - z ^-M ) / (M · (1 - z ^-1 )), where M is the sub-sampling factor and z is a delay element.

A mean value filter with the above-mentioned transfer function is likewise a tracking filter, which can be adapted to the network period via the control signals of the phase locked loop. Such a mean value filter enables an exact determination of the DC component of the Rogowski coil signal with respect to a network period interval.

In a further preferred embodiment of the device according to the invention it is provided that after the integrator unit, a multiplier for multiplying the discrete-time voltage signal u (n) with the integrator signal i (n) is arranged, and that after the multiplier, a third average filter is arranged wherein the phase-locked loop transmits the sub-sampling time p and the sub-sampling factor M as control signals for use in the filtering to the third average filter.

A multiplier allows the determination of composite measures, such as power, work or energy. The products from the samples of current and voltage at the output of the multiplier are combined by the third average filter to the average power output of individual grid periods. The sum of M products at time p forms a measurement result which corresponds to the average power (power _p = energy _p * f _a / M) of the network period p.

wobei j eine Laufvariable, M der Unterabtastfaktor und z ein Laufzeitglied ist.It has proven useful if the third mean value filter has a transfer function that obeys the following mathematical expression:

where j is a running variable, M is the sub-sampling factor and z is a delay element.

A mean value filter with the above-mentioned transfer function is likewise a so-called tracking filter, which can be adapted to the network period via the control signals of the phase-locked loop. Such a mean value filter enables an exact determination of composite measured quantities in relation to a network period interval.

In a further, likewise preferred embodiment of the device according to the invention, a graphical output unit is provided for the graphical detection of AC parameters, preferably of three-phase parameters, which represents the reactive energy of the alternating current per network period interval as a function of the active energy of the alternating current per network period interval.

A graphical output unit enables a timely graphical representation of the complex power states in the AC network. This graph of analysis results facilitates detection of the power state in the AC grid; it is easy and quick to judge. The graphical representation is shown in a polar diagram, which represents the per-grid period converted apparent energy of the alternating current as a function of the phase angle (phase difference between current and voltage). Such a phase state diagram has four quadrants and can be seen in Cartesian coordinates as a representation of the reactive energy of the alternating current converted per network period over the active energy of the alternating current converted per network period. For each of the up to three network phases analyzed, one measurement point is entered in the phase state diagram per detected network period.

In the graphical output unit, the mean power output of a grid period is represented as the measuring point S (φ, r). The distance r, ie the distance between the origin of the phase state diagram and the measuring point, corresponds to the apparent power. The angle φ of the measuring point furthermore forms the phase position of the measuring signal. The angle function cos (φ) = reactive power / active power applies.

The general equation apparent power = voltage · current (S = U · I) defines the current signal of the conductor I as a modulation signal Imod. This modulation signal Imod may be characterized by four specific resistance operators and vectorially linked in the quadrant of the phase state diagram. If the registered measuring point lies, for example, on the positive axis of abscissa, the associated modulation signal Imod _{1 can be represented} by the mathematical expression Imod ₁ = U * (1 / R). If, on the other hand, the measuring point lies on the positive ordinate axis, the modulation signal Imod ₂ and the relationship Imod ₂ = U * (j * 2π * f * C) apply.

The equation Imod ₃ = U * (-1 / R) describes a measurement point on the negative axis of abscissa and the mathematical expression Imod ₄ = U * (-j / (2π * f * L)) a measurement point on the negative ordinate axis, where j the imaginary unit, f the line frequency, R an ohmic resistance, C is a capacitance and L is an inductance.

It has proven useful if the first, the second or the third average filter by the transfer function H _mean (z ^-1 ) = 1 / M (1-z ^-M ) / (1-z ^-1 ) is realized, where M is the filter length and z - 1 is a delay element with the basic delay 1 / f _a .

An average filter with the above-mentioned transfer function is a so-called tracking filter, which can be adapted to the network period via the control signals of the phase-locked loop. Such a mean value filter enables an exact determination of the DC component of the Rogowski coil signal with respect to a network period interval.

