JPH10322246A - Distribution line carrying signal receiving method using wavelet transformation - Google Patents

Abstract

PROBLEM TO BE SOLVED: To exactly decode a signal even when carrier frequency is coincident with noise frequency. SOLUTION: In a distribution line carrying signal receiving method to receive a modulated high frequency carrying signal superimposed on AC power supply wave of a distribution line, the signal is decoded by separating the carrying signal from the AC power supply wave by a band pass filter 3, generating a transformation pattern by performing wavelet transformation to waveform of the carrying signal by a wavelet transformer 10 and collating the transformation pattern with a predetermine reference signal pattern by a pattern collating device 13.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an improved distribution line carrier signal receiving method for receiving a modulated high-frequency carrier signal superimposed on an AC power supply wave of a distribution line.

[0002]

2. Description of the Related Art In a conventional distribution line carrier signal transmission in which a signal wave is superimposed on an AC power supply wave of a distribution line to perform communication, compared to a signal transmission performed using a dedicated communication line, a signal present in the distribution line exists. Since it is directly affected by noise, a filter that passes only a specific frequency band is used to remove the effect of noise existing in the power distribution system, so that desired communication performance is obtained.

[0003]

In the conventional signal detection method using a filter, the presence of a signal is determined by focusing only on the frequency at which the signal is to be conveyed. When the frequencies are close to each other, it is very difficult to determine the signal for the noise.

A first object of the present invention is to provide a distribution line carrier signal receiving method using a wavelet transform, which can accurately decode a signal even when the carrier frequency and the noise frequency match. It is.

A second object of the present invention is to provide a distribution line carrier signal receiving method using a wavelet transform that can cope with signal fluctuations.

A third object of the present invention is to provide a distribution line carrier signal receiving method using a wavelet transform, which can accurately decode a signal even if a state of a distribution line or a transmission side changes. That is.

[0007]

In order to achieve the first object, the present invention according to claim 1 receives a modulated high-frequency carrier signal superimposed on an AC power supply wave of a distribution line. In the distribution line carrier signal receiving method, a carrier signal is separated from an AC power supply wave, a wavelet transform is performed on a waveform of the carrier signal to generate a conversion pattern, and the conversion pattern is compared with a predetermined reference signal pattern. And decoding the signal.

In order to achieve the second object,
According to a second aspect of the present invention, prior to matching the conversion pattern with the reference signal pattern, the conversion pattern is weighted in accordance with a region of interest of the reference signal pattern. .

Furthermore, in order to achieve the third object, the present invention according to claim 3 is characterized in that the weighting is selected according to a decoding result of a signal.

[0010]

DESCRIPTION OF THE PREFERRED EMBODIMENTS FIG. 1 is a block diagram of an example of a distribution line carrier receiving apparatus using a wavelet transform for implementing the method of the present invention.

In FIG. 1, reference numeral 1 denotes a distribution line of commercial power in which a signal wave is superimposed on a power supply wave, and a power supply waveform and a signal waveform are extracted through an insulating filter 2 for electrically isolating a power distribution system and a receiver. And the signal waveform passes through a band-pass filter 3 that passes near the carrier frequency.
The signal is amplified by the amplifier 4 to an appropriate level.

On the other hand, a power supply waveform is amplified by an amplifier 5 to generate a power supply sine wave without distortion.
Oscillation is forcibly synchronized with the power supply frequency. With this oscillation waveform, the synchronous pulse generator 7 outputs a trigger pulse for capturing a signal to the A / D converter 8, and the A / D converter
The converter 8 sequentially stores the digitized received signals in the signal memory 9 at a fixed sampling interval. When the signal memory 9 becomes full, the wavelet converter 10 starts the conversion operation and stores the conversion result in the collation memory 11.

The result of wavelet conversion of a noise-free signal is previously written in the signal pattern memory 12 as a reference signal pattern, and the pattern matching unit 13 converts the reference signal pattern of the signal pattern memory 12 into the converted signal of the comparison memory 11. The pattern is compared with the pattern, and if the contents match, it is determined that there is a signal, and if the contents do not match, it is determined that there is no signal, thereby confirming whether or not the transmitting side has transmitted a signal to this position, and outputting a signal bit string. The portion surrounded by the dashed line is actually constituted by the microcomputer COM.

When signal transmission is performed by the distribution line carrier system, a burst signal F (t) modulated by a digital signal and using a carrier frequency f as shown in FIG. 2 is often used as a carrier signal. As the characteristic amounts of this waveform, three typical elements are a frequency f, a burst width T, and a carrier phase φ. With respect to such a burst signal F (t), frequencies a,
Multiply a time width T '(which does not have to be equal to the burst width T of the burst signal) by a wavelet (wave packet) W (t) having a shift amount b, and sequentially convolution this on the time axis while changing the shift amount b. The integration (convolution operation) defines the wavelet intensity (integral value) on the time-frequency plane. By connecting points having the same intensity, the frequency after the wavelet transform as shown in FIG. As an ab plan view of a versus the shift amount b, a contour line concentric pattern unique to the signal is obtained. Note that the shift amount b in FIG.
It represents the time to the center of (t).

