KR100892260B1 - Electric fire prevention equipment - Google Patents

Abstract

The present invention relates to a device capable of preventing an electric fire by measuring a current and a voltage signal on an electric line to detect a short circuit, a short circuit, an arc, and a tracking. The apparatus includes: a voltage detector for detecting a voltage signal on a line from a voltage sensor; A current detector for detecting a current signal on a line from the current sensor; Calculates one or more of an effective value, a delay time, a harmonic, and a shoulder value from the digitally converted signal, and compares the slope of the effective value of the current with a reference value to determine whether a short circuit occurs; Determination of whether or not tracking occurs by comparing the harmonic growth rate with the reference value in the short-circuit state, and determining the arc occurrence by comparing the effective value of the current, the width of the shoulder, the delay time, and the harmonic growth rate with each reference value. A signal processor for performing any one or more of the following functions; A display unit which displays a processing result of the signal processor; A transmitter for remotely transmitting the processing result of the signal processor; And an alarm unit for generating an alarm when at least one of short circuit, short circuit, tracking, and arc occurs. When comparing the harmonic increase rate with the reference value when determining whether the tracking and the arc is generated, it is characterized by comparing the magnitude of the high-frequency components extracted through the wavelet transform with the Dobitis function as a reference value and the reference value. Identify in advance whether short circuits, short circuits, arcs, and tracking are the direct causes of a fire, and accurately and quickly extract features and determine abnormalities from waveforms containing various types of harmonics, especially in arc and tracking detection. It has the effect of doing it.

Electricity, fire, short circuit, short circuit, arc, tracking, harmonic, wavelet transform, doviches

Description

Electric fire prevention equipment

1 is a system block diagram of an electric fire prevention apparatus according to the present invention.

2 is a flow chart showing a line abnormality detection method of the electric fire prevention apparatus according to the present invention.

3 is a diagram illustrating a conversion process for wavelet decomposition according to the present invention.

4 is a diagram illustrating a conversion process for wavelet synthesis in the present invention.

5 is a flowchart illustrating a tracking detection method according to the present invention.

6 is a flowchart illustrating an arc detection method according to the present invention.

** Description of symbols for the main parts of the drawing **

11: voltage detector 12: current detector

20: A / D conversion unit 30: signal processing unit

31: memory 41: display unit

42: transmitter 43: alarm

The present invention relates to an electric fire prevention device, and more particularly, to a device capable of preventing an electric fire by measuring a current and a voltage signal on a line to detect a short circuit, a short circuit, an arc, and tracking.

Fires can be caused by cigarettes, arson, fire, or fire, but the biggest cause is electricity, especially when there is a fire caused by electricity in crowded places such as hotels or department stores. Since damage is caused, the importance of preventing fire by electricity cannot be emphasized enough. The main causes of such electric fires are mainly classified into short circuits and short circuits caused by short circuits and overloads and poor insulation and poor wiring, and arcs due to poor contact. Among them, overcurrent and leakage can be prevented by installing overcurrent and leakage circuit breaker, but devices that can prevent fire by tracking and arc have not been put to practical use. In general, the damage to the coating material of the indoor wiring is moist and dirty, and the wire coating is deteriorated naturally, or the leakage current flows through the wire coating material between the wires or between the wires and the ground under the influence of the surroundings. When leakage current flows between the conductors through the wire covering material, the insulating layer between the conductors is carbonized by microdischarge to form a carbonized conductive path, and the leakage current gradually increases along the carbonized conductive path, finally tracking between the conductors. ) With destruction, the wire covering material is ignited.

However, the magnitude of the current ignited by the tracking breakage is only several amperes (A) at several tens of milliamps (mA), whereas the rated current of the circuit breaker is 20 amperes (A) or more, so that the ignition current due to tracking cannot be detected. There is this. In general, fast Fourier transform (FFT) is used to analyze waveforms of current or voltage, but it is difficult to detect structural changes in waveforms on the time axis, which is inappropriate for detecting time-varying tracking leakage. It is not available.

It is an object of the present invention to provide a device capable of preventing short circuits, short circuits, short circuits, arcs and tracking, which are direct causes of an electric fire, by measuring current and voltage signals on a line in advance and generating an alarm to prevent an electric fire.

Another object of the present invention is to determine the presence of arc and tracking through the wavelet transform by measuring a fine current waveform when the arc and tracking occurs, and to classify and detect the presence or absence of abnormality by time series analysis by the waveform DB construction To provide.