With regard to the method for measuring an alternating current in a conductor, the abovementioned object is achieved on the basis of a method of the type mentioned in the introduction in that after the method step (b) of digitizing the voltage signal u (t) the network period of the time-discrete voltage signal u (n ) is detected via a phase locked loop and transmitted to the integrator unit, wherein the Grid period in the determination of the integrator signal i (n) is used.

In conventional methods for measuring an alternating current in a conductor, it is provided that the voltage signal u (t) of the conductor is detected by a voltage sensor and subsequently converted by a first analog-to-digital converter into a time-discrete voltage signal u (n). Furthermore, the current signal i (t) of the conductor is detected via a Rogowski coil which supplies a Rogowski coil voltage r (t) proportional to the differential of the current signal i (t). The Rogowski coil voltage r (t) is subsequently converted into a time-discrete Rogowski coil signal by means of a second analog-to-digital converter and integrated with an integrator unit, so that an integrator signal i (n) proportional to the current signal i (t) is obtained. is obtained.

Since a Rogowski coil detects only AC currents, the Rogowski coil voltage output by the Rogowski coil should ideally not have a DC component. In practice, however, this is not achievable. Thus, due to the circuit, the Rogowski coil voltage, the Rogowski coil signal and the integrator signal are often impressed with a DC component which can falsify the measurement result. In order nevertheless to ensure an exact determination of the current signal and the measured variables derived therefrom, it is provided in the method according to the invention that the alternating current to be measured is network-cycle-based, that is to say determined with reference to one or more network periods. Since both the current signal and the voltage signal have a network frequency, the determination of a network period can be based on both the voltage signal and the current signal of the conductor.

In the method according to the invention, it is therefore provided that the network period of the time-discrete voltage signal u (n) is determined via a phase locked loop and transmitted to the integrator unit. The determined network period is subsequently available as a control signal.

In the integrator unit, the grid period is used in determining the integrator signal i (n). In this way, a network period-based evaluation and subsequently a synchronized evaluation of the measurement signals detected by the device according to the invention, for example the voltage signal and the current signal, are made possible, so that an exact measurement is ensured. This applies in particular to the determination of composite measured variables, for example a service or a work.

Suitable method modifications result from the above explanations of the device according to the invention.

embodiment

The invention will be explained in more detail with reference to embodiments and some drawings. This shows in a schematic representation

1 a first embodiment of the device according to the invention for measuring an alternating current in a conductor, in which a phase-locked loop is arranged in the voltage channel for detecting equidistant network periods, which are used as synchronized analysis windows,

2 a prior art embodiment of a power meter for alternating current,

3 an embodiment of the integrator unit of the device according to the invention 1 in which both a suppressed average value filter is used for suppressing the DC signal component of the Rogowski coil voltage as well as for minimizing the proportion of the integration constants in the regenerated current signal,

4 a second embodiment of the device according to the invention, which is suitable for measuring three-phase current,

5 a time course of an input signal for both the integrator unit of the inventive device according to 1 as well as for the integrator unit of the prior art AC current measuring device according to 2 as well as the respectively associated output signal of the respective integrator unit,

6 a time course of an input signal for the integrator unit of the device according to the invention 1 with a DC signal component which is suppressed over time as well as the output signal of the integrator unit assigned to the input signal and its DC signal component,

7 a network power state diagram that graphically records the AC parameters reactive energy per grid period as a function of the active energy per grid period,

8th one of the first network phase 7 underlying current / voltage signal,

9 one of the second network phase 7 underlying current / voltage signal, and

10 one of the third network phase 7 underlying current / voltage signal.

1 schematically shows a first embodiment of the inventive device for measuring an alternating current, the total reference numeral 1 assigned. The AC measuring device 1 has a flow channel 2 for detecting a current signal i (t) of a conductor and a voltage channel 3 for detecting the voltage signal u (t) assigned to the current signal i (t).

In the stream channel 2 is a Rogowski coil as a current sensor 4 intended. During the measuring process is the Rogowski coil 4 toroidally arranged around the current conductor, being in the Rogowski coil 4 a Rogowski coil voltage r (t) = di (t) / dt proportional to the differential of the current signal i (t) is formed.

