DE3045033A1 - Measuring time average values - is for improved control response, using digital sampling and continuous integration, within limits - Google Patents

Abstract

The arrangement is for measuring the time average of analogue measurement values and enables rapid continuous measurement of effective values, real and imaginary components, apparent power and time dependent Fourier coefficients. It enables the dynamic response of control loops to be improved, esp. in the initial build-up period to be made. The measurement value or Fourier coefficient, if required after multiplication by analogue input parameters, is converted into time discrete form using a clocked (2) sample and hold circuit (1) with a certain sampling rate. The output of the sample and hold circuit goes to an analogue to digital converter (3) which drives a first adder (5) and a shift register (4) whose number of stages equals the measurement period divided by the sampling rate. The circuit contains further summation and register circuits, a factor element (6) and a feedback loop.

Description

Arrangement for determining the mean values over time

of analog measured values The invention relates to an arrangement according to the preamble of the patent claim.

There are a number of electrical quantities in AC voltage networks, the determination of which requires averaging. When calculating the effective value of Current and voltage or when determining active and reactive power, a mean value must be calculated can be made through integration over the period. Is it at Current and voltage around non-sinusoidal, periodic quantities exist alongside the Fundamental harmonics, their amplitudes and phase positions through the Fourier series expansion are set. To determine the Fourier coefficients must also averaging can be carried out.

In control loops, there is a low-delay to improve the control dynamics Actual value acquisition desired. Is this actual value acquisition with averaging connected, transducers have so far been used that operate on the principle of multiplication of the number of phases work. The single-phase Actual value synthetically expanded to a multi-phase variable. This measure leads in addition, that the measuring time constant decreases proportionally with the number of phases, see references: "ETZ-A" Vol. 97 (1976), Issue 12, pages 759-768, article: Gretsch, R. "Methods for the rapid measurement of mains voltage fluctuations" and "Technische Communications AEG-TELEFUN-KEN "vol. 61 (1971) issue 6, pages 327-329, article: Breitenberger, G. "Low-delay AC-DC converter. For the representation of Effective active and reactive power ".

However, there are limits to reducing the measurement constants because In the case of an unfavorable input variable configuration, there is a strong overshoot of the output variables makes noticeable that increases with increasing number of phases. Responsible for this behavior, which is undesirable in practice, is the differentiating proportions in the circuit structure.

Fast actual value generation is particularly important when there is compensation of reactive power and harmonics.

The invention is based on the object of providing an arrangement for fast, correct and continuous determination of the above also in the transient range Indicate mean values that do not have the disadvantages mentioned.

According to the invention, this object is achieved by the Claim specified features solved.

The arrangement according to the invention is illustrated below with reference to the drawing described embodiment shown in more detail.

The invention is based on the consideration that in the event that how here the time function of the mean values is required, a continuous integration of the function for which the time average is formed, what is done by Introducing time-dependent integration limits is achievable.

If u (t) is a periodic function with the period T, then the mean value u can be applied with The following applies in particular to the active power p (t): The differentiation of this equation according to the time gives If one writes the differential quotient in the form of a difference quotient, then with t = nht with n = 0,1,2, ...

or rewritten: This means that the current value p (n. Tt) is made up of the previous value and a correction value.

In the circuit according to the figure, the analog input variables the voltage u (t) and the current i (t) are fed to a multiplier 9. On his The active power p (t) appears at the output.

This value is fed to a sample-and-hold element 1, in which it is carried out the supply of clock signals of a clock generator 2 with a sampling rate At is time-discretized will. In the following A / D converter 3, which is also controlled by the clock signals from the clock generator is clocked, an amplitude quantization takes place.

The output of the A / D converter leads on the one hand to an adder 5, on the other hand to a T / dS t-stage shift register 4, its content is pushed further in time with the clock. The output of this shift register is fed to adder 5 negated. A value appears at its output, the corresponds to the correction value mentioned above.

The output value of the adder is used in the subsequent factor generator 6 5 multiplied by the constant factor ort t 1 T.

The digital multiplication with the constant factor # t / X is particularly simple if T = Edt and a natural number is chosen for K. When using the dual number system, K can be equal to a power of two, because then the multiplication with the constant factor by shifting the digits of the binary number (cf. shifting the Eomma in the decadal number system). In this way, no additional circuit complexity is required for the multiplication.

The output signal of the factor generator 6 is used by a further adder 7, the output of which leads to a single-stage shift register 8. This The shift register is also controlled by the clock pulses from the clock generator 2.

Its output is fed back to the input of the adder 7.

So if the value p ((n - 1). Li t), it is added to the correction value with the next clock pulse and results as the storage value p (n ast). At the output of the shift register (terminal 8a) appears the current value p (n ß) at the next cycle.

If an analog output is required, the output signal of the Shift register to a D / A converter 10 with a subsequent low-pass filter 11 for smoothing fed to the stepped output variable of the D / A converter. At the output (terminal 11a) the corresponding analog signal then appears.

The arrangement described can also telwertbi accordingly for Mib Charging for the effective value as well as for reactive and apparent power can be used.

It can also be used to determine time-dependent Fourier coefficients be used. As mentioned at the beginning, their determination becomes necessary when it if current and voltage are non-sinusoidal, periodic quantities, see above that in addition to the fundamental, there are harmonics, their amplitudes and Phase positions are determined by the Fourier series expansion. The measurement period T in this case is a multiple of the measuring period for the pure sinusoidal Course of voltage and current.

A particular advantage of this digital measuring arrangement is that it is simple Adaptation of the calculation algorithm to the respective period duration of the measurement signal. In addition only the frequency of the clock (2) has to be changed.

Blank page

Claims (1)

Claim: Arrangement for rapid continuous determination the temporal mean values of analog measured values, such as rms values, active, reactive or apparent powers, as well as time-dependent Fourier coefficients, characterized in that that the measured value or Fourier coefficient, possibly after multiplying analog input variables, in a sample-and-hold element (1) by supplying Clock signals of a clock generator (2) with a sampling rate bt is time-discretized, that the sample-and-hold element (1) is also an A / D converter which is clocked by the clock signals (3) connected downstream, the first whose output on the one hand to an / adder (5), on the other hand leads to an X, stage shift register (4), where T is the measuring period and where the content of the stages of the shift register in time with the clock signals of the clock is shifted so that the output of the shift register is negated is fed to the adder, and that the output of the adder via a factor generator (6) is fed to a second adder (7), the output of which is set to a single-stage Shift register (8), which is also clocked by the clock signals of the clock generator, leads, the output of the single-stage shift register to the input of the second Adder is returned.