JPH08331752A - Detecting method of ground fault - Google Patents

Abstract

PURPOSE: To detect a ground fault and to know the direction thereof by incrementing a condition establishment counter on the basis of a phase angle of a zero-phase current to a zero-phase voltage and by detecting the ground fault from a value of the counter. CONSTITUTION: Based on an output signal of a digital filter of a filter part 11, an arithmetic part 12 carries out computation of the respective effective values for a commercial frequency of zero-phase voltage and current, computation of a phase difference and determination of a lead/delay of the zero-phase current in relation to the zero-phase voltage, makes determination as to whether each effective value is a set value or above and as to whether a phase angle of the zero-phase current in relation to the zero-phase voltage is within a set range, and increments a condition establishment counter (f) by one only when all of these AND conditions are established, while it clears a condition failure counter (g) at the same time. A time point when the condition establishment counter (f) reaches a specified value comes to be the point of detection of a ground fault and a value set initially in the condition establishment counter (f) comes to be the time of detection of the fault.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a digital protection relay for protecting an electric power system and a method of detecting a ground fault and grasping a direction of occurrence of the accident, which is used for the relay.

[0002]

2. Description of the Related Art Conventionally, in a digital type protection relay, its input part is composed of a hardware and has an anti-aliasing filter (analog filter) and a software structure for the purpose of removing a folding error due to sampling during digital calculation. The transient DC component and harmonics are removed by serializing a digital filter consisting of
Basically, only the fundamental wave component is extracted and the subsequent protection calculation is performed. However, the sampling frequency is 600Hz
Because of this, the ability of the digital filter is limited to the removal ability up to the third harmonic, and in order to compensate for it, the dependency on the analog filter is still high.

FIG. 11 shows an example of a configuration using a conventional digital filter. Generally, the relay operation of an analog filter is delayed as its gain is increased. If a digital filter is added to this, the delay becomes more noticeable and it becomes difficult to operate the relay at high speed. The purpose of this solution is to select a filter to be used according to the presence or absence of harmonics included in the input.

Reference numeral 21 in the figure denotes an analog filter, which has low-gain analog filters AF1 to AF2 composed of a high-pass filter and a low-pass filter, and AF3 composed of a high-gain band-pass filter. Then, a digital filter unit 22 is provided on the output side of the low gain analog filter, and the output of this digital filter unit 22 and the output of the high gain analog filter are respectively applied to the respective one input terminals of the AND gate 24. Further, a judgment operation block 23 for judging presence / absence of harmonics is provided on the output side of the digital filter unit 22, and an input selection signal obtained according to the presence / absence of harmonics is applied to each other input terminal of the AND gate. Are configured to selectively open these AND gates.

[0005]

However, in the above-mentioned example, it cannot be said to be a complete digital signal processing because it relies heavily on the analog filter, and the ratio of hardware is high, and miniaturization of equipment cannot be expected. In addition, since several kinds of digital filters are required depending on the harmonic order, the load on software is heavy. Considering that the above-mentioned configurations are required for zero-phase current and zero-phase voltage, if the harmonic components included in the voltage and current are different, the phase difference between zero-phase voltage and current and the filter output Since the phase difference between the two signals is different and the phase determination becomes difficult, the correction or the like is necessary in the calculation after the AND gate, and the processing becomes complicated.

In order to eliminate these drawbacks in the present invention, the analog filter is an anti-aliasing filter only for the purpose of removing sampling folding error during digital calculation, and a band pass type digital filter is provided at the output of both filters. FIR with phase linear characteristics (F
Init Impulse Response (finite impulse response) type of basic configuration solves the above-mentioned problems and grasps the ground fault accident and its direction by using the quantized digital filter output as it is. It is intended to provide a completely digital method of doing so.

[0007]

In order to achieve the above object, the method of detecting a ground fault according to the present invention provides a zero-phase voltage and zero-phase current of a non-grounded distribution system to obtain the magnitude and phase difference of the zero-phase voltage and zero-phase current. In the method of detecting a ground fault, which detects the occurrence of a ground fault and grasps the direction of the accident when the voltage / current is above the set value and the phase difference is within the set range for more than the set time, It has an anti-aliasing filter that removes sampling aliasing error during digital calculation and a bandpass digital filter that removes frequency components other than the commercial frequency contained in the zero-phase voltage and zero-phase current, based on the output signal of the digital filter. Commercial frequency 1
/ 2 cycles as one unit, the RMS value calculation of the zero-phase voltage and zero-phase current is compared with the preset ground fault detection reference value, and the phase angle of the zero-phase current with respect to the zero-phase voltage is preset. If it is determined that both are within the range of the specified accident detection operation phase angle, and both are within the set value or more,
The unit has a counter that increments the number of the established units, assuming that the ground fault accident detection condition is satisfied, so that the time point when the number becomes equal to a preset value is the ground fault detection. .