In addition, in the flow channel 2 an analog-to-digital converter 5 and an integrator unit 6 arranged. The analog-to-digital converter 5 operates with a sampling frequency f _a , which determines the number of sampling times n per network period and thus also the measurement accuracy of the method via the ratio f _a / f. In the realized example f _a = 100 · f was chosen. The distance between the sampling times n is the sampling interval. The analog-to-digital converter 5 converts the Rogowski coil voltage r (t) into a discrete-time signal r (n) passing through the integrator unit 6 is fed back to an integrator signal i (n) proportional to the current signal of the conductor i (t).

In the voltage channel 3 is a voltage sensor (not shown) for recording the time course of the voltage signal u (t), an analog-to-digital converter 7 and a phase locked loop 8th arranged. Through the analog-to-digital converter 7 the voltage signal u (t) is converted into a time-discrete voltage signal u (n). Analog to digital converter 7 and analog-to-digital converter 5 operate in phase with the same sampling frequency f _a .

For detecting the network period of the digitized voltage signal u (n), the digital phase locked loop 8th (phase-locked loop; PLL) with a controllable oscillator used. The digitized voltage signal u (n) serves as an external reference signal. The phase locked loop 8th compares the phase position of the external reference signal with the oscillator signal and generates a control signal for adjusting the phase position and frequency of the oscillator signal to the external reference signal. In the steady state phase and frequency of the oscillator signal are adapted to the external reference signal, that is, except for a small deviation, in the as a control loop 2 , Order-sized PLL arises, the controllable oscillator is tracked to the reference signal in phase and frequency.

The phase locked loop 8th provides the output signals subsampling time p and sub-sampling factor M. The oscillator signal forms a sinusoidal signal whose polarity changes in regular repetition. The subsampling instant p is the instant at which the polarity of the oscillator signal changes from minus to plus. He sets the analysis window over the network period. The sub-sampling factor M denotes the number of sampling times n that are between the current positive zero crossing and the previous positive zero crossing; it corresponds to the difference between the sum of the sampling times at time n _p and the sum of the sampling times at time n _p-1 .

The signal processing algorithm underlying the device operates as a multirate system. The subsampling times p and sub-sampling factors M determined by the phase-locked loop are used as control signals in the integrator unit 6 and in the mean filter 9 used.

In the stream channel 2 arranged integrator unit 6 For each network period, the DC component of the time discrete Rogowski coil signal r (n) is determined and subtracted from the Rogowski coil signal r (n) to form a signal Rogowski free coil signal r ⁰ (n). This DC-free Rogowski coil signal r ⁰ (n) is used as a basis for subsequent integration and generates a regenerated phase-current signal in the form of an integrator signal i (n) proportional to the current signal i (t). In the integration process, any DC component present in the integrator signal i (n) is evaluated in a network-period-supported manner by the DC component of the integrator signal i (n) from the preceding network period at the start of the new network period being used to correct the integrator signal i (n). This feedback ensures that the integrator does not overflow and the integrator signal i (n) only reflects the alternating current component of the current in the current conductor.

To determine other AC parameters, such as power, or work, is after the integrator unit 6 the multiplier 11 for the multiplication of the integrator signal i (n) and the voltage signal u (n). The time difference (z-1) between the integrator signal i (n) and the associated voltage signal u (n) caused by the bilinear integrator is already taken into account in the synchronization of the two analog-to-digital converters. Further, after the multiplier 11 a mean value filter 9 and a sub-scanner 10 arranged. The mean value filter 9 is _realized with the transfer function H _MEAN (z ^-1 ) = 1 / M (1-z ^-M ) / (1-z ^-1 ). Since filter length and sub-sampling factor M on the Phase locked loop are coupled, are in each case M in a mean incoming samples between two adjacent Unterabtastzeitpunkten.

As a result of the fact that the integrator signal i (n) and the voltage signal u (n) are determined by network periods, the measured variables derived therefrom can also be determined for each network period. In 1 is an example of an arrangement for determining the energy per network period E _p · f _a / M 12 indicated.

