JPS6350932B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a ground fault detection device used in systems such as high resistance grounding systems or ungrounded distribution lines.

Conventionally, there has been a device of this type as shown in FIG. In the figure, 1a, 1b, 1c are 3-phase balanced power supplies, 2a, 2b, 2c are 3-phase distribution lines connected to this 3-phase balanced power supply, and 3a, 3b, 3c are 3-phase distribution lines, respectively. Ground capacitance with 4
5a, 5b, 5c are voltage dividers for detecting the ground voltage of each phase of the distribution line; 6a, 5c are voltage dividers for detecting the ground voltage of each phase of the distribution line;
6b and 6c are operational amplifiers that perform addition, and 7 is a grounding resistor at the neutral point of the power supply.

Next, the operation will be explained. As shown in Figure 1,
When phase a has a ground fault through resistor Rg at ground fault point 4, the ground potential changes from O to O' as shown in the vector diagram in Figure 2, and the voltage to ground of each phase after the ground fault is v _a , It becomes v _b , v _c . If the capacitance and ground fault resistance of the line change at the time of an accident, point O' moves on the circle diagram 8. Voltage dividers 5a, 5b, 5 in FIG.
The voltages v _a , v _b , v _c detected by c are added to the adder 6
By passing through a, 6b, and 6c, the following output is obtained.

v ₁ = v _a + v _b/2 , v ₂ = v _b + v _c/2 , v ₃ = v _c + v _a/2 These voltages are shown in the vector diagram in Figure 2. As can be seen, the magnitude of the vectors synthesized in this way is the smallest for the one corresponding to the accident phase. To be more specific, a
When a ground fault occurs in a phase, the following relationships hold between the amplitudes of the vectors: |v ₁ |<|e _a |<|v ₂ |, |v ₃ |. Similar relationships hold when accidents occur in other phases. Therefore, by using an appropriate logic circuit, the fault phase can be determined.

As mentioned above, conventional ground fault phase detection devices essentially detect variations in the absolute value of each relative ground voltage before and after an accident, and determine the ground fault phase based on the magnitude between them. However, if both the ground capacitance Co and ground fault resistance Rg of the line are large at the time of an accident, the difference in voltage between normal and faulty conditions becomes extremely small, making detection difficult. In order to increase the temperature and sensitivity, it is necessary that the partial pressure ratio of the voltage divider and the amplification rate of the operational amplifier are highly balanced between each phase, and even if this were achieved, However, the disadvantage was that malfunctions were more likely to occur due to noise.

This invention was made in order to eliminate the drawbacks of the conventional ones as described above, and it is possible to directly detect the zero-sequence voltage generated in proportion to the fault current and determine the ground fault phase from the entire waveform. The purpose is to provide a ground fault phase detection device.

Hereinafter, in FIG. 3 for explaining one embodiment of the present invention, 1a, 1b, 1c are three-phase balanced power supplies, and 2a, 2b, 2c are three-phase power supplies connected to the three-phase balanced power supply 1. The electric wires 3a, 3b, 3c are
The ground capacitance of the 3-phase distribution line 2, 4 is the ground fault point that may exist somewhere on the 3-phase distribution line, 7 is the earthing resistance of the power supply neutral point, and 9 is the star-shaped phase of each phase. Two sets that generate reference voltages u _a , u _b , u _c corresponding to vectors delayed by a certain phase from the voltage, and reference voltages u _a ′, u _b ′, u _c ′ whose phase is 90° ahead of these voltages. a three-phase phase-shifting transformer with a secondary winding of
10 is a capacitive voltage divider that detects the zero-phase voltage v _p ; 11
a, 11b, 11c are the zero-phase voltage v _p and the reference voltage u _a ,
A multiplier that multiplies u _b and u _c , 12 is the zero-sequence voltage v _p
The differentiators for determining the differential v _p ′=1/ω dv ₀ /at, 1 4a, 14b, and 14c are multipliers for multiplying v _p ′ and u _a ′, u _b ′u _{c ′} , 15a, 15b, 15
c is 11a, 11b, 11c and 14a, 14
an integrator 16a, 16b, which adds together the outputs of two sets of multipliers b and 14c and then integrates the sum;
16c is a comparator that compares the output of the integrator with a preset threshold -Wth.