Further, in the same method, a counter for incrementing the unit for not satisfying the condition is provided, and if the number becomes equal to or larger than a preset value, the value of the condition satisfied counter is set to 0.
If the value of the condition satisfied counter is other than 0 and the value of the condition not satisfied counter is within a preset value, the condition satisfied counter is also incremented at the same time even if the condition is not satisfied, and the occurrence and recovery of the accident are performed. Repeated intermittent ground faults are also detected in a fixed time.

[0009]

According to the present invention, the zero-phase voltage and the zero-phase current are respectively 1
It is composed of two anti-aliasing filters (hardware) and a digital filter, and one ground fault accident detection calculation unit (software) common to zero-phase voltage and current. In the anti-aliasing filter, the analog amount of zero-phase voltage or current is used as an input signal, and the purpose is to remove sampling folding error during digital calculation. Also, by using a filter of FIR type, each order included in the input is made to have a phase linear characteristic. The object is to make the filter passage time of the input signal always constant without being affected by the harmonics, and to make the phase difference between the input signal and the output signal due to the filter passage constant. As a result, the ground fault detection time is stabilized and the phase comparison between the two is stable due to the zero-phase voltage / current phase comparison principle.

In this digital filter, the output of the anti-aliasing filter is received, the FIR type is used for the same purpose as that of the filter, and the DC component superimposed on the input signal is removed.
A bandpass filter is used for the purpose of removing harmonics other than the above. With the conventional digital filter, the sampling frequency of the input signal is about 600 Hz, so sufficient ability to remove harmonics cannot be obtained, and it was necessary to rely on the analog filter in the previous stage. By devising the wear structure, the sampling frequency can be set to 4800 Hz or higher, and even the 40th harmonic or higher can be removed. According to the combination with the anti-aliasing filter in the previous stage,
It is possible to remove harmonic components and DC components with one digital filter.

In the ground fault accident calculation unit, all ground fault accident detection and direction determination are performed by software processing based on the quantized output data from the digital filter. This calculation unit calculates each effective value for the commercial frequency of the zero-phase voltage / current every 1/2 cycle, calculates the phase difference, and determines the advance / delay of the zero-phase current with respect to the zero-phase voltage, and determines each effective value. Is greater than or equal to each set value, and whether the phase angle of the zero-phase current with respect to the zero-phase voltage is within the set range. Only when all these AND conditions are satisfied, one condition satisfaction counter is set. Increment and at the same time clear the condition counter.

If the condition is not satisfied, the condition not-achieved counter is incremented by one, and when the value reaches a designated value, the condition-satisfied counter is cleared and the accident disappears. The time at which the condition satisfaction counter reaches the specified value is the time at which the ground fault is detected, and the value initially set in advance in the condition satisfaction counter is the time at which the ground fault is detected. In this way, by monitoring digitally with 1/2 cycle as one unit, the stability of the accident detection time is secured, and in principle the shortest time is 1
High-speed operation of accident detection in / 2-1 cycle is possible.

With a sampling frequency of 4800 Hz, the sampling interval is 1/4800 = 208.33.
33 μs, and assuming that the commercial frequency is 60 Hz, the phase difference detection resolution is 360 / (16.6666 ms / 208.
It has a high accuracy of 3333 μs) = 4.5 degrees.

[0014]

EXAMPLES The present invention will be specifically described below with reference to examples. FIG. 1 shows the minimum hardware configuration when the present invention is used. In the figure, 1 is an anti-aliasing filter. Since it is a circuit consisting only of a resistor R and a capacitor C, the filter response time is as fast as several ms, and by setting the cutoff frequency as low as about 160 Hz, it is assumed that it will sufficiently function to remove the sampling folding error at 4800 Hz. There is.

Reference numeral 2 denotes an analog amplifier circuit, which is composed of one operational amplifier. The MPU unit 3 is a one-chip microcomputer. The output of each operational amplifier is input to the A / D converter incorporated in the microcomputer and sampled at a sampling frequency of 4800 Hz. A
Only the commercial frequency (fundamental wave) component of the data quantized by the / D converter is extracted by the digital filter using the product-sum operation unit built in the microcomputer. After that, it is processed by the CPU and the RAM according to the procedure of the program written in the ROM. As described above, the present invention can be applied to the minimum scale hardware, and the size of the device can be reduced.

FIG. 2 shows the flow of signal processing in the present invention.
Reference numeral 11 in the figure denotes a filter unit, which is intended to extract only the commercial frequency (fundamental wave) as described above. The output from the digital filter is output in synchronization with the sampling cycle, and when the sampling frequency is 4800 Hz, the cycle is 208.3333 μs, so 1/2
8.3333 ms / 208.3333μ during the cycle
Since s = 40, the number of data will be 40. Reference numeral 12 denotes a calculation unit for detecting a ground fault accident based on the quantized data after filtering, which is a main body of software processing. In this arithmetic unit, the following processing is performed with 1/2 cycle as one arithmetic unit. a calculates the effective value based on 40 data.