FIG. 4 is a diagram showing a reference signal pattern obtained as a result of a wavelet transform of a noise-free burst signal. FIG. 5 is a diagram illustrating an example in which a signal and noise are separated by wavelet transform. FIG. 5 (a)
Is a signal waveform on the time axis, and even when there is noise as shown in FIG.
As shown in FIG. 5 (b), when viewed from an ab plane view of the frequency versus shift amount after the wavelet transform, the conversion pattern Ws for the left signal is wide and has a specific frequency band. , Whereas the noise conversion pattern Wn after the wavelet conversion on the right has a narrow width and deviates from the frequency band, so that the pattern matching unit 13 By comparing the reference signal pattern with the conversion pattern in the matching memory 11, only the signal can be accurately discriminated and detected.

FIG. 6 is a diagram showing an example in which the signal and the noise are separated by the wavelet transform when there is almost no level difference between the signal and the noise.

FIG. 6 (a) shows a case where sine wave noise having almost no level difference from the signal exists before and after the signal. If this is simply processed by a low-pass filter, FIG. As shown in b), it is difficult to separate signals even when the threshold level is changed. However, FIG.
As a result of the wavelet transform of the signal and noise of (a), the signal is discriminated as a conversion pattern Ws of a concentric specific frequency band as shown in an ab plan view of the frequency versus shift amount in FIG. On the other hand, noise appears as a (frequency) = constant completely different shape conversion pattern Wn.
As a result, it is possible to accurately receive only the signal even when the sine wave noise level is high.

In the distribution line conveyance, a carrier frequency of 300
I want to use ~ 400 Hz, but the noise frequency on the distribution line is a harmonic of 50 or 60 Hz, which matches the carrier frequency band. In such a case, the present receiving method is extremely effective.

FIG. 7 is a block diagram showing a case where a fuzzy template (a fuzzy matching filter with a clear boundary) is added to the distribution line carrier receiving apparatus using the wavelet transform shown in FIG. 1 in order to cope with signal fluctuations. is there.

Since an error is unavoidable in the carrier frequency of the carrier signal transmitted from the transmitting side, it is necessary to provide a certain width for discriminating the presence or absence of the signal. For this purpose, the fuzzy template 14 shown in FIG. And a fuzzy pattern collator 15 are used.

The fuzzy template 14 is, for example,
A fuzzy matching filter representing a weighted membership function having a two-dimensional spread as shown in FIG. 8, which weights the conversion pattern of the real signal stored in the matching memory 11, that is, matches all the reference signal patterns. It is not necessary to perform the comparison, and only the characteristic region of interest needs to be compared. Therefore, the portion of the region of interest of the conversion pattern of the actual signal is extracted.

The fuzzy template 14 is shown in FIG.
The shape is determined by using the length in the AA ′ direction, the length in the BB ′ direction, and the center position as parameters, and the setting of these parameters is first determined according to the attention area of the reference signal pattern Approximately determined, then manually adjusted to best experimentally. For example, in an actual device, the reference signal pattern and the conversion pattern of the actual signal completely match even in a situation where there is no noise due to a reference clock error, an analog-digital conversion calculation error, and other various factors. Do not. Further, since the power supply frequency is not constant, even in an ideal case, the conversion pattern of the actual signal becomes a figure having fluctuation around the reference signal pattern. If the weighting by the fuzzy template 14 is set so as to cover a wide area, the pattern will not be missed even with large fluctuations, but instead, the two kinds of slightly different patterns cannot be distinguished. Conversely, if the width of the weighting area is set to be narrow, even a slight fluctuation will result in a different signal being determined. Since the optimum weighting shape does not have a property that can be calculated in advance, it is experimentally and practically adjusted and determined so as to reduce the decoding error of the original signal.

As shown in FIG. 9, the fuzzy pattern matching unit 15 performs weighting by performing a fuzzy AND operation on the membership function of the fuzzy template 14 and the conversion pattern of the real signal, and performs a fuzzy template The integration in the fuzzy template 14 (three-dimensional volume) with the reference signal pattern in the signal pattern memory 12 is performed by performing integration in the signal pattern memory 12. If the degree of coincidence is larger than a specific value, a signal is present, otherwise a signal is absent, so that the signal can be discriminated even when there is some deviation in the carrier frequency on the transmitting side.

In FIG. 8, in the case of 1-bit collation, only the template on the left side of FIG. 8A is used, and in the case of collating 2-bit (1, 1), [1, 1] shown in FIG.
1) is used. It is also possible to match three or more bits at once, in which case the number of templates increases.