2 shows a prior art embodiment of a power meter, the overall reference numeral 20 assigned.

The power meter 20 is used to measure an alternating current in a conductor. It includes a flow channel 22 and a voltage channel 23 for detecting the current or voltage signal of the conductor. In the stream channel 22 is a Rogowski coil 24 for detecting a Rogowski coil voltage r (t) proportional to the differential of the current signal of the conductor i (t), an analog-to-digital converter 25 for digitizing the Rogowski coil voltage to form a time-discrete Rogowski coil signal r (n) and an integrator unit 26 for integrating the Rogowski coil signal r (n), which outputs an integrator signal i (n) proportional to the current signal i (t). The integrator unit consists of a high-pass filter 26a and an integrator element 26b ,

In the voltage channel 23 is a voltage sensor for recording the time course of the voltage signal of the conductor u (t), an analog-to-digital converter 27 and a high-pass filter 21 arranged. Through the analog-to-digital converter 27 The voltage signal u (t) is first converted into a discrete-time signal u (n), which then the high-pass filter 21 passes.

The high pass filter 21 serves to filter in the current channel through the high-pass filter 26a Conditional signal changes in the amplitude and phase response in the voltage channel 23 to realize. High Pass Filter 21 and high-pass filters 26a therefore correspond to each other.

In addition, the power meter includes 20 a multiplier 28 , a low-pass filter 29 and an accumulator 27 , A network period-based evaluation of the power meter 20 determined measured variables is not provided.

3 shows a schematic representation of an embodiment of the integrator unit 6 out 1 , The integrator unit 6 has an entrance 301 for control signals, that is, for the parameters subsampling time p and sub-sampling factor M, and one input 302 for the time-discrete Rogowski coil signal r (n). It also includes a first tracked average filter 304a and a second tracked average filter 305 ,

The mean value filter 304a is part of the DC offset module 304 which performs a first reduction of DC components in the Rogowski coil signal r (n). Starting from the sub-sampling time p, the mean value filter sums 304a for a first network period p the samples of the discrete-time Rogowski coil signal r (n + 1) _p , r (n + 2) _p ... r (n + M) _p to the DC component r _mean [p] together. The sub-scanner 304b makes the DC component r _mean [p] determined for the subsampling instant p available for the following network period p + 1. This DC component r _mean [p] becomes available in the following network period in the module 304c subtracted from the individual samples of the Rogowski coil signal r (n). The following applies: r ° (n) _{p + 1} = r (n) _{p + 1} - r _mean [p].

The thus obtained equal signal-free Rogowski coil signal r ° (n) is combined with a bilinear integrator 307 integrated, so that an integrator signal i (n) arises in the form of a regenerated current signal. The bilinear integrator comprises the building blocks 307a and 307c , as well as the building block 307b which allows a seed correction. For the bilinear integrator, the following transfer function applies: H _INT (z ^-1 ) = π / M (1 + z ^-1 ) / (1 - z ^-1 ).

When the integrator signal i (n) is generated, its DC component is determined for each network period and by the switch 306 at the beginning of each successive network period as a correction value to the start value correction input 307b of the bilinear integrator. That is, the mean value filter 305 forms for the duration of the network period p from all samples of the integrator signal i (n + 1) _p , i (n + 2) _p ... i (n + M) _p the DC component i _mean [p] and the switch 306 selects the correction value i _mean [p] for the sub-sampling time p, so that this is the starting value correction input 307b at the start of the subsequent network period p + 1 is subtracted from the current integrator signal i (n). For the first sample of the network period p + 1, the following applies: i (n) _{p + 1} = i (n) _{p + 1 -i mean} [p].

output 303 the integrator unit 6 is the integrator signal i (n).

