JP2000180484A - Apparatus for measuring harmonic wave - Google Patents

Abstract

PROBLEM TO BE SOLVED: To correctly measure harmonic waves to a high order at an arbitrary time with a simple circuit configuration by sampling an input signal at a predetermined fixed frequency, obtaining number of sampling for a plurality of cycles by rounding, operating discrete Fourier transforming(DFT) of its data and measuring harmonic wave components. SOLUTION: A frequency of a fundamental wave of a input signal is measured, and sampled by a predetermined fixed frequency. A sampling number N used for its measurement is, for example, four cycles of the input signal. Then, a sampling point near a zero-crossing position of two sampling points sandwiching a zero-crossing position of a fourth cycle is used as an N-th point. Thus, N-th sampling point is decided by a rounding method to minimize an error. Thus, the harmonic wave components can be measured based on the DFT calculation based on N pieces of the obtained sampling data.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a harmonic measuring device for measuring a harmonic content and a phase thereof in an electric power system.

[0002]

2. Description of the Related Art Conventionally, the following method has been adopted for measuring harmonics in a power system.

(1) A method in which a sampling frequency of an input signal from a system is locked in a hardware manner by a PLL circuit, and a harmonic is measured by synchronizing the sampling frequency with a fluctuation of a signal waveform.

(2) A method of measuring a frequency of an input signal from a system and then measuring a harmonic by adjusting a sampling frequency to the measured frequency.

(3) A method in which a sampling frequency is fixed and a harmonic is measured by multiplying an input signal by a window function.

(4) A method in which the sampling frequency is fixed and DFT is performed while shifting by one sample, and an average value of half cycle results is used (this method is disclosed in JP-A-6-194395). It is shown).

[0007]

In the variable sampling methods (methods in which the sampling frequency fluctuates) of the above (1) and (2), the measurement accuracy is high because the sampling frequency is adjusted to the measurement frequency. In the method of (1), a PLL circuit is required in hardware, harmonics cannot be measured at the time of transient change (because the response cannot catch up), and oscilloscope measurement (waveform observation by capturing a transient waveform) is performed simultaneously. There is a drawback in that a hardware circuit including an A / D converter different from that for oscilloscope measurement is required when trying to perform the measurement, and the method (2) has a disadvantage in frequency measurement and harmonic measurement. Variable sampler that cannot be used during transitions due to different timing, requires a hardware timer for accurate sampling timing To be grayed scheme, the hardware circuit is necessary to include a separate A / D converter and for oscilloscope measurement circuit, there are drawbacks like.

Further, the method (3) has a problem that it is not possible to correctly obtain even higher-order harmonics, and that the window function itself becomes an error.

Further, the method of the above (4) is the same as the method of the above (3).
In the same way as above, it is not possible to correctly measure up to higher harmonics, the effective frequency fluctuation range is very narrow, up to 0.5 Hz, and furthermore, it is not possible to share the device for 50/60 Hz measurement. There are inconveniences.

SUMMARY OF THE INVENTION It is an object of the present invention to provide a harmonic measuring apparatus capable of correctly measuring harmonics up to a high order at any time including a transition time with a simple circuit configuration.

[0011]

According to a first aspect of the present invention, there is provided a frequency measuring means for measuring a frequency of an input signal, a fixed sampling means for sampling the input signal at a predetermined fixed frequency, and m cycles (m) of the measuring frequency. > 1) N calculation means for calculating the sampling number N by rounding, and DFT calculation means for measuring a harmonic component by DFT calculation based on the data of the sampling number N.

In the present invention, the sampling method is a fixed sampling method. That is, sampling is performed at a fixed frequency even if there is a change in the input signal. The frequency of the input signal is separately measured by frequency measuring means. The frequency is measured by, for example, shaping the waveform of the input signal into a square wave through a comparator, dividing the frequency by two, and counting the clock with a high frequency (eg, 1 MHz) clock during the H level. Further, it can also be obtained by using interpolation between zero crossings of the signal waveform. This method will be described later.

A sampling number N is calculated for m cycles (m is 2 or more, for example, m = 4) of the measurement frequency obtained by the frequency measurement means. At this time, if the last position of the m cycle does not coincide with the sampling timing, the last sampling timing is determined by rounding. For rounding, for example, a rounding method can be adopted.

Here, if m = 4, and the fixed frequency of sampling is a fixed frequency at which sampling is performed at 128 points per cycle at 55 Hz, when the measurement frequency is f, N = 128 points × 4 cycles × 55 Hz / f And the last sampling point is determined by rounding. It should be noted that 55 Hz was used for 60 Hz.
This is due to the choice of the intermediate frequency so that it can be shared for both Hz and 50 Hz signals.

The N sampling data thus obtained is subjected to a DFT operation to measure the n-th (n is an arbitrary) harmonic component, that is, the harmonic content and the phase difference with respect to the fundamental signal are obtained.

According to this device, since the fixed sampling method is employed, there is no need to adjust the sampling frequency to the measurement frequency, so that there is no problem of responsiveness. Since the harmonic measurement is performed based on the data, it is possible to prevent a decrease in measurement accuracy. This enables accurate measurement even for higher-order harmonics.