Next, the principle of operation of the apparatus shown in FIG. 3 according to the present invention will be explained. The phase voltages of the three-phase balanced power supplies 1a, 1b, and 1c of the system under consideration can be expressed as follows.

e _a = E・Sinωt e _b = E・Sin (ωt−2/3π) e _c = E・Sin (ωt−4/3π) The phase shift transformer 10 lags the upper phase voltage by an angle α Reference voltage u _a corresponding to the vector,
Suppose we generate u _b , u _c and reference voltages u _a ′, u _b ′, u _c ′ each with a phase lead of 90°.
These can be expressed as follows.

u _a =E・Sin(ωt−α) u _b =E・Sin(ωt−α−2/3π) u _c =E・Sin(ωt−α−4/3π) u _a ′=E・cos(ωt −α) u _b ′=E・cos(ωt−α−2/3π) u _c ′=E・cos(ωt−α−4/3π) Since these reference voltages are related to the line voltage, There is no change even when there is an unbalanced ground fault point of only the a phase as shown in FIG. 3 and FIG. However, a
When a fault occurs in the ground fault resistance Rg in a phase, the potential at the neutral point of this three-phase circuit fluctuates and a zero-sequence voltage is generated. This zero-sequence voltage v _p appears at the output of the capacitive voltage divider. v _p is expressed as follows in relation to Rg, the capacitance c _p of the three-phase line, and the grounding resistance R _N at the neutral point.

v _p = −V _p・Sin(ωt−θ) However, θ=tan ^-1 3ωCo/1/R _N +1/Rg Figure 4 shows the above voltages e _a , e _b , _c ; u _a , u _b , u _c ;
This is a vector diagram showing the relationship between u _a ′, u _b ′, u _c ′ and v _p . When the value of Rg changes, the vector v _p
The foot of moves on the circle diagram 8. This zero-sequence voltage v _p
and the reference voltages u _a , u _b , u _c are multiplied by the multiplier 11 in FIG.
When inputting into a, 11b, and 11c, the following product is obtained.
w _a1 , w _b1 , w _c1 are obtained.

w _a1 =v _p・u _a =−1/2V _p E{cos(θ−α)
−cos(2ωt−θ−α)} w _b1 =v _p・u _b =−1/2V _p E{cosθ−α−2
/3π−cos(2ωt−θ−α−2/3π)} w _c1 =v _p・u _c =−1/2V _p E{cos(θ−α−
4/3π)-cos(2ωt-θ-α-4/3π)} In these equations, the DC component in the first term on the right side is an amount proportional to each reference voltage component of the zero-sequence voltage in the vector diagram, and the The second term is an AC component that changes at a frequency twice the power supply frequency. Information regarding the accident phase is included in the DC component of the first term, and the second term, which oscillates at 2ω, becomes an obstacle in determining the accident phase, so we would like to cancel this term if possible.

Therefore, the zero-phase voltage v _p is input into the differentiator 12 shown in FIG. 3 to obtain v _p '=1/ω dv _p /dt. This voltage v _p ′ has a phase that is 90° ahead of v _p and can be expressed as follows.

v _p ′=−V _p・cos(ωt−θ) Now, if we put v′ _p and u _a ′, u _b ′, u _c ′ into the multiplier 14 in FIG. 3 as before, we get the following A product like
w _a2 , w _b2 , w _c2 are obtained.