In 4 schematically is a second embodiment of a device according to the invention 400 shown, which is particularly suitable for the measurement of three-phase alternating currents in three conductors. The device 400 points for each of the three Mains phases have a separate voltage channel and a separate current channel. In each stream channel is an integrator unit 404 . 405 . 406 arranged. In addition, each of the voltage channels has a phase locked loop 407 . 408 . 409 on, which determines the mains period of the respective voltage signal and these to the integrator unit 404 . 405 . 406 of the associated current channel for use in integrating the Rogowski coil signal. For this purpose, the phase-locked loop respectively determines a sub-sampling time p and a sub-sampling factor M and transmits these control parameters to the associated integrator unit 404 . 405 . 406 , The respective integrator unit 404 . 405 . 406 determines from the Rogowski coil signal r (n) an integrator signal i (n) proportional to the current signal of the conductor. The device 400 moreover, has its own signal evaluation unit for each of the three network phases 401 . 402 . 403 on, the signal evaluation unit 401 for evaluation of the network phase 1 associated signals, the signal evaluation unit 402 for the evaluation of the network phase 2 associated signals and the signal evaluation unit 403 is provided for evaluation of the network phase 3 associated signals.

The phase locked loops 407 . 408 . 409 determine the frequencies f1 (n), f2 (n), f3 (n) and the phase curves ph1 (n), ph2 (n), ph3 (n) of the individual network phases. The thus determined phase profiles are transmitted to the respectively associated signal evaluation unit. In addition, the phase curve ph1 (n) of the first network phase is also sent to the signal evaluation unit provided for the second network phase 402 transmitted, this determines the phase shift between the first and second network phase. In addition, the phase curve ph2 (n) is also sent to the signal evaluation unit 403 transmit, so that the signal evaluation unit 3 determines the phase shift between the second and third network phase. The phase shift between the first and third network phase is performed by the signal evaluation unit 401 calculated from the previously measured phase shifts calculated by their subtraction of 360 °.

Finally, the device points 400 for the individual phase locked loops 407 . 408 . 409 a common feedback branch 410 on, the one adder 410a and a forward integrator 410b includes. Those from the phase locked loops 407 . 408 . 409 determined frequency values f1 (n), f2 (n) and f3 (n) are determined by the parent PLL feedback branch 410 combined into a common frequency value f (n) and to all signal evaluation units 401 . 402 . 403 transmitted.

5 shows by way of example a digitized input signal for an integrator unit in the form of a Rogowski coil signal r (n). The Rogowski coil signal r (n) has a phase angle. The DC component of the Rogowski coil signal r (n) is zero, that is to say positive and negative signal components cancel each other within a network period. Such a Rogowski coil signal is produced when the current signal i (t) of the conductor monitored by the Rogowski coil has only one positive half-wave but no negative half-wave per network period. In order for no unwanted DC component to occur in the conductor, the DC component of the Rogowski coil test signal described above could be eliminated by a simple potential shift. As a result, the signal shape of the conductor current i (t) to be detected via a Rogowski coil would be identical to the signal form of the output signal i ₁ (n) of the device according to the invention 1 ,

5 also shows the output i ₂ (n) of the device for comparison purposes 2 that corresponds to the state of the art.

The output signal i ₁ (n) of the integrator unit of the device according to the invention 1 is after a few grid periods in a stable, steady state. In the steady state, the device according to the invention provides an output signal which is proportional to the current signal of the conductor i (t) and, like the current signal of the conductor, has only one positive half-wave. The thus obtained output signal of the integrator unit of the device according to the invention maps the current signal to be measured with a high accuracy.

The output signal of the prior art device according to 2 Although it also shows a current waveform similar to the current signal of the conductor i (t). However, this has a lower accuracy with respect to the current signal of the conductor i (t). This can be seen in the course of the output signal, in particular, that the sections with constant current waveform of the current signal of the conductor i (t) in the output signal of this device i ₂ (n) have a slope.

To explain the function of the integrator unit 6 out 3 are in 6 by way of example, an input signal r (n) of the integrator unit 6 , the input signal r ⁰ (n) for the bilinear integrator 307 , as well as the associated output of the integrator unit 6 , namely the integrator signal i (n), and that of the mean value filter 305 determined DC signal component i _mean (p) shown in their time course. The input signal r ⁰ (n) is created by the block 304b supplied direct signal component r _mean [p] of the Rogowski coil signal r (n) is subtracted from this. In this implementation example of the device according to the invention, the individual samples of the Rogowski coil signal r (n) with the DC signal r _mean [p] from the previous period corrected r ° (n) _{p + 1} = r (n) _{p + 1} - r _mean [p].