According to a second aspect of the present invention, the frequency of the input signal is obtained by averaging an arbitrary cycle.

Although it is possible to extract the frequency of one cycle of the input signal to obtain the frequency, the measurement result may be deteriorated due to noise or the like. Therefore, by measuring the frequency by averaging a plurality of cycles, it is possible to prevent adverse effects due to noise and the like.

According to a third aspect of the present invention, the DFT operation means includes a table in which values of the sine wave n-order harmonic component and the cosine wave n-order harmonic component corresponding to the sampling number N within a predetermined range are stored in advance. The DFT operation is performed using this value.

In the DFT operation, as is well known, the input signal is multiplied by a sine wave harmonic component and a cosine wave harmonic component to obtain a correlation to obtain a harmonic component of each order.
The calculated values of the first-order harmonic component and the nth-order cosine wave harmonic component always correspond to the sampling number N. Therefore,
By storing those values corresponding to the sampling number N in the predetermined range in the table, the time of the DFT operation can be shortened.

According to a fourth aspect of the present invention, there is provided a device for storing data sampled by the fixed sampling device as waveform data.

When an oscilloscope measuring circuit is incorporated in the harmonic measuring apparatus, by providing means for storing data sampled by the fixed sampling means as waveform data, the sampling means can be used for both oscilloscope measurement and harmonic measurement. The configuration can be simplified.

[0023]

FIG. 1 is a block diagram of a harmonic measuring apparatus according to an embodiment of the present invention. Low-pass filter 1
a, 1b, and 1c are three-phase current detection signals (Ia) and (Ia).
b) and (Ic) are input, and low-pass filtering is performed. The multiplexer 2 selects one of the low-pass filters, samples and holds each signal by the sample and hold circuit 3, and converts the signal into digital data by the AD converter 4 at the next stage. CPU6
Executes a program stored in the ROM 7 and performs harmonic measurement by a process described later. CPU 6 has I /
A switching signal of the multiplexer 2, a sample signal of the sample and hold circuit 3, and an A / D control signal for the A / D converter 4 are output via the O port 5, and the digital data obtained by the A / D converter 4 is read. R
Expand to AM8. The operation panel 10 inputs various measurement conditions such as setting the maximum harmonic order to be measured. The printer 12 outputs a transient waveform of a predetermined order in an oscilloscope measurement mode for outputting a measurement result or outputting waveform data at the time of transition.

In the harmonic measurement mode, the CPU 6 measures the frequency of the fundamental wave of the input signal, samples the input signal at a predetermined fixed frequency, and analyzes the sampled data by a DFT operation to analyze the harmonics of each order. I do.

The frequency of the fundamental wave can be measured by a known method. For example, as shown in FIG. 2, the input signal waveform is shaped into a square wave through a comparator, this square wave is frequency-divided by 2, and the H section is counted by a high frequency clock (for example, 1 MHz). The reciprocal of the count value multiplied by the clock cycle is the frequency of the input signal. Also, as shown in FIG. 3, the fundamental wave signal is sampled in an appropriate section (in the figure, a triangular wave waveform is shown for easy understanding), and a zero-cross portion (for example, between 1-2 in FIG. 3). (N- (N + 1)) to obtain a zero cross position by performing linear interpolation. In the example shown in FIG.
It is obtained as an added value of the times of (between 2-N), (between zero-cross position-2), and (between N-zero-cross position).

The sampling fixed frequency is, for example, 5
A frequency at which 128 points of sampling data can be obtained for 5 Hz, that is, 7040 Hz is adopted. The sampling number N used for harmonic measurement is, for example, 4
Cycles. At this time, depending on the relationship between the zero-cross position in the fourth cycle and the sampling position, which position is set as the N-th sampling point changes. That is, of the two sampling points sandwiching the zero-cross position in the fourth cycle, the sampling point close to the zero-cross position is the N-th sampling point. This determination can be made by a rounding method such as rounding. For example, in the case shown in FIG. 4A, the last sampling point in the fourth cycle is the Nth one, and in the case shown in FIG. 4B, the first sampling point for the five cycles is the Nth one. It will be. Thus, by determining the N-th sampling point by the rounding method, the error can be minimized.

Based on the N pieces of sampling data obtained in this way, the harmonic components are measured by a DFT operation (division Fourier transform operation).

Note that the frequency of the input signal (the frequency of the fundamental wave component) can be obtained by an average value for a plurality of cycles. By doing so, it is possible to prevent adverse effects due to noise and the like. However, if the number of cycles is increased, an error will occur in the harmonic calculation in the transition period. Therefore, it is better to limit the number of cycles to several.

A method for calculating the sampling number N by rounding can be expressed by the following formula. N = 128 points × 4 cycles × 55 Hz / f In this equation, f is the measurement frequency, and 55 Hz / f is obtained by rounding off if it is not an integer.

The DFT calculation is performed by the following equation.