Electric fire prevention device according to the present invention for achieving the above object,

A voltage detector for detecting a voltage signal on a line from the voltage sensor;

A current detector for detecting a current signal on a line from the current sensor;

Calculates one or more of an effective value, a delay time, a harmonic, and a shoulder value from the digitally converted signal, and compares the slope of the effective value of the current with a reference value to determine whether a short circuit occurs; Determination of whether or not tracking occurs by comparing the harmonic growth rate with the reference value in the short-circuit state, and determining the arc occurrence by comparing the effective value of the current, the width of the shoulder, the delay time, and the harmonic growth rate with each reference value. A signal processor for performing any one or more of the following functions;

A display unit which displays a processing result of the signal processor;

A transmitter for remotely transmitting the processing result of the signal processor; And

An alarm unit that generates an alarm when one or more of short circuit, short circuit, tracking, and arc occur; Characterized in that comprises a.

Here, in comparing the harmonic increase rate with the reference value when determining whether the tracking and arcing are generated, it is preferable to compare the magnitude of the high frequency component extracted through the wavelet transform using the Dobitice function as the reference value.

Hereinafter, with reference to the accompanying drawings will be described the configuration and operation of the electric fire prevention device according to an embodiment of the present invention.

1 is a system block diagram of an electric fire prevention apparatus according to the present invention.

When the device starts to operate, each component's function is set, the buffer is initialized, and the sensor accepts voltage and current signals that are measured in real time.

The voltage detector 11 is a voltage input circuit for determining whether a line is abnormal by detecting a change in voltage input from a voltage sensor.

The current detector 12 is a current input circuit for determining whether a line is abnormal by analyzing a current waveform input from a current sensor.

The voltage detector 11 and the current detector 12 convert a voltage / current amplifier that amplifies the voltage / current signal detected by the sensor and an input voltage / current signal into a signal corresponding to the rating of the A / D converter 20. It consists of a bias circuit that adjusts.

A filter unit (not shown) removes a noise signal from the voltage and current signals.

The A / D converter 20 converts the detected analog voltage signal and current signal into a digital signal and transmits the converted digital signal to the signal processor 30.

The signal processor 30 calculates an input value and a frequency of a voltage and current waveform converted into a digital signal, and compares the result with a reference set value to determine whether a short circuit, a short circuit, an arc, and a tracking occurs. The memory 31 is composed of RAM, FLASH, and the like and stores the set value, a program for calculation and comparison.

The processing result of the signal processing unit is displayed on the display unit 41 such as an LCD or transmitted remotely through the transmission unit 42 configured as a communication port. The line manager may selectively search for information displayed on the display unit through key manipulation, and may remotely monitor the line state by receiving data transmitted through the transmitter to a computer.

On the other hand, when a line abnormality is detected due to a short circuit, a short circuit, an arc, or a tracking occurrence, the line manager can easily recognize that the abnormality has occurred by operating an alarm unit 43 composed of a lamp or a buzzer as an information relay contact output. Make sure

2 is a flow chart showing a line abnormality detection method of the electric fire prevention apparatus according to the present invention.

The signal processor 30 obtains an effective value, a duration, a delay time, a harmonic, a shoulder size, and the like from the data converted into a digital signal, and then determines whether the line is abnormal in the following manner. Here, the shoulder is a waveform having a shape similar to the shape of a human shoulder, and refers to a form of an intrinsic waveform that appears when an arc is generated.

In the case of a short circuit, a current of tens to hundreds of amperes flows in a short duration, and then the slope of the current is tracked within the sampling time, and then the short circuit is determined by comparing with the set values S1 and S2.

When a short circuit occurs, a leakage current (Io) of several milliamperes or less is generated, and it is compared with the set value S5 to determine whether there is a short circuit.

If the increase in harmonics occurs at the same time when determining that the leakage current is determined by tracking.

The arc is an input parameter using the effective value of the current, the width of the shoulder, the delay time, etc., and the arc is determined when the effective value of the current decreases, the width and the delay time of the shoulder are larger than the set value, or the harmonics increase within the sampling period. The harmonics and shoulder widths are calculated to detect the direction of the arc and set the arc reference.

In order to determine the harmonic increase in tracking and arc determination, the present invention uses an algorithm through a wavelet transform. That is, since the tracking and arc signals have fine signal components, it is impossible to determine whether they are generated by effective values, and the data are processed by an analysis algorithm using wavelet transform having characteristic elements of frequency and time components. The wavelet transform can also analyze current waveforms on the time and frequency axes to interpret the time series distribution of the waveforms, thus identifying weak tracking leakage.

In order to analyze the arc and tracking signals using discrete wavelet transforms, the choice of mother function and the stage of multiple decomposition must be determined. In the present embodiment, Dobechies is selected as a mother function in consideration of the ability to extract features and remove harmonic components from current signals of arc and tracking. In addition, the decomposition step is influenced by the sampling frequency (Fs) of the original signal. In this embodiment, the filter bank is used to divide the high frequency filter component into three frequency bands (Fs / 2, Fs / 4, Fs / 8). That is, the discrete signal is divided into signals A1, A2 and A3 passing through the low pass filter L and signals D1, D2 and D3 passing through the high pass filter H using discrete wavelet transform. The signal passing through the low pass filter (L) is a high-frequency low-frequency component, and the unique characteristics of the arc signal show only the changed state of the waveform under each condition when changing the size and resolution. On the other hand, the signal passing through the high pass filter (H) is a low-scale (low-scale) high-frequency components, and represents the fine characteristics of the abnormal signal and provides important information in signal analysis.