In the present example, the Rogowski coil voltage r (n) has a DC component. This can be seen, inter alia, from the level difference of the input signal r (n) between the first two network periods p and p + 1, which results from that of the first mean value filter 304a during the first network period p determined DC component r _mean [p] is taken into account only in the second network period p + 1 in the generation of the DC signal-free Rogowski coil signal r (n). For the determination of the low-signal Rogowski coil signal r ⁰ (n) in the first network period, it is assumed that r _mean [p-1] is zero. This transient response could be prevented in an extended implementation example by an additional delay element in the DC component correction module.

In 6 is also the associated output signal i (n) of the integrator unit 6 and the DC component i _mean [p] attributed to the starting value correction input of this unit is represented. To achieve a steady state, two network periods p and p + 1 are required. Thereafter, as the output of the integrator unit 6 a time-stable integrator signal i (n) can be tapped.

7 schematically shows one of a graphical output unit 710 an AC meter according to the invention output network power state diagram 700 , The network performance state diagram 700 has a Cartesian coordinate system in which the reactive energy per grid period is represented as a function of the active energy per grid period. For a better overview, a unit circle is in the coordinate system 720 entered. For each grid period and phase phase, the AC meter determines in real-time operation, that is in time with the detected grid frequency, the AC parameters Reactive Power and Active Power and displays these in the form of a measuring point in the grid power status diagram 700 represents.

This in 7 illustrated network power state diagram 700 also shows by way of example a measurement result of a three-phase alternating current with the three mains phases (indicated by the numbers 1, 2 or 3), wherein the mains phases to each other a phase shift (which is not shown in this diagram) of 120 °. The measurement results for each network phase are included as measurement points in the network power state diagram 700 entered. To identify the individual network phases, the measurement points of a network phase are identified by a number, namely the measurement points of the first network phase with "1", the measurement points of the second network phase with "2" and the measurement points of the third network phase with "3". Depending on the size of the analysis time window, for each measured network period and for each network phase, one measurement point each is entered in the network power state diagram, so that finally a measurement point cloud is obtained. In addition, in order to ensure a representation of the frequency distribution of the determined measuring points, it is additionally provided that the measuring points are displayed in color coded form, depending on their frequency of occurrence, so that frequently determined measuring results can be easily recognized by their color. In the test example for the network performance status diagram 700 the digit size of the measuring points shown represents the number of directly superimposed measuring points. In another embodiment, it is provided that the actuality of the individual points in the diagram is indicated by a change in the brightness of the measuring points, for example, by currently measuring points with a high brightness and older measuring points with a lower brightness being displayed. Network power state diagrams obtained in this way 700 are meaningful for both short and long analysis time windows.

The network performance state diagram 700 allows a simple assessment of the power state of the individual network phases through their recorded point clouds 701 . 702 . 703 ,

For example, from the point cloud 701 in the network power state diagram 700 to recognize that in the first network phase via a variable resistive load active energy is fed.

The point cloud 702 shows the load of the second mains phase by a consumer, which consists of the series connection of a fixed inductive resistor and a variable ohmic resistance. The measuring point where the ohmic and inductive resistances are identical is indicated by the large number "2" in the point cloud 702 excellent. The distance between the individual measuring point and the unit circle is determined by the ratio between the ohmic and the inductive resistance at this load.

The third network phase becomes the phase state diagram in the test example 700 loaded by a variable ohmic resistor, which additionally regulates a phase control by passing the current only at the beginning of each half-period. Here is the resulting point cloud 703 detect a capacitive shift of the network load.

For a more detailed description of the point clouds 701 . 702 and 703 are the corresponding current / voltage signals for the individual network phases in the schematic diagrams 8th . 9 and 10 displayed.