w _a2 =v _p ′・u _b =−1/2V _p E{cos(θ−α)+
cos(2ωt−θ−α)} w _b2 =v _p ′・u _b ′=−1/2V _p E{cos(θ−α−
2/3π) + cos (2ωt−θ−α−2/3π)} w _c2 = v _p ′・u _c ′=−1/2V _p E{cos(θ−α−
4/3π) + cos (2ωt-θ-α-4/3π)} The AC components of the second term on the right side of these equations are w _a1 , w _b1
Since the sign is opposite to the alternating current component of w _c1 , the vibration term of 2ωt can be eliminated by adding these together.

w _a = w _a1 + w _a2 = -V _p・E・cos (θ−α) w _b = w _b1 +w _b2 = −V _p・E・cos (θ−α−2/3 π) w _c = w _c1 +w _c2 = -V _p・E・cos (θ−α−2/3 π) Actually, the integrators 15a, 15b, 15c in Fig. 3
By adding w _a1 and w _a2 , w _b1 and w _b2 , w _c1 and w _c2 and integrating them at the same time, Wa, Wb,
Looking for Wc. Expressing these in formulas is as follows. However, t=tg is the time when the accident occurred.

Wa=∫w _a dt=-V _p・E・cos(θ−α)・(t−tg) Wb=∫w _b dt=−V _p・E・cos(θ−
α−2/3π)・(t−tg) Wc=∫w _c dt=−V _p・E・cos(θ−
α-4/3π)・(t-tg) As mentioned above, the integrands w _a , w _b , w _c are the directional components of each reference voltage u _a , u _b , u _c of the zero-sequence voltage in the vector diagram. It is a proportional amount, and if there is no ground fault and the three phases are in balance, all three will be zero, and the integral
Wa, Wb, and Wc also become zero. However, as can be seen from the vector diagram in Figure 4, if a ground fault occurs in phase a, the foot of the vector of v _p will be somewhere on the circle diagram 8, so α can be set from 0 to If you choose an appropriate angle between 90 degrees, v _p and u _a will be almost opposite, and
The value of w _a above is negative. Here, an appropriate angle α
This is because when the resistance value Rg at the grounding accident point is small, it can be detected with high sensitivity around α = 0°, and when Rg is large, it can be detected with high sensitivity around α = 90°, so α can be determined in advance according to the expected accident point. can. Wb and Wc, which are not ground fault phases, have positive or small negative values, and can be distinguished from Wa, which has a large negative value.

Figure 5 shows these integral quantities before and after the occurrence of a ground fault.
It shows how Wa, Wb, and Wc change over time. In this way, the integral quantities Wa, Wb, and Wc increase without oscillating in the negative direction immediately after a ground fault, while the components corresponding to the fault phase increase in the positive direction. Increase or change little. Therefore, by setting a certain negative threshold value -Wth in advance, FIG.
If the integrated value of any phase reaches this threshold value, the signal a corresponding to the ground fault phase will be generated.
b, c can be generated. Although other suitable logic circuits may be considered, using a method that uses a fixed threshold value for the integral quantity will reduce the integral quantity in the event of a ground fault that causes a large zero-sequence voltage. Since it changes rapidly, it takes a long time to make a decision, but they all have one thing in common: it takes a certain amount of time to make a decision for small zero-sequence voltage accidents.

Above, we considered the case where the integrators 15a, 15b, and 15c perform complete time integration, but the multiplier 14
If there is even a slight DC component due to the calculation accuracy of a, 14b, 14c, etc., this will accumulate. In order to avoid this, integrators 15a, 15b, 15c
It is necessary to make the characteristics of the integral with an appropriate time constant, that is, the transfer function of the first-order lag element 1/(1+sT). Note that the above-described function can be maintained as long as the integration time constant T is set to be appropriately long compared to the ground fault phenomenon to be detected.