The wavelet inverse transform reconstructing the decomposed arc signal into the original signal passes through the decomposed signal at H 'and L', which are mirror filters of the high pass filter H and the low pass filter L, as shown in FIG. To do it.

In the present embodiment, the mother function is selected as db 5 having five nodes among the dobiches capable of discrete wavelet analysis, and the sampling frequency is selected as 256 sample / cycle. The three-step decomposition extracts the low-frequency regions D1, D2 and D3 and the high-frequency region A3. When the current of each phase to be measured increases beyond the reference, the magnitude of the high frequency components of D1 extracted through the wavelet transform is compared.

5 is a flowchart illustrating a tracking detection method according to the present invention.

The current input from the current sensor is discrete wavelet transform after waveform filtering. The discrete wavelet transform applies a wavelet transform equation by applying a scaling function, and extracts the characteristics of the current waveform in real time by finding the low frequency component and the high frequency component of the input signal using the wavelet function. Wavelet coefficients, average values, dispersion values, skewness, and the like may be calculated as data representing characteristics by tracking. Analyze each unique waveform for tracking leakage power source and type it in advance, and analyze the extracted waveform on the basis of the built-in tracking DB to diagnose the cause of tracking occurrence. In addition, it is stored in the tracking DB for tracking waveform analysis to be input in the field in the future.

FIG. 6 is a flowchart illustrating an arc detection method according to the present invention, and shows a fault detection by comparing the magnitude of the arc fault current with the magnitude of the D1 component of each phase extracted through wavelet transform.

When an arc occurs, the following three characteristics are shown. Using this point, the arc current is classified and the waveform is analyzed.

1) When arc occurs, frequency component between specific bands is increased by several times more than normal case.

2) When the arc continuously occurs, it forms a value smaller than the current effective value.

3) Shoulder discharge occurs near zero point due to continuous discharge.

The D1 component is sensitive to the arc signal, enabling faster fault detection.

As shown in Equation 1, if the sum of the absolute values of one period D1 component exceeds the set value, it is determined as an event state (arc generation). The set value k is arbitrarily adjustable here.

When the event state, as shown in Equation 2, the size of the D1 component of each phase is compared, and if the size of one phase in the RST phase is larger than the size of the other two phases to determine the abnormal (abnormal) state. The set value k was chosen here. However, in order to minimize the malfunction due to temporary load fluctuation or operation of open / close switch, when the sample is higher than the set value, 32 samples are maintained.

In addition, when the measured current waveform is accumulated and monitored, a pattern formed by accumulating signals may be analyzed. By analyzing the cumulative pattern, it is possible to determine whether the current arc is generated, the frequency of occurrence, and the size of the generated arc. By analyzing the shape of a particular phase, it is possible to provide a more reliable solution by determining the occurrence frequency and the size of the generated arc as well as the presence of the arc.

Electric fire prevention device according to the present invention by measuring and detecting the current and voltage signal on the line to determine in advance whether the leakage, arc and tracking that directly causes the electric fire to be able to effectively deal with the fire risk .

In addition, the detection method of discrete wavelet transform of frequency spectral components in arc and tracking detection enables accurate and rapid extraction of features and determination of abnormalities from waveforms containing various types of harmonics.

Claims (5)

A voltage detector for detecting a voltage signal on a line from the voltage sensor; A current detector for detecting a current signal on a line from the current sensor; An A / D converter converting the detected analog signal into a digital signal; Calculates one or more of an effective value, a delay time, a harmonic, and a shoulder value from the digitally converted signal, and compares the slope of the effective value of the current with a reference value to determine whether a short circuit occurs; Determination of whether or not tracking occurs by comparing the harmonic growth rate with the reference value in the short-circuit state, and determining the arc occurrence by comparing the effective value of the current, the width of the shoulder, the delay time, and the harmonic growth rate with each reference value. A signal processor for performing any one or more of the following functions; And It comprises a display unit for displaying the processing result of the signal processing unit, In comparing the harmonic increase rate with the reference value when determining whether the tracking and arcing occurs, the magnitude of the high frequency component extracted through the wavelet transform is compared with the reference value. delete The electric fire prevention apparatus according to claim 1, wherein the dobitice function is a mother function of wavelet transform. The electric fire prevention apparatus according to claim 1, further comprising a transmission unit for remotely transmitting the processing result of the signal processing unit. The electric fire prevention apparatus according to claim 1, further comprising an alarm unit for generating an alarm when at least one of short circuit, short circuit, tracking, and arc occurs.

Images