RU2074473C1 - Method for automatic control of compensation of capacitance and resistance constituents when arc short-circuits to earth occur - Google Patents

Abstract

FIELD: electric engineering. SUBSTANCE: method involves measurement of bias voltage of neutral line and linear voltage in one phase, synchronous detection by two orthogonal reference signals, filtration of results of synchronous detection and control of compensation of resistance and capacitance constituents. First operation of synchronous detection involves setting phase of reference signal equal to phase of electromotive force between short circuit point to earth and neutral wire. Result is used for compensation of resistance constituent. Second operation of synchronous detection involves use of neutral bias voltage for control of compensation of capacitance constituent. EFFECT: increased precision of automatic control, improved dynamic characteristics. 5 dwg

Description

 The invention relates to the electric power industry, specifically to relay protection and automation in three-phase high-voltage distribution networks with non-grounded neutral.

 A known method of auto-compensation of the capacitive component (Bryzgin OR, Zelensky, VA, Device for automatic regulation of the compensation current. Aut. St. USSR N 330509, class N 02 J 3/12), which includes recognition of a single-phase earth fault (ОЗЗЗ), recognition of the damaged phase, measurement of the voltage of the damaged phase and the bias voltage of the neutral, as well as the formation of the compensation control of the capacitive component, is proportional to the projection of the voltage vector of the damaged phase on a vector orthogonal to the voltage vector eniya neutral (using the synchronous detection operation).

 The disadvantage of this method is the impossibility of controlling the compensation of the active component of the OZNZ current and, as a result, the incomplete suppression of arc faults if the breakdown voltage of the insulation at the fault location is below the amplitude of the phase voltage of the network. In addition, this method is not applicable in cases where an earth fault (ZNZ) did not occur in the phase conductor of the network, but in the winding of the motor or transformer (power or load).

 The closest in technical essence to the claimed method is a method for automatically adjusting the compensation of the capacitive component (CES) and the active component (CAS) of the ZNZ currents described in (Efimov Yu.K. Obabkov V.K. Tseluevsky Yu.N. Shishkina OG System automatic suppression of arc faults in auxiliary networks of power units of 500 MW. Power stations, 1992, N 5, p. 71 75; Obabkov VK, Tseluevsky Yu.N. Efimov Yu.K. Device for compensating the total current of a single-phase fault in short networks VNIIGPE decision on the issue of copyright certificate by application N 4827316/07 of 01/28/91). This method consists in measuring the instantaneous values of the phase and linear voltages of the network and the neutral bias voltage, recognizing the arc OZZZ, recognizing the damaged phase, generating reference signals in the form of a linear voltage between the undamaged phases and its orthogonal signal, as well as synchronous detection of the voltage of the damaged phase by the indicated signals with subsequent filtering, after which CAS and CES are controlled. The prototype method is capable of completely suppressing the arc processes during OZNZ in cases where the breakdown voltage of the arc gap is quite high.

 The described method has the following disadvantages. Firstly, it is unsuitable for auto-tuning in cases where the ZNZ did not occur in the phase conductor of the network, but in the load winding or the transformer supplying the network, since it is principally focused on minimizing the voltage between one of the network phases (namely, the damaged phase) and ground. Secondly, the correct auto-tuning of CAS is achieved only with the fine-tuning of CES. When CES detuning in that direction, the CAS is tuned with overcompensation, which is greater, the greater the detuning according to CES. This adversely affects the dynamics of two-channel auto-tuning, delays transients and reduces the efficiency of the system as a whole. Thirdly, the speed of the prototype method depends on the breakdown voltage of the arc gap and decreases significantly with its decrease, and the presence of a dead zone in the channel for measuring the voltage of the damaged phase (which positively affects the dynamic characteristics of the system at high breakdown voltages) at low voltages leads to the opening of both auto-tuning circuits, since the voltage on the damaged phase ceases to go beyond the mentioned dead band.

 The present invention solves the following technical problems. Firstly, accurate autoconfiguration of CAS and CES is provided regardless of where the fault occurs: in the phase conductor of the network, in the load winding or in the winding of the supply transformer, and with any connection scheme of these windings. In this case, the voltage is minimized precisely at the ZNZ site, which leads to the cessation of arc breakdowns. Secondly, the CAS regulation process turns out to be largely independent of CES regulation, which improves the dynamic characteristics of auto-tuning. Thirdly, the speed of the method with a decrease in the breakdown voltage at the place of damage does not decrease, but increases, and the opening of the auto-tuning circuits in the presence of arc breakdowns does not occur. And, finally, for the implementation of this method, it is not necessary to recognize the arc ZNZ mode, recognize the damaged phase and measure the voltage across it, thereby achieving some simplification of the mentioned implementation.

The solution of these problems is ensured by the fact that in a method that includes measuring the neutral bias voltage and one of the line voltage of the network, synchronous detection with two mutually orthogonal reference signals, filtering the results of synchronous detection and controlling the compensation of capacitive and active components in accordance with the results obtained, additionally determine the value of the phase of the EMF, acting between the point of the fault and the neutral of the network, set the phase of the reference signals, respectively, the synchronous detection operation is equal to the phase of said EMF, and for the second synchronous detection operation, which is 90 ^° different from it, the neutral bias voltage is filtered, differentiated, limited and subjected to the synchronous detection operations, the results of the first synchronous detection operation are used to control the active component compensation , and the result of the second synchronous detection operation to control the compensation of the capacitive component of the earth fault currents.

 Figure 1 shows the timing diagrams of processes in the network and the signals generated during the implementation of this method in conditions of fine-tuning the CES and undercompensation for the active component; figure 2, figure 3, figure 4 similar timing diagrams for (respectively) fine tuning CES and overcompensation for the active component (figure 2), overcompensation for the capacitive component and fine tuning CAS (figure 3), undercompensation for capacitive component and fine-tuning CAS (figure 4). In addition, figure 5 shows an example of a functional circuit diagram of a device that implements the inventive method.