In the above embodiment, the reference voltages u _a , u _b , u _c and
Although u _a ′, u _b ′, and u _c ′ are used as sine waves, it is possible to obtain a device having the same function as the above embodiment even if these are made into rectangular waves with the same phase and constant amplitude as the sine waves. can. A waveform representing this is shown in FIG. The figure shows the waveforms of various parts after a ground fault occurs in the a phase, and the reference voltages u _a and u _a ′ are set to a,
b, the zero-sequence voltage v _p is c, v _p ′ is d, the product of the reference voltage and zero-sequence voltage v _p・u _a , v _p ′・u _a ′ is e, f, v _p・
w _{a, which is the sum of u a} and v _p ′・u _{a ′} , is shown in g. Figure 7 also shows the situation before and after the ground fault accident.
The time change of the integral Wa of w _a is b, the integral amount Wb of the c phase,
It is shown in comparison with the change in Wc over time. Although the waveform has a vibration component to some extent compared to the waveform in FIG. 5, if thresholds are provided for these three components, the ground fault phase can be determined in almost the same way.

In the above example, the reference voltage is a rectangular wave, but the zero-phase voltage v _p and its differential v _p ′ are formed into a constant amplitude rectangular wave with only their phase information and input to the multiplier. Also, it is possible to obtain a device having the same functions as those of the above embodiment.

In addition, when attempting to detect a ground fault phase by increasing the zero-sequence voltage detection sensitivity, due to the dynamic range constraints of the arithmetic circuit, v _p or v _p ′ must be determined for a large zero-sequence voltage signal. Saturation may occur in the circuit, but considering that this extreme case corresponds to the one using the zero-sequence voltage of the rectangular wave above, even if saturation occurs in the arithmetic circuit, the phase of the zero-sequence voltage It can be seen that since the information remains, it is possible to detect the ground fault phase without any problem.

In the above embodiment, a slight imbalance in the system
Detector imbalance and multipliers 11a, 11b, 11
Depending on the accuracy of c, 14a, 14b, 14c, etc.
It is conceivable that a DC component is accumulated in the outputs of the integrators 15a, 15b, and 15c. In order to avoid this, the characteristics of the integrators 15a, 15b, and 15c are changed from the transfer function 1/S for pure integration to the transfer function 1/S for integration with an appropriate time constant, that is, the transfer function 1/S for the first-order lag element.
It is necessary to change it to (1+ST). This is because at t→∞ (corresponding to S→0), the output becomes infinite in pure integration (1/S), whereas,
This is because 1/(1+ST) is finite. Therefore, although the small DC component is not rapidly integrated by the integral amount, this DC component increases the integrator output.
DC bases with different values are generated in Wa, Wb, and Wc even before the ground fault occurs, so
It is conceivable that this may have an adverse effect on ground fault phase detection using the threshold value. In order to eliminate this, the output of the integrator (first-order lag element with time constant T) may be passed through a capacitor C as shown in FIG. A resistor R placed after the capacitor C is used to always set the output base to zero, and the overall transfer function becomes S/(S+CR). In the entire system, the transfer function of the integrating circuits 15a, 15b, 15c is 1/(1
+ST) (which becomes a finite value as S→0 at t→∞) is in a cascade with the above, so the value of the time constant CR, like the time constant T of integration, is The choice should be made taking into consideration the level of network disturbance and the level of regular grid disturbance. Also CR
Even if the positions of the circuit and the integrating circuit are swapped back and forth, the operation will not change.

In the above embodiment, the reference voltages u _a , u _b , u _c and
Although a three-phase phase shift transformer was used to calculate u _a ′, u _b ′, and u _c ′, it is also possible to calculate the reference voltage using a capacitance voltage divider. Figures 9 and 10
The figure shows an example where the phase difference α=30°. In FIG. 9, the outputs of the capacitive voltage dividers 5a, 5b, 5c are input into adders 17a, 17b, 17c, and the reference voltages u _a ,
At the same time as calculating u _b and u _c , the differentiators 18a and 18
b and 18c generate u _a ′, u _b ′, and u _c ′.
On the other hand, in FIG. 10, adders 17a, 17b, 17
The outputs u _a , u _b , u _c of c are sent to adders 19a, 19 again.
b, 19c to obtain u _a ′, u _b ′, u _c ′.