[0017]

[Equation 1] Here, S _i is the i-th data in 1/2 cycle.

The respective effective values of the calculated zero-phase voltage / current are
It is compared with the preset value and it is 1 if it exceeds the set value.
However, if not exceeded, 0 enters the input terminal of the AND gate of e. Reference numeral b is a phase difference calculation unit that obtains a phase difference by simultaneously inputting zero-phase voltage / current data in synchronization with the sampling cycle.

FIG. 3 shows an example of the phase difference between zero-phase voltage and current. The element indicating the phase difference between the two waves is indicated by the polarities of the respective waves, that is, the phase difference is the period in which the polarities do not match. According to FIG. 3, when the zero-phase voltage and the current are in the same phase, the polarities all match, when there is a 180-degree phase difference, they all do not match, and when there is a 90-degree difference, the matching portion and the non-matching portion Is the same.

In the 1/2 cycle operation unit, the output from the digital filter is compared with the code bit included in both data, and the counter on the RAM is incremented only when there is a mismatch. Since 40 samplings are performed in 1/2 cycle as described above, the phase angle resolution is 180 degrees / 40 = 4.5 degrees, and at the end of 1/2 cycle (at the end of the 40th comparison from the start of comparison). Then, the phase difference is obtained by multiplying the value of the counter by 4.5 degrees. The calculated phase difference is input to the determination unit of the set phase angle range of d.

Reference numeral c denotes a direction (advance / delay) determination calculator for the zero-phase current with respect to the zero-phase voltage. Each time the zero-phase voltage and current data are input, the sign bit is checked for each and the result is compared with the immediately preceding sign bit to check whether the polarity has changed. When there is a change in the polarity of either the voltage or the current, it is possible to determine whether the lead or the lag by looking at the polarity of the other data. The judgment conditions are shown in Table 1.

[0022]

[Table 1] In Table 1, * indicates the case where there is no change, and since it is not possible to determine whether it is a lead or a lag, it is decided to lead.

In this case, the phase difference is either in-phase or 180-degree difference. In the phase difference calculation of b, if the phase is in-phase, the calculated phase difference is 0 degree, and if it is 180-degree difference. , The calculated phase difference is also 180 degrees, and can be easily distinguished. The result here is input to the determination unit of the set phase angle range of d. d is a determination part of the set phase angle range. The phase angle of the zero-phase current with respect to the zero-phase voltage is obtained by inputting the phase difference calculation result of b and the direction determination result of c.

FIG. 4 shows an example of the set operating phase angle range. If the judgment result of c is advanced, it is judged whether the phase difference calculation result of b is within the load side ground fault accident range of FIG. If the determination result of c is delayed, it is determined whether the phase difference calculation result of b is within the power source side ground fault accident range of FIG. The judgment result and the preset load side /
Power supply side Compared with the information indicating in which direction the ground fault accident should be detected, if they match, the output of e to the AND gate is set to 1.

The operating phase angle range of the protection relay shown in FIG. 4 should be a range that sufficiently covers the operating phase angle range if it is attempted to cooperate with the substation protection relay. By providing the dead zone shown in the figure, it is possible to eliminate erroneous determination, and since the phase angle is handled in digital form, stable characteristics can be obtained as a protection relay. e is a judgment unit of the AND result, which is 3 of AND gates.
When all one inputs are 1, the output becomes 1, and the 1/2 cycle which is the object of determination means that the ground fault accident detection condition is satisfied. If even one of the three inputs is 0, the condition is not satisfied. When the condition is satisfied, the g counter is incremented.

Here, f is a condition satisfaction counter and g is a condition satisfaction counter. When the condition satisfaction counter of f becomes a preset value, it becomes the ground fault accident detection, and the preset value becomes the monitoring time from the occurrence of the ground fault accident to the detection in 1/2 cycle (8.3333 ms) units. It can be set.

When the ground fault accident is assumed to be an intermittent ground fault in which occurrence and disappearance are repeated and the fault occurs again in a preset value (time) after disappearance of the ground fault,
It is considered that the accident continued while the erasure disappeared, and the condition satisfaction counter of f is incremented every 1/2 cycle. However, if it has disappeared for more than the set value (time), the condition satisfaction counter of f is cleared. When a ground fault is detected, both the counters f and e are cleared, and
The operation unit 12 does not function until there is an instruction for re-monitoring.
Here, if time coordination with the substation protection relay is to be taken, f and g may be selected from preset values.