напряжения e(t) смещения нейтрали, 5 результат синхронного детектирования указанной производной опорным сигналом, пропорциональным ЭДС E(t), 6 среднее значение U_a сигнала 5, 7 опорный сигнал операции синхронного детектирования, сдвинутый на 90^o, по отношению к ЭДС E(t), 8 результат синхронного детектирования производной 4 опорным сигналом 8, 9 среднее значение U_т.In FIG. 1 4 are marked: 1 EMF E (t), acting between the point of the SCZ and the neutral of the network, 2 voltage e (t) of the neutral bias, 3 voltage U _g (t) between the point of the SCZ and ground, 4 derivative

voltage e (t) of the neutral bias, 5 the result of synchronous detection of the indicated derivative with a reference signal proportional to the EMF E (t), 6 the average value U _{a of the} signal 5, 7 the reference signal of the synchronous detection operation, shifted by 90 ^o , relative to the EMF E ( t), 8 is the result of synchronous detection of the derivative 4 by the reference signal 8, 9; the average value of U _t .

In FIG. 5 are marked: 9 a network with a non-grounded neutral N, powered by EMF sources E _A (t), E _B (t), E _C (t), having capacities C _A , C _B , C _C between the phases of the network and ground and working on load 10 with earth fault at point G. The network is equipped with an arc suppression device 11 DGA (for example, a combination of "connecting transformer - arc suppression reactor" or a Bauch transformer with a thyristor key in the open triangle circuit of the secondary windings) associated with the compensator 12 of the active component of the UAN, and three-phase measuring transformer voltage 13 th ITN. The voltage e (t) of the neutral bias from the output of the ITN 13 is supplied to the series-connected blocks 14 of filtering F1, differentiation 15 D and restrictions 16. In addition, the voltage e (t) is supplied to the inputs of the conversion circuit 17 into the time interval of the instantaneous value e (t _i + δ) of the neutral bias voltage e (t) that takes place at the time t _i + δ, where δ is a small value (that is, immediately after the abrupt change in the neutral bias voltage e (t) due to an arc breakdown at the ZNZ site). Circuit 17 consists of an integrator built on an operational amplifier 18 and a capacitor 19, an analog switch 20, resistors 21 and 22, which turn the integrator (with the switch position 20 opposite to that shown in FIG. 5) into an inertial unit with a small time constant, a circuit resistor 23 transformations "initial conditions of the integrator time interval" and relay link 24 with a small (about 10 30 mV) hysteresis. Together with the phase voltages U _A (t), U _B (t) and U _C (t), the neutral bias voltage e (t) is also supplied to the additional signal coupling unit 25 (DBSS). In addition, the device includes a processor 26 with read-only memory and RAM devices, a bus 27 of data, addresses and control signals, an input (receiving information) port 28 P1, an output (output information) port 29 P2, a logic element 30 "DISAQUALITY" , controller 31 interruptions KP and the scheme of the overall synchronization of the device with industrial frequency. The mentioned general synchronization scheme includes a zero-comparator 32 of the line voltage signal U _AB (t) of the network, pulse shapers 33 and 34 (FI1 and FI2), eliminating the bounce, respectively, of the falling edge of the output signal of the comparator 32 and the programmable timer 35 T1, to the counting input "c" of which a high-frequency (for example 2 MHz) clock signal is applied. The bus 27 is also connected to blocks 35 and 36 of the formation of the BFUA and BFUR departments according to the active and reactive components, respectively, the outputs of which are connected to the executive bodies 37 and 38 (IOR and IOA). The structure of blocks 37 IOR and 38 IOA depends on the specific implementation of the DGA 11 and the method of introducing CAS. For example, if the DGA is a Bauch transformer, then the IOR unit 37, by analogy with a device that implements the prototype method, is a thyristor switch installed in the secondary circuit of the specified transformer. If the introduction of controlled artificial asymmetry from an external source of EMF through an additional inductor is used for CAS, then the IOA block 38, by analogy with a device that implements the prototype method, is a thyristor switch connected in series with the said additional inductor.

 The action of the method is as follows.

где e(t) напряжение смещения нейтрали;
I_g(t) ток в месте повреждения;
θ(t) ток несимметрии сети, в состав которого входит, в частности регулируемый ток искусственной несимметрии, вводимый для компенсации активной составляющей;
U_g(t) напряжение между точкой повреждения сети и землей;
E(t) ЭДС, действующей между точкой ЗНЗ и нейтралью сети;
F(U_g) вольт-амперная характеристика места повреждения;
L индуктивность в нейтрали сети (определяющая КЕС);
С С_A + C_B + C_C суммарная емкость между фазами сети и землей;
g проводимость, характеризующая активные потери в КНПС;
В основу изложения принципа действия предлагаемого способа положим рассмотрение свойств решений данной системы уравнений, причем начать указанное изложение целесообразно с наиболее простых случаев точной настройки КЕС при условиях точной настройки, недо- и перекомпенсации активной составляющей с относительно высоким напряжением пробоя дугового промежутка в месте ЗНЗ.Low-frequency processes in the circuit of the zero sequence of the network (KNPS) are described simplistically by the following system of differential equations:

where e (t) is the neutral bias voltage;
I _g (t) current at the fault location;
θ (t) the asymmetry current of the network, which includes, in particular, an adjustable artificial asymmetry current introduced to compensate for the active component;
U _g (t) voltage between the point of damage to the network and ground;
E (t) EMF, acting between the point ZNZ and the neutral network;
F (U _g) current-voltage characteristic of the place of damage;
L inductance in the neutral of the network (determining CES);
С С _A + C _B + C _C total capacity between network phases and ground;
g conductivity, characterizing active losses in KNPS;
The basis of the description of the principle of action of the proposed method is to consider the properties of solutions to this system of equations, and it is advisable to start the above statement with the simplest cases of fine-tuning the CES under conditions of fine-tuning, under- and overcompensation of the active component with a relatively high breakdown voltage of the arc gap at the ZNZ site.