As described above, according to the present invention, the multiplier processes information from the entire zero-sequence voltage of the system, and the integrator integrates information for a certain period to determine the fault phase, thereby reducing noise. Even if it exists, it has the effect of obtaining something that operates stably.

[Brief explanation of drawings]

Fig. 1 shows the structure of a conventional ground fault phase detection device, and Fig. 2 shows the voltage of each phase when a ground fault occurs, and
FIG. 3 is a diagram showing the vector relationship of zero-sequence voltages. FIG. 3 is a diagram showing the structure of an embodiment of the ground fault phase detection device based on the present invention. FIG. The figure shows that when the vibration term that oscillates at 2ωt is removed, the integral amount of each reference voltage direction component of the zero-sequence voltage is
Figure 6 shows how the reference voltage, zero-sequence voltage, product and sum of the reference voltage and zero-sequence voltage change over time when the reference voltage is a rectangular wave voltage. Figure 7 is a diagram showing how the sum of the reference voltage and zero-sequence voltage changes over time;
The figure shows an example of a method for removing the influence of zero-sequence voltage that occurs during normal operation, Figure 9 is a circuit diagram showing an example of calculating a reference voltage using a capacitive voltage divider, and Figure 10 shows a capacitive voltage divider. FIG. 2 is a circuit diagram showing an example of calculating a reference voltage using a voltage divider. In the figure, 2a, 2b, 2c are three-phase distribution lines,
9 is a phase shift transformer, 10 is a capacitance voltage divider, 11a, 1
1b, 11c are multipliers, 12 is a differentiator, 14a,
14b, 14c are multipliers, 15a, 15b, 15
c is an integrator, and 16a, 16b, 16c are comparators. Note that the same reference numerals in each figure indicate the same or corresponding parts.

Claims (1)

[Claims] 1. In a three-phase power system, a zero-sequence voltage is detected from the ground voltage of each phase, and a differentiator generates a differentiation of the zero-sequence voltage.
Generate three first reference voltages corresponding to constant angularly delayed vectors between 90 degrees and a second reference voltage of the same amplitude that is 90 degrees ahead of them in phase, and use a multiplier to calculate the above Multiply the instantaneous value of the zero-sequence voltage, the instantaneous value of the first reference voltage, the instantaneous value of the differential of the zero-sequence voltage, and the second reference voltage for the corresponding phases, and integrate the sum of these products. Monitor the 3 integral quantities obtained by integrating with the device, and if any of the above 3 integral quantities increases in the negative direction immediately after a ground fault, a ground fault has occurred in the phase corresponding to that integral quantity. A ground fault phase detection device characterized by determining that a ground fault has occurred. 2. As a reference voltage, a rectangular wave voltage that has the same phase as three sine wave voltages corresponding to a vector delayed by a certain angle between 0° and 90° from the star-shaped phase voltage of each phase, and a rectangular wave voltage that has a phase of 90° from them. 2. The ground fault phase detection device according to claim 1, wherein a rectangular wave voltage with an advanced voltage is used. 3. Claim 1, characterized in that a circuit that generates a rectangular wave voltage having the same phase as the zero-sequence voltage and its differential is used in the portion that detects the zero-sequence voltage and its differential from the ground voltage of each phase. Or the ground fault detection device described in paragraph 2. 4. Claim 1, characterized in that the part that detects the zero-sequence voltage and its differential from the ground voltage of each phase is configured such that the output of the detection part is saturated for inputs above a certain level. Or the ground fault phase detection device according to item 2. 5. The ground fault phase detection device according to any one of claims 1 to 4, characterized in that a capacitor for removing a DC component is connected before or after the integrator.