FIGS. 7 to 10 are oscilloscope waveform diagrams at the time of high-voltage verification test. Zero-phase voltage detection sensitivity of the protection relay is 160 at a charging current of 5 A and a ground fault resistance of 5 Ω.
V, zero phase current detection sensitivity 0.34A, detection time 117m
s, the zero-phase voltage waveform V0, the zero-phase current waveform I0, and the detection signal when the ground fault on the load side of the protection relay is set to be detected. Here, FIG. 7 shows a metallic ground fault, FIG. 8 shows an intermittent ground fault, FIG. 9 shows a cable ground fault due to a pinhole, and FIG. 10 shows a ground fault due to a porcelain crack, and there is little variation in detection time depending on the type of ground fault. I understand. The detection time of the oscilloscope is added with a mechanical delay of about 4 ms because the detection signal is output to the oscilloscope.

[0029]

As described above, according to the present invention, the hardware required for the ground fault accident detection function is as simple as shown in FIG. 1, and the downsizing of the equipment, the cost reduction and the reduction of the number of parts are achieved. It is possible to improve reliability. Also, except for the anti-aliasing filter and the analog amplifier circuit, all are processed by software, so there is little effect on changes in ambient temperature.

Moreover, since the detection time, phase angle, etc. are all handled as digital quantities, detection that depends on the magnitude of the zero-phase voltage and current that the conventional protection relay has and variation in the recovery time, and detection that depends on the phase difference between the two. Variations in sensitivity and non-linearity of operating phase angle are eliminated, and constant characteristics can always be obtained.

Accordingly, the operation coordination with the substation protection relay is facilitated, and the power failure in the healthy section is achieved by transmitting the protection relay information on the distribution line and the substation protection relay information to the computer installed in the sales office in real time. It is possible to construct an advanced distribution line accident handling system that separates only the ground fault accident section.

[Brief description of drawings]

FIG. 1 is a block diagram showing a hardware configuration example of a ground fault accident detection unit of a protection relay using the present invention.

FIG. 2 is a block diagram showing a flow of signal processing in the present invention.

FIG. 3 is a waveform diagram showing an example of a phase difference between zero-phase voltage and current.

FIG. 4 is an explanatory diagram showing an example of a set operating phase angle range.

FIG. 5 is an explanatory diagram showing a frequency response of a digital filter.

FIG. 6 is an explanatory diagram showing a step response of a digital filter.

FIG. 7 is an oscilloscope waveform diagram at the time of a high-voltage demonstration test of a metallic ground fault.

FIG. 8 is an oscilloscope waveform diagram at the time of a high voltage demonstration test of an intermittent ground fault.

FIG. 9 is an oscilloscope waveform diagram at the time of a high voltage verification test of a cable ground fault due to a pinhole.

FIG. 10 is an oscilloscope waveform diagram at the time of a high-voltage verification test of a ground fault due to insulator cracking.

FIG. 11 is a block diagram showing an example of a ground fault accident detection device using a conventional digital filter.

[Explanation of symbols]

1 anti-aliasing filter, 2 analog amplifier circuit, 3 MPU section, 11 filter section, 12 arithmetic section

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Vice Hironobu 2-82 Watanabe-dori, Chuo-ku, Fukuoka City, Fukuoka Prefecture Kyushu Electric Power Co., Inc.

Claims (2)

[Claims] 1. The magnitude and phase difference of the zero-phase voltage and zero-phase current of an ungrounded power distribution system are calculated, and the zero-phase voltage and current are equal to or greater than a set value and the phase difference is within a set range for a set time or more. In a ground fault detection method that detects the occurrence of a ground fault and determines the direction of the fault if it continues, an anti-aliasing filter that eliminates sampling aliasing error during digital calculation and zero-phase voltage and zero-phase current Has a band-pass digital filter that removes frequency components other than the commercial frequency included in 1), and 1/2 cycle of the commercial frequency is set based on the output signal of the digital filter.
As a unit, the effective value calculation of the zero-phase voltage and zero-phase current is compared with the preset ground fault accident detection reference value, and the phase angle of the zero-phase current with respect to the zero-phase voltage is preset to the accident detection operation phase. If it is within an angular range, and if the magnitude comparison result and the phase angle range are equal to or greater than a set value, the unit is determined to be the ground fault accident detection condition, and the number of the established units is determined. A ground fault accident detection method, which is characterized in that a ground fault accident is detected when a condition is satisfied by a counter, and the number of times when the count is equal to a preset value. 2. A unit of condition not satisfied is counted by a condition not satisfied counter, and when the number is equal to or greater than a preset value, the value of the condition satisfied counter is reset to 0, and the value of the condition satisfied counter is other than 0. If the value of the unsatisfied counter is within a preset value, the condition-satisfied counter is also incremented at the same time even if the condition is not satisfied. The ground fault accident detection method according to claim 1.