Здесь равенство (4) означает, что напряжение e(t) смещения нейтрали к моменту деионизации дугового промежутка становится практически равным ЭДС E(t), действующей между точкой ЗНЗ и нейтралью сети, а равенство (5) означает, что ток нулевой последовательности (ток дугогасящего аппарата ДГА) за время существования проводящего состояния в месте ЗНЗ не успевает сколько-нибудь заметно измениться (а именно, изменяется на величину δ•e(t_o)/L,, которая мала вследствие малости δ.In FIG. 1, the voltage 3 U _g (t) at the damage site (which is shown in the graph as curve 1) according to (3) corresponds to the distance from curve 1 depicting the EMF E (t) to curve 2 depicting the neutral bias voltage e (t) . When the voltage U _g (t) of the voltage U _d reaches the breakdown of the arc gap at the ZNZ (moment t _o in figure 1), the latter is ionized and its conductivity increases many times. In this case, the instantaneous current value I _g (t) rapidly increases from zero to a very large value (possibly kiloamperes), which means the appearance of a significant value on the right side of equation (1). A similar effect on the left side of equation (1) leads to a fast (almost stepwise) change in the neutral displacement voltage e (t) toward its approaching the value of E (t _o ), which entails a decrease in the voltage U _g (t) at the damage site , as well as, according to (2), a decrease (in absolute value) of the current I _g (t) in the place of the ZNZ. After a period of time δ (usually a short one, determined by the time of recharging the capacitance between the phases of the network and the ground), the current I _g (t) decreases to a certain threshold value at which the arc gap is deionized and the conductivity in the SCZ decreases to almost zero, i.e. there is a shockless pause. Thus, the effect on the left side of equation (1) at the moment of the arc breakdown is of a nature close to the influence of the d-function, the result of which can be considered a change in the initial conditions of this equation. The mentioned initial conditions with which further processes develop (after the extinction of the arc) have the following form:

Here, equality (4) means that the neutral bias voltage e (t) at the moment of deionization of the arc gap becomes almost equal to the EMF E (t) acting between the point of the short-circuit current and the network neutral, and equality (5) means that the zero sequence current (current during the existence of the conducting state in the place of the ZNZ does not have time to noticeably change (namely, it changes by the value of δ • e (t _o ) / L, which is small due to the smallness of δ.

и фазой

где θ_m и ν соответственно, амплитуда и фаза тока
θ(t) = θ_mcos(ωt+ν)
несимметрии сети.Processes in KNPS during a dead time pause is determined by the sum
e (t) e _c (t) + e _b (t) (6)
two oscillatory components: free
e _C (t) = e $\genfrac{}{}{0ex}{}{\text{m}}{\text{C}}$ • e- ^λt cos (ω _C t + Φ _C ), (7)
with frequency ω (LC) ^-1/2 , damping λ = g / 2C and forced
e _b (t) = e _o cos (ωt + Φ), (8)
with industrial frequency ω, amplitude

and phase

where θ _m and ν, respectively, the amplitude and phase of the current
θ (t) = θ _m cos (ωt + ν)
network asymmetries.

When fine tuning the CAS, the amplitude θ _m (determined by the artificial asymmetry introduced for the CAS) is such that the equality e _o E _o , where E _{o is the} amplitude of the EMF
E (t) = E _o cos (ωt + β),
acting between the GSS point and the neutral of the network, and when fine-tuning the CES, the value of L is such that the equality Φ = β is observed, where β is the phase of the mentioned EMF E (t). If the arc breakdown occurred at the extremum point of the function E (t), then, as is easy to see from (1) (8), the amplitude e $\genfrac{}{}{0ex}{}{\text{m}}{\text{C}}$ the free component will be equal to zero, therefore, the curve e (t), up to higher harmonics, coincides with the curve E (t). At the same time, the voltage Ug (t) at the damage site will be identically equal to zero, arc breakdowns will no longer be repeated, and therefore there will be no need to change the settings of the CES and CAS (with constant KNPS parameters).

в момент времени t₁ напряжения U_d происходит очередной дуговой пробой и процесс повторяется (моменты t₂, t₃ и т.д. на фиг.1). Дифференцирование сигнала e^*(t), который представляет собой напряжение e(t) смещения нейтрали, предварительно отфильтрованное, согласно формуле изобретения, от колебательных высокочастотных составляющих, дает последовательность однополярных (положительных или отрицательных для каждого из дуговых пробоев) импульсов, возникающих в моменты t_i дуговых пробоев. Указанные импульсы значительно превосходят по абсолютной величине амплитуду сигнала рассматриваемой производной, имеющего место в промежутках между дуговыми пробоями (то есть во время бестоковых пауз). Поэтому на фиг.1 4 производная

(под которой всюду подразумевается производная именно отфильтрованного сигнала e^*(t)) во время бестоковых пауз изображается равной нулю. Ниже показано, что знаки рассматриваемых импульсов несут информацию о настройке КАС и КЕС.If in the network there is an undercompensation for CAS (under conditions of fine tuning of the CES), then for the forced oscillations (8) the relations e _o <E _{o are} valid; Φ = β .. The solution of equation (1) is such that the curve e (t) slowly (in accordance with the decrement λ of the damping of free vibrations (7)) deviates from the curve E (t) in the direction of decreasing the amplitude of the oscillatory process, which causes , according to (3), a gradual increase in the amplitude of the voltage U _g (t) at the site of damage. Upon reaching

at the time moment t _{1 of the} voltage U _d , another arc breakdown occurs and the process repeats (moments t ₂ , t ₃ , etc. in FIG. 1). Differentiation of the signal e ^* (t), which is the neutral bias voltage e (t), previously filtered, according to the claims, from high-frequency vibrational components, gives a sequence of unipolar (positive or negative for each of the arc breakdowns) pulses that occur at times t _i arc breakdowns. The indicated pulses significantly exceed in absolute value the amplitude of the signal of the derivative under consideration, which takes place in the gaps between the arc breakdowns (that is, during dead time pauses). Therefore, in Fig.1 4 derivative

(by which everywhere is meant the derivative of the filtered signal e ^* (t) itself) during current-free pauses is shown to be equal to zero. It is shown below that the signs of the considered pulses carry information on the tuning of CAS and CES.