The network power state diagram (in the apparent power level) can be used to effectively differentiate different states in the AC power grid. It is similar to the phase state diagram (in the complex signal plane) used to distinguish modulation types in message transmission.

QUOTES INCLUDE IN THE DESCRIPTION

This list of the documents listed by the applicant has been generated automatically and is included solely for the better information of the reader. The list is not part of the German patent or utility model application. The DPMA assumes no liability for any errors or omissions.

Cited patent literature

• US 7548054 B2 [0007, 0009]

Claims (9)

Device for measuring an alternating current in a conductor or measured variables derived therefrom by an AC voltage with a mains frequency, comprising a current channel for detecting a current signal i (t) of the conductor having a mains period, a voltage channel for detecting a voltage signal associated with the current signal i (t) u (t) of the conductor, wherein in the voltage channel at a sampling frequency f _a working first analog-to-digital converter for converting the voltage signal u (t) into a discrete-time voltage signal u (n) and in the current channel a Rogowski coil, in by the current signal i (t) forms a Rogowski coil voltage r (t) proportional to the differential of the current signal i (t), a second analog-to-digital converter connected to the Rogowski coil for converting the Rogowski coil voltage r (t) into a discrete-time Rogowski coil signal r (n) and an integrator unit which, based on the Rogowski coil signal r (n), enters the current i gnal i (t) proportional integrator signal i (n) are determined, are arranged, characterized in that in the voltage channel, a phase locked loop is arranged, which determines the network period and the latter to the integrator unit for use in the determination of the integrator signal i (n) transmits. Device according to Claim 1, characterized in that the integrator unit comprises a bilinear integrator supplying an integrator raw signal and a DC signal correction unit downstream of the bilinear integrator with a second mean value filter for determining a DC signal component of the integrator raw signal, the DC signal correction unit determines the DC component of the integrator raw signal for a first network period interval and subtracts this DC component of an integrator raw signal of a subsequent network period interval. Apparatus according to claim 2, characterized in that the bilinear integrator has a transfer function which obeys the following mathematical expression: (π / M) · ((1 + z ^-1 ) / (1-z ^-1 )) where M is the sub-sampling factor and z is a delay element. Device according to any one of the preceding claims, characterized in that the phase-locked loop determines a sub-sampling time p and a sub-sampling factor M for determining a mains-period interval and transmits them to the integrator unit for use in determining the integrator signal i (n). Device according to one of the preceding claims, characterized in that the integrator unit comprises a correction unit with a first average filter for determining a DC component of the Rogowski coil signal r (n), wherein the correction unit the DC component of the Rogowski coil signal r (n). n) is determined for a first network period interval and subtracts this DC component from a Rogowski coil signal r (n) determined for a subsequent network period interval with the output of a low-equivalent Rogowski coil signal r ⁰ (n). Device according to one of the preceding claims 2 to 5, characterized in that the integrator unit comprises a correction unit with a first average filter for determining a DC component of the Rogowski coil signal r (n), wherein the correction unit the direct component of Rogowski Coil signal r (n) is determined for a first network period interval and this DC component of a determined for the first network period interval Rogowski coil signal r (n) with the output of a low-signal Rogowski coil signal r ⁰ (n) subtracted. Device according to one of the preceding claims, characterized in that after the integrator unit, a multiplier for multiplying the time-discrete voltage signal u (n) with the integrator signal i (n) is arranged, and that after the multiplier, a third average filter is arranged, wherein the phase-locked loop transmits the sub-sampling time p and the sub-sampling factor M as control signals for use in the filtering to the third average filter. Device according to one of the preceding claims, characterized in that for the graphical detection of AC parameters, preferably of three-phase parameters, a graphical output unit is provided which represents the reactive power of the alternating current as a function of the active power of the alternating current. Device according to one of the preceding claims, characterized in that the first, the second or the third mean value filter by the transfer function H _mean (z ^-1 ) = 1 / M (1-z ^-M ) / (1-z ^-1 ) is realized, where M is the filter length and z ^{-1 is} a delay element with the basic delay 1 / f _a .

Images