(Equation 1) Di: the number of sampling data a _n: n order cosine component amplitude b _n: n order sine component amplitude C _n: n order amplitude [Phi _n: n order phase Figure 5 is a flowchart showing a procedure for performing harmonic measurement of the It is.

In step ST1, an input signal is sampled at a fixed frequency. Subsequently, in ST2, the fundamental frequency is measured. Next, in ST3, a sampling number N corresponding to the measured frequency is obtained. Then, in ST4, a DFT operation is performed using the obtained N sampling data strings to perform harmonic measurement.

The data sampled in ST1 is recorded on the printer 12 in ST5 when the oscilloscope measurement mode is set. This oscilloscope measurement mode is used when analyzing a transient phenomenon in order to record the measured input signal as it is. In addition, it is also possible to output on a monitor instead of a printer.

As shown in FIG. 5, since the data sampled in ST1 is used for oscilloscope measurement as it is,
There is no need to prepare separate sampling means for oscilloscope measurement and harmonic measurement.

In the above DFT calculation, if the sampling number N is within a certain range, the values of all the sine wave n-order harmonic components and cosine wave n-order harmonic components within those ranges are stored in a table in advance. By remembering,
The calculation itself can be reduced. Therefore, it is conceivable that the values of the sine wave component and the cosine wave component corresponding to the sampling number N in a predetermined range are stored in a table in advance. For example, N = 4 per 4 cycles
By storing the values of the n-th harmonic component from 33 points to N = 626 points as a table, the processing time of the DFT operation can be greatly reduced.

FIGS. 6 to 8 show simulation results according to the above embodiment. The fixed frequency at which sampling is performed is 55 Hz × 128 = 7040 Hz, and the number of cycles used for harmonic measurement is four. FIGS. 6 and 7 show simulation results of two frequencies (f = 44.948 Hz and 45.020 Hz) near a measurement frequency of 45 Hz, and FIGS. 8 and 9 show two frequencies (f = 64. 9596Hz and 65.1098H
A part of the simulation result of z) is shown. If the frequency is above, the number of sampling data N
Is as follows. In this example, when the value of the fractional part of N or less is 0.5, either +1 or -1 is allowed. For convenience, the frequency N is 626.5 or 6 for convenience.
25.5 and 433.5 or 432.5.

44.948 Hz = 626.5 points ≒ = 626 points 45.020 Hz = 625.5 points ≒ = 626 points 64.9596 Hz = 433.5 points ≒ = 433 points 65.1098 Hz = 432.5 points ≒ = 433 The point harmonic shows 5th to 40th harmonic content and phase difference when 5% is contained in the 39th order and the fundamental phase and the harmonic phase are changed by 30 °. The 38th and 40th phases are indefinite due to the content rate of ≒ 0%.

As shown in FIGS. 6 to 9, it can be understood that the content and the phase can be measured fairly accurately even at the very high harmonic of the 39th order.

[0038]

According to the first aspect of the present invention, responsiveness does not deteriorate due to the fixed sampling method.
Since the sample data for a plurality of cycles is used, the accuracy of the harmonic measurement does not deteriorate. In addition, since the sampling data can be used for oscilloscope measurement as it is, the sampling circuit can also be used, and there is an advantage that the circuit configuration is simplified.

According to the second aspect of the invention, it is possible to prevent adverse effects due to noise and the like.

According to the third aspect of the present invention, the speeding up of the DFT operation can be easily realized.

According to the fourth aspect of the present invention, when the oscilloscope measuring circuit is incorporated, there is an advantage that the circuit configuration of the entire harmonic measuring apparatus can be simplified.

[Brief description of the drawings]

FIG. 1 is a configuration diagram of a harmonic measurement device according to an embodiment of the present invention.

FIG. 2 is a diagram illustrating a method for measuring the frequency of an input signal.

FIG. 3 is a diagram illustrating another method for measuring the frequency of an input signal.

FIGS. 4A and 4B are diagrams for explaining a method of obtaining a sampling number N by rounding;

FIG. 5 is a flowchart showing a schematic operation of the harmonic measuring device.

FIG. 6 is a diagram showing a simulation result.

FIG. 7 is a diagram showing a simulation result.

FIG. 8 is a diagram showing a simulation result.

FIG. 9 is a diagram showing a simulation result.

Claims (4)

[Claims] 1. A frequency measuring means for measuring a frequency of an input signal, a fixed sampling means for sampling an input signal at a predetermined fixed frequency, and a sampling number N for m cycles (m> 1) of a measured frequency is obtained by rounding. A harmonic measurement device comprising: N calculation means; and DFT calculation means for measuring a harmonic component by DFT calculation using data of the sampling number N. 2. The harmonic measuring device according to claim 1, wherein the frequency of the input signal is obtained by averaging an arbitrary cycle. 3. The DFT operation means includes a table in which values of a sine wave n-order harmonic component and a cosine wave n-order harmonic component corresponding to a sampling number N within a predetermined range are stored in advance.
3. The harmonic measurement device according to claim 1, wherein a DFT operation is performed using this value. And means for storing data sampled by the fixed sampling means as waveform data.
The harmonic measuring device according to claim 1.