, и, следовательно, напряжение U_g(t) синфазно (при недокомпенсации по КАС, фиг.1) или находится в противофазе (при перекомпенсации по КАС, фиг.2) с данной ЭДС. Поэтому дуговые пробои в указанных условиях возникают вблизи экстремумов напряжения U_g(t), а следовательно вблизи экстремумов e(t) и ЭДС E(t), то есть в моменты времени, когда

, i 1, 2, 3, что и отражено на фиг.1. Далее, из того обстоятельства, что

при каждом t следует, что при недокомпенсации по КАС (то есть в ситуации, отображенной на фиг.1) знаки

в дискретные моменты t_i, i 1, 2, 3 дуговых пробоев совпадают со знаками ЭДС E(t) в эти же моменты t_i. Поэтому синхронное детектирование сигнала 4, то есть, производной

сигналом, пропорциональным ЭДС E(t), даст последовательность однополярных (положительных) импульсов (позиция 5 на фиг.1), результат 6 осреднения которых также будет положительным. Результат же синхронного детектирования упомянутой производной сигналом, сдвинутым на 90^o относительно ЭДС Е(t), будет близким к нулю, поскольку дуговые пробои приходятся на такие моменты времени, когда опорный сигнал близок к нулю. После предусмотренного формулой изобретения осреднения по времени он тем более будет близок к нулю, так как импульсы, полученные в результате синхронного детектирования, скорее всего не будут однополярными.With a fine adjustment of the CES (i.e., with ω _C ≃ω and Φ-β), during a current-free pause, the voltage e (t) is almost in phase with the EMF E (t), and

, and, therefore, the voltage U _g (t) is in phase (when the CAS is undercompensated, FIG. 1) or is out of phase (when the CAS is overcompensated, FIG. 2) with this EMF. Therefore, arc breakdowns under the indicated conditions occur near the voltage extremes U _g (t), and therefore near the extremes e (t) and EMF E (t), that is, at times when

, i 1, 2, 3, which is reflected in figure 1. Further, from the fact that

for each t it follows that in case of undercompensation by UAN (that is, in the situation shown in Fig. 1), the signs

at discrete moments t _i , i 1, 2, 3 of arc breakdowns coincide with the signs of EMF E (t) at the same moments t _i . Therefore, the synchronous detection of signal 4, that is, the derivative

a signal proportional to the EMF E (t) will give a sequence of unipolar (positive) pulses (position 5 in FIG. 1), the result of 6 averaging of which will also be positive. The result of synchronous detection of the aforementioned derivative by a signal shifted by 90 ^o relative to the EMF E (t) will be close to zero, since arc breakdowns occur at such times when the reference signal is close to zero. After the time averaging provided by the claims, it will be all the more close to zero, since the pulses obtained as a result of synchronous detection will most likely not be unipolar.

превысит напряжение U_d пробоя поврежденной изоляции, то в момент t₁ произойдет очередной дуговой пробой и процесс повторится (дальнейшие пробои следуют в моменты времени t₂, t₃ и т. д. на фиг. 2). Согласно изложенному выше, дуговые пробои происходят вблизи экстремумов ЭДС E(t), то есть при

, k 0, 1, 2, 3 ( фиг.2), однако, поскольку в рассматриваемой ситуации

в отличие от ранее рассмотренной недокомпенсации по активной составляющей, знаки

в моменты t₁ дуговых пробоев противоположны знакам ЭДС E(t) (кривая 1) в эти же моменты. Поэтому синхронное детектирование сигнала производной

сигналом, пропорциональным ЭДС E(t), даст последовательность однополярных, отрицательных импульсов (позиция 5 на фиг.2), результат 6 осреднения которых также будет отрицательным. Результат же синхронного детектирования сигнала упомянутой производной опорным сигналом, сдвинутым на 90^o относительно ЭДС E(t) так же, как и в предыдущем случае, будет близок к нулю по тем же причинам.In the case of overcompensation according to the CAS (as before in the conditions of fine tuning of the CES), for forced oscillations in the voltage e (t) of the neutral bias, the relations e _o > E _{o are} valid; Φ = β .. This case is illustrated in figure 2. During each dead-time pause, the solution of equation (1) is such that e (t) (curve 2) slowly (in accordance with the parameter λ) deviates from the curve E (t) in the direction of increasing amplitude of the oscillatory process, which, as in the previous case , leads to a gradual increase in the amplitude of the voltage U _g (t) at the site of damage. If the value

exceeds the breakdown voltage U _{d of the} breakdown of the damaged insulation, then at the moment t ₁ another arcing breakdown will occur and the process will repeat (further breakdowns follow at times t ₂ , t ₃ , etc. in Fig. 2). According to the above, arc breakdowns occur near the extrema of the EMF E (t), i.e., at

, k 0, 1, 2, 3 (Fig. 2), however, since in this situation

in contrast to the previously considered undercompensation for the active component, the signs

at moments t _{1 of} arc breakdowns are opposite to the signs of EMF E (t) (curve 1) at the same moments. Therefore, synchronous signal detection of the derivative

a signal proportional to the EMF E (t), will give a sequence of unipolar, negative pulses (position 5 in figure 2), the result of 6 averaging which will also be negative. The result of synchronous signal detection of the aforementioned derivative by a reference signal shifted by 90 ^o relative to the EMF E (t), as in the previous case, will be close to zero for the same reasons.

, k 0, 1, 2, 3, и поэтому отмеченные выше свойства результатов осреднения U_a и U_т полностью сохраняются.It is easy to see that, at a low voltage, the breakdown of the arc gap at the ZNZ site in both cases can result in several arc breakdowns during each half-period, which, however, are located almost symmetrically with respect to the moments

, k 0, 1, 2, 3, and therefore the properties of the averaging results U _a and U _t noted above are completely preserved.

(производная 4) в момент скачка положительна на падающих участках E(t) и отрицательна на восходящих участках. Так как падающим участкам E(t) соответствуют участки с положительными значениями опорного сигнала 7, сдвинутого на 90^o по отношению к ЭДС E(t), синхронное детектирование производной 4 опорным сигналом 7 даст последовательность однополярных, отрицательных импульсов 9 (фиг.3), среднее значение которых также отрицательно. Кроме того, как видно из фиг.3, большинство дуговых пробоев с положительной производной 4 происходит при положительных значениях ЭДС E(t), а большинство дуговых пробоев с отрицательной производной 4 при отрицательных значениях ЭДС 1. Поэтому результатом синхронного детектирования сигнала 4 (фиг.3) опорным сигналом 1 является последовательность импульсов 5, среди которых преобладают положительные, что определяет положительный знак среднего значения U_a (фиг.3), как это и должно быть в случае недокомпенсации (или отсутствия компенсации) активной составляющей. При введении КАС, то есть при появлении в решении (6) уравнений (1) (3) ненулевой вынужденной составляющей (8), растет число дуговых пробоев с положительной производной 4, попадающих на отрицательные значения ЭДС 1 E(t) и наоборот (то есть с отрицательной производной 4, попадающих на положительные значения ЭДС 1 E(t)). В этом случае соотношение между положительными и отрицательными импульсами 5 в сигнале U_a смещается в сторону отрицательных импульсов, что приводит к уменьшению среднего значения U_a, которое достигает нуля при e_o≃ E_o,, то есть вблизи точной настройки КАС. В случае дальнейшего роста КАС отрицательные импульсы 5 начинают преобладать над положительными, то есть при перекомпенсации активной составляющей среднее значение U_a становится отрицательным.Next, we consider the case of overcompensation in the capacitive component in the absence of UAN, that is, for e _o ≃0; ω _C > ω, Φ> β, which is illustrated in FIG. 3. Under these conditions, the free component (7) dominates in the solution e (t) of equations (1) (3). Since ω _C > ω ,, after each arc breakdown (at the moments t _o , t ₁ , t ₂ , etc.) and the occurrence of the next dead time, the voltage e (t) (curve 2) deviates from E (t) (curve 1) in the advance direction, that is, down from the curve E (t) in its falling sections and up from it in the ascending sections. As a result, as before, the voltage U _g (t) appears and gradually increases (in absolute value) at the site of damage, while being positive in the falling sections of the EMF E (t) and negative in its ascending sections. When this voltage reaches the threshold U _{d of} breakdown of the arc gap of the ZNZ site due to the action of the next arc breakdown, as before, an abrupt setting of relation (4) occurs (while maintaining equality (5)). Wherein

(derivative 4) at the time of the jump is positive in the falling sections of E (t) and negative in the ascending sections. Since the falling sections of E (t) correspond to sections with positive values of the reference signal 7 shifted by 90 ^° with respect to the EMF E (t), synchronous detection of the derivative 4 by the reference signal 7 will give a sequence of unipolar, negative pulses 9 (Fig. 3), whose average value is also negative. In addition, as can be seen from FIG. 3, the majority of arc breakdowns with a positive derivative 4 occurs at positive EMF values E (t), and the majority of arc breakdowns with a negative derivative 4 at negative EMF values 1. Therefore, the result of synchronous detection of signal 4 (FIG. 3) the reference signal 1 is a sequence of pulses 5, among which positive prevail, which determines the positive sign of the average value U _a (Fig. 3), as it should be in the case of undercompensation (or lack of compensation) active with leaving. With the introduction of CAS, i.e., when equations (1) (3) appear in the solution (6)) of a nonzero forced component (8), the number of arc breakdowns with a positive derivative 4 that fall on negative EMF values 1 E (t) and vice versa increases ( is with the negative derivative 4, falling on the positive values of the EMF 1 E (t)). In this case, the ratio between positive and negative pulses 5 in the signal U _a is shifted towards negative pulses, which leads to a decrease in the average value of U _a , which reaches zero at e _o ≃ E _o ,, that is, near the fine tuning of the CAS. In the case of further growth of CAS, negative pulses 5 begin to prevail over positive ones, that is, when the active component is overcompensated, the average value of U _a becomes negative.

в момент скачка отрицательна на падающих участках E(t) и положительна на восходящих участках ЭДС E(t) (позиция 4 на фиг.4). Результат синхронного детектирования импульсов 4 опорным сигналом 7 показан на фиг.4 в позиции 8 и представляет собой последовательность положительных импульсов 9 (что влечет за собой положительный знак среднего значения U_т). Кроме того, на фиг.4 так же, как и на фиг.3 большинство дуговых пробоев с положительной производной 4 происходит при положительных значениях ЭДС E(t) и наоборот, что определяет преобладание положительных импульсов в результате 5 синхронного детектирования сигнала 4 опорным сигналом 1 и как следствие положительный знак среднего значения U_a (что и должно происходить при недокомпенсации активной составляющей). Как и ранее, при введении КАС происходит смещение соотношения положительных и отрицательных импульсов в сигнале 5 и изменение знака среднего значения 6U_a вблизи точки точной настройки КАС. Описанные свойства средних значений сохраняются при изменениях напряжения пробоя дугового промежутка в месте ЗНЗ.Finally, in the case of undercompensation in the capacitive component in the absence of CAS, the following conditions are satisfied: e _o ≃ 0, ω _C <ω, Φ <β. This case is illustrated in figure 4. Due to the fact that ω _C <ω, after each arc breakdown (at times t _o , t ₁ , t ₂ , etc.), the curve e (t) (position 2 in FIG. 4) deviates from curve 1 E ( t) in the direction of the lag, that is, up from the curve E (t) in its falling sections and down from it in the ascending sections. As a result, the derivative

at the time of the jump, it is negative in the falling sections of E (t) and positive in the ascending sections of the EMF E (t) (position 4 in FIG. 4). The result of the synchronous detection of pulses 4 by the reference signal 7 is shown in figure 4 at position 8 and represents a sequence of positive pulses 9 (which entails a positive sign of the average value of U _t ). In addition, in Fig. 4, as in Fig. 3, the majority of arc breakdowns with a positive derivative 4 occurs at positive EMF values E (t) and vice versa, which determines the predominance of positive pulses as a result of 5 synchronous detection of signal 4 by reference signal 1 and as a result, a positive sign of the average value of U _a (which should happen when the active component is undercompensated). As before, with the introduction of CAS, the ratio of positive and negative pulses in signal 5 shifts and the sign of the average value 6U _a changes near the point of CAS fine tuning. The described properties of the average values are preserved with changes in the breakdown voltage of the arc gap at the ZNZ site.

It follows from the foregoing that the sign of the averaged signal U _t corresponds to the detuning according to the CES, and the sign of the averaged signal 6 U _a corresponds to the detuning according to the CAS and, therefore, the above variables can be used to automatically adjust the CES and CAS (using, for example, integral, proportional-integral , proportional-integral-differential or more complex methods of regulation).

, возникающие в моменты дуговых пробоев, синхронны и синфазны с импульсами тока I_g(t) в месте повреждения. Поэтому результаты операций синхронного детектирования функции

несут ту же информацию, что и измерение фазы первой гармоники тока I_g(t) (в том случае, если дуговые пробои возникают периодически), который ненаблюдаем со стороны системы автокомпенсации. Фаза же первой гармоники указанного тока несет, как известно, информацию о величинах и знаках расстроек компенсации.The operation of this method can be further explained in frequency terms. In this case, attention should be paid to the fact that the momenta of the function

arising at the moments of arc breakdowns are synchronous and in phase with current pulses I _g (t) at the site of damage. Therefore, the results of synchronous function detection operations

carry the same information as the measurement of the phase of the first harmonic of the current I _g (t) (in the event that arcing breakdowns occur periodically), which is unobservable by the auto-compensation system. The phase of the first harmonic of the indicated current carries, as is known, information on the magnitudes and signs of the compensation disturbances.

находим:

При этом считаем, что мгновенная фаза ωt_o приведена к отрезку [0, 2π]. Таким образом, для первоначального нахождения фазы β необходимо выполнить следующие действия: выявить дуговой пробой в месте ЗНЗ, определить мгновенную фазу wt_o момента t₀ этого пробоя и вычислить искомую фазу β согласно (12). Для распознавания моментов дуговых пробоев можно воспользоваться, например импульсами, возникающими в эти моменты при дифференцировании напряжения e(t) смещения нейтрали согласно заявляемому способу.In order to determine the value of the phase β of the EMF E (t) acting between the point of the fault and the neutral of the network, the voltage property e (t) noted above can be used, the neutral bias becomes equal to the instantaneous value of the EMF E (t) at the moment of termination of the arc process in place ZNZ (which is reflected in the ratio (4)). If we assume that the first arc fault occurs at the moment t _{0 of the} maximum (in absolute value) voltage U _g (t) at the damage site, that is, at the time of the maximum (in absolute value) EMF
E (t) = E _o cos (ωt + β), (11)
acting between the point of the ZNZ and the neutral of the network, then from the above maximum condition:

we find:

Moreover, we believe that the instantaneous phase ωt _{o is} reduced to the segment [0, 2π]. Thus, to initially find the phase β, it is necessary to perform the following steps: to identify the arc breakdown at the ZNZ site, determine the instantaneous phase wt _{o of the} moment t _{0 of} this breakdown, and calculate the desired phase β according to (12). To recognize the moments of arc breakdowns, you can use, for example, pulses arising at these moments during differentiation of the neutral bias voltage e (t) according to the claimed method.

Данная система совместима в том случае, если
sinω(t_i-1-t_i)≠0, (15)
то есть, если величина ω(t_i-1-t_i) не кратна π.. Ее решение дает следую- щее выражение для фазы β:::

Таким образом, если в сети произошло более одного дугового пробоя (в моменты времени t_i-1 и t_i, для которых выполняется условие (15)), то для определения фазы β следует зафиксировать мгновенные фазы wt_i-1 и ωt_i моментов пробоя, определить мгновенные значения e(t_i-1) и e(t₁) напряже- ния смещения нейтрали в эти моменты и вычислить упомянутые фазу сог- ласно выражению (16).If more than one arc breakdown has been registered since the occurrence of the fault, it is possible to increase the accuracy of determining phase b by abandoning the assumption that arc breakdowns occur at the moments of the maxima of the absolute value of the EMF E (t) and using instead additional information on the instantaneous phases wt _i moments t _i other arc breakdowns and about the instantaneous values e (t), voltage e (t) of the neutral bias at the indicated moments. Indeed, having written relation (4) for two arc breakdowns that occurred at neighboring times t _i-1 and t _i , we obtain the following system of equations:

This system is compatible if
sinω (t _i-1 -t _i ) ≠ 0, (15)
that is, if the quantity ω (t _i-1 -t _i ) is not a multiple of π .. Its solution gives the following expression for the phase β :::

Thus, if more than one arc breakdown has occurred in the network (at time t _i-1 and t _i , for which condition (15) is satisfied), then to determine the phase β one should fix the instantaneous phases wt _i-1 and ωt _i of the breakdown time , determine the instantaneous values e (t _i-1 ) and e (t ₁ ) of the neutral bias voltage at these moments and calculate the mentioned phase according to expression (16).

 If a series of several arc breakdowns occurred during the ZNZ, it is advisable to determine the phase β as described above for each pair of arc breakdowns and then average the obtained values of phase b .. The simplest and most effective methods of averaging are to determine the arithmetic mean over all that occurred to a given moment of arc breakdowns or the determination of a moving average from a predetermined number of arc breakdowns.

действующей между точкой ЗНЗ и нейтралью сети должен располагаться либо на одной из сторон треугольника линейных ЭДС, если ЗНЗ произошло или в одной из обмоток, соединенных в треугольник, либо на одном из лучей звезды фазных ЭДС, если ЗНЗ произошло или в одной из обмоток, соединенных в звезду, или в фазном проводнике сети. Привлечение ука- занной априорной информации означает, что полученная в результате пре- дыдущих действий величина β должна быть скорректирована путем прое- цирования вектора

, параметры которого определены в ходе описанных выше действий, на ближайшую из сторон треугольника линейных ЭДС или на ближайший из лучей звезды фазных ЭДС. Для этого, при каждом из дуговых пробоев, дополнительно к перечисленным выше действиям, определяют амплитуду

, действующей между точкой ЗНЗ и нейтралью сети, следующим образом:
если имеется информация только лишь о первом пробое в момент времени t₁, то полагают:

если имеется информация о двух дуговых пробоях в моменты времени t_i-1 и t_i, для которых выполняется условие (15), то по- лагают

если имеется информация более чем о двух пробоях, то для каждой пары из них определяют величину

по формуле (18), после чего полученные величины осредняют.An additional increase in the accuracy of determining the phase β is achieved by using a priori information that the end of the vector

acting between the ZNZ point and the neutral of the network should be located either on one side of the triangle of linear EMF, if the ZNZ occurred or in one of the windings connected in a triangle, or on one of the beams of the star of the phase EMF, if the ZNZ occurred or in one of the windings connected in a star, or in the phase conductor of a network. Attraction of the indicated a priori information means that the value β obtained as a result of previous actions should be corrected by projecting the vector

, the parameters of which are determined in the course of the above steps, to the nearest of the sides of the triangle of linear EMF or to the nearest of the rays of the star phase emf. For this, with each of the arc breakdowns, in addition to the above actions, the amplitude is determined

acting between the ZNZ point and the network neutral, as follows:
if there is information only about the first breakdown at time t ₁ , then consider:

if there is information about two arc breakdowns at times t _i-1 and t _i for which condition (15) is satisfied, then

if there is information on more than two breakdowns, then for each pair of them determine the value

by the formula (18), after which the obtained values are averaged.

отклонения Φ=β-α_i значения фазы β ЭДС, действующей между точкой замыкания на землю и нейтралью сети от значения α_i фазы фазной ЭДС E_i=E_mcos(ωt+α_i) не превышает π/3, а затем, если удовлетворяется условие

где E_m амплитуда фазной ЭДС сети, то скорректированную величину фазы β ЭДС, действующей между точкой замыкания на землю и нейтралью сети, считают равной фазе a_i упомянутой фазной ЭДС; в противном слу- чае, то есть, если условие (19) не удовлетворяется, то проверяют вы- полнение другого условия:

и, если данное условие выпол- няется, то также полагают β=α_i. Если же условия (19), (20) не выполняет- ся, то величину β определяют по следующей формуле:

Для исключения влияния на автонастройку компенсации высокочастотных помех, присутствующих в напряжении смещения нейтрали, последнее перед выполнением операции дифференцирования должно быть подвергнуто фильтрации для подавления составляющих, лежащих вне спектра процессов в КНПС при дуговых ЗНЗ (ориентировочно с частотами свыше 10 кГц). Поскольку форма сигнала

в течение дугового пробоя не несет информации о знаках и величинах расстроек по КАС и КЕС, данный сигнал целесообразно ограничить сверху и снизу. Ограничение же при достаточно большом усилении приводит, по сути дела, к превращению данного сигнала в логический, что упрощает реализацию операций синхронного детектирования.Next, determine the phase (A, B or C) of the network for which the absolute value

deviations Φ = β-α _{i of the} value of phase β of the EMF acting between the point of earth fault and the neutral of the network from the value α _{i of the} phase phase EMF E _i = E _m cos (ωt + α _i ) does not exceed π / 3, and then, if the condition is satisfied

where E _{m is the} amplitude of the phase EMF of the network, then the adjusted value of the phase β EMF acting between the earth fault point and the neutral of the network is considered equal to the phase a _{i of the} mentioned phase EMF; otherwise, that is, if condition (19) is not satisfied, then the fulfillment of another condition is checked:

and if this condition is satisfied, then β = α _{i is} also assumed. If conditions (19), (20) are not satisfied, then the value of β is determined by the following formula:

In order to exclude the influence of high-frequency interference present in the neutral bias voltage on the auto-tuning, the latter must be filtered before suppressing the differentiation operation to suppress components that are outside the spectrum of processes in the SSC at arc faults (approximately with frequencies above 10 kHz). Since the waveform

during the arc breakdown does not carry information about the signs and magnitude of the detuning according to CAS and CES, it is advisable to limit this signal from above and below. The limitation with a sufficiently large gain leads, in fact, to the conversion of this signal into a logical one, which simplifies the implementation of synchronous detection operations.

, имеющей импульсный характер (в режиме перемежающегося дугового ЗНЗ), их абсолютные величины растут с ростом частоты пробоев, например из-за больших расстроек или низкого напряжения пробоя дугового промежутка. Данное обстоятельство делает весьма эффективным комбинирование заявляемого способа с другими известными способами автонастройки компенсации, включая способ-прототип.The inventive method does not use the inherent in all known methods of auto-compensation assumptions that ZNZ occurred precisely between the phase of the network and the ground. Therefore, it does not require recognition of the damaged phase. Due to this feature, the method is operable in the event of an SCZ not only in the phase conductors of the network, but also in the load windings or transformers supplying the network, connected both to a star and to a triangle. Information on the EMF phase E (t) can be obtained, for example, by measuring the instantaneous phases of the moments of arc breakdowns and the instantaneous values of the neutral bias voltage e (t) at the indicated times. Moreover, this method fundamentally does not need to recognize the network operating modes, since in the absence of arc breakdowns in the normal network operation mode or in arc-free ZNZ modes (for which the method is not intended) it generates zero equations. Then, however, that mode recognition may be needed in cases where the mode information requires the used methods of CAS or CES implementation, or when the claimed method is combined with other methods of automatic compensation adjustment. Due to the relative independence of the CAS control sign from the CES detuning (and vice versa), the dynamic properties of systems implementing the inventive method are improved. Due to the fact that the controls in the proposed method are formed by processing functions

having a pulsed nature (in the intermittent arc fault mode), their absolute values increase with increasing breakdown frequency, for example, due to large detunings or low breakdown voltage of the arc gap. This circumstance makes it very effective to combine the proposed method with other known methods of automatic compensation, including the prototype method.

 Due to the relative complexity of the actions provided by the claimed method, it is advisable to implement it on a microprocessor basis.

 Consider the operation of a device that implements the inventive method and shown in Fig.5. The basis of the device is a processor 26 with a constant (ROM) and operational (RAM), memory device (PRZ), which as follows interacts with other blocks of the device. The processor 26 reads the next program command from the ROM, decodes it, and then executes, referring to the contents of the command, to the RAM or ROM of block 26, and through bus 27 also to the additional block 25 for pairing the DBSS signals, timer 35, to the controller 31 interrupts, to ports 28, 29 and to blocks 35 and 36 of the formation of the control of BFUV and BFUR according to the active and reactive components. When a signal edge arrives at one of the inputs of Ir1.Ir3 of controller 31 of interrupt 1, the latter, interacting with block 26 of the ROM, forces it to interrupt the execution of the current program, perform one of the procedures for processing interrupts recorded in the ROM of block 26 of the RAM, and then return to execution interrupted program.

Let us further consider the process of general synchronization of a device with industrial frequency. The signal U _AB (t), which is proportional to the line voltage of the network between phases A and B, is generated by a voltage measuring transformer 13 (Fig. 2), fed to the input of the null-comparator 32 of the BUK unit (Fig. 7) and then passes through the former 33 and 34 pulses excluding bounce, respectively, on the decline and on the front of this signal. The front of the signal at the output of the shaper 33, which occurs almost simultaneously with the transition of the linear voltage U _AB (t) through zero (from negative values to positive), restarts timer 35. Timer 35 starts counting clock pulses that arrive at its counting input “s” s frequency, 3 5 orders of magnitude higher than the network frequency. The count begins with a code approximately corresponding to the half-period of the network. As a result, using the code read from this timer by the RAM block 26 at any time t, it is not difficult to restore the instantaneous phase ωt of the reading time t, and the time difference between these events can be determined from the difference between the codes read synchronously with any events. With the described synchronization method, all phase shifts are counted relative to the linear voltage U _AB (t) (which has a zero phase shift in this connection). At the moment the timer code 35 passes through the zero value, a pulse is generated at its output, which initiates (via the interrupt controller 31) a call by the block 26 of the ROM of the interrupt processing routine at the input Ir1 of the controller 31. This procedure involves polling (via port 28 P1) the status of the comparator 32 the moment in time when the instantaneous phase ωt according to the readings of timer 35 is π. Since at this moment the signal U _AB (t) must pass through zero, by the state of the comparator 32 (that is, by the signal at the output of the driver 34), we can judge the correspondence of the time delay of the timer 35 to the actual duration T of the half-period at a given network frequency, as well as adjust it, that is, slightly increase this shutter speed if a logical zero is present at the output of the shaper 34 (U _AB <0) or reduce it otherwise (which should also be provided for by the interrupt processing procedure at the input Ir1 of the controller 31). The tracking system thus formed, tracking changes in the network frequency, eliminates their influence on the functioning of the device.

 In the normal mode of operation of the network, there are no abrupt changes in the neutral bias voltage signal e (t), therefore, there are also no pulses at the input Ir2 of the interrupt controller 31. Block 17 in the considered mode is transferred (by the signal z of port 29) to the state of the inertial link with a small time constant (for which the analog switch 20 connects the input of the operational amplifier 18 to the connection point of the resistors 21, 22) and therefore monitors the voltage e (t ) neutral bias. The interrupt at the input Ir3 of the controller 31 is caused if the signal g at the output of port 2 does not correspond to the signal h at the output of the comparator 24 (as a result of which the logical "unit" appears at the output of the logic element 30 "DISCHARGE", which exactly causes this interrupt ) The interrupt processing routine at input Ir3 maps the output signal g of port 31 to the output signal h of comparator 24. Such a correspondence is a normal state of the considered example of the device. A device in normal network operation mode must either support some preset CES configuration, for example, that remained from the previous ZNZ mode, or carry out preliminary (resonant) KNPS tuning, implementing any of the known methods. The necessary signals for this are generated by block 25 DBSS. The executive bodies 37 and 38 are controlled by block 26, according to the program implemented in the normal mode, by means of interface blocks 35 and 36 (the device of which is determined both by the selected CAS and CES methods, and by the device of blocks 37 and 38).

 When an arcing fault occurs in the network, an abrupt change in the neutral bias voltage e (t), measured by the ITN block 13 and transmitted through a filter 