FIG. 10 shows a distribution line carrier / reception device using the wavelet transform shown in FIG. 7 for automatically correcting the parameters of the fuzzy template 14 for long-term fluctuations of the distribution line which is a signal line. FIG. 13 is a block diagram in a case where an adaptive processing function by a genetic algorithm is added.

The shape of the above-mentioned fuzzy template 14 is determined by using the length in the AA 'direction, the length in the BB' direction and the center position in FIG. 8A as parameters. The optimum value is not fixed, and needs to be changed depending on the state of the distribution line and the manufacturing error of the transmission device.

For this purpose, the transmitting side transmits a signal including the same combination of bit strings every time the signal is transmitted, and the receiving side prepares a plurality of templates each having slightly different parameters. That is, the genetic algorithm execution engine 16 in FIG. 10 performs a test on the reproducibility of the bit string for all templates each time a block signal having 8 bits as one block is received, and the parameter (member One set of (ship function) is placed at the top of the array in the fuzzy template array 17, and a set of parameters having poor reproducibility for a fixed pattern bit sequence is discarded. Also, for some parameter sets, some of their values are sometimes exchanged (referred to as mating), and for some parameter sets,
The parameter is changed to a parameter slightly different from the previous parameter by random number processing by the microcomputer. By repeating such an operation, the membership function of the fuzzy template 14 can always take an optimal value.

The method of the present invention is not limited to frequency modulation.
It can be applied to other modulation schemes.

[0029]

As described above, according to the first aspect of the present invention, the shape of the temporal change of the signal simultaneously with the frequency change of the signal is extracted as the characteristic amount of the signal by using the wavelet transform. Thus, even when the carrier frequency and the noise frequency match, the signal can be decoded accurately.

According to the second aspect of the present invention, the wavelet conversion pattern is weighted in accordance with a predetermined region of interest of the reference signal pattern, thereby coping with signal fluctuation. be able to.

Further, according to the third aspect of the present invention,
Since the weight is selected according to the decoding result of the signal, even if there is a change in the state of the distribution line or the transmission side,
The signal can be decoded correctly.

[Brief description of the drawings]

FIG. 1 is a block diagram of an example of a distribution line carrier receiving apparatus using a wavelet transform that implements the method of the present invention.

FIG. 2 is a diagram illustrating an example of a burst signal.

FIG. 3 is a diagram illustrating an example of a waveform of a wavelet.

FIG. 4 is a plan view of a frequency versus shift amount of a conversion pattern after performing a wavelet conversion on a burst signal.

FIG. 5 is a diagram illustrating an example in which a signal and noise are separated by a wavelet transform.

FIG. 6 is a diagram illustrating an example in which a signal and a noise are separated by a wavelet transform when there is almost no level difference between the signal and the noise.

7 is a block diagram in a case where a fuzzy template is added to the distribution line carrier receiving device of FIG. 1 in order to cope with signal fluctuations.

FIG. 8 is a diagram showing an example of a fuzzy template of FIG. 7;

FIG. 9 is a diagram illustrating an example of weighting of an actual signal using a fuzzy template.

FIG. 10 adds an adaptation processing function using a genetic algorithm to the distribution line carrier receiving apparatus of FIG. 7 for automatically correcting the parameters of a fuzzy template for long-term fluctuations of a distribution line as a signal line. It is a block diagram in case.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 Distribution line 2 Insulating filter 3 Bandpass filter 4 Amplifier 5 Amplifier 6 Power supply synchronous oscillator 7 Synchronous pulse generator 8 A / D converter 9 Signal memory 10 Wavelet converter 11 Matching memory 12 Signal pattern memory 13 Pattern matching device 14 Fuzzy template Reference Signs List 15 fuzzy pattern collator 16 genetic algorithm execution engine 17 fuzzy template array a frequency b shift amount COM microcomputer F (t) burst signal f frequency T, T ', T ₀ time width W (t) wavelet Ws signal conversion pattern Wn noise conversion pattern φ phase

Claims (3)

[Claims] 1. A distribution line carrier signal receiving method for receiving a modulated high-frequency carrier signal superimposed on an AC power supply wave of a distribution line, wherein the carrier signal is separated from the AC power supply wave to form a waveform of the carrier signal. A method for receiving a distribution line carrier signal using a wavelet transform, wherein a transform pattern is generated by performing a wavelet transform, and the signal is decoded by comparing the transform pattern with a predetermined reference signal pattern. . 2. The wave according to claim 1, wherein before comparing the conversion pattern with the reference signal pattern, weighting is performed on the conversion pattern in accordance with a region of interest of the reference signal pattern. A distribution line carrier signal receiving method using let transform. 3. The distribution line carrier signal receiving method using wavelet transform according to claim 2, wherein the weighting is selected according to a signal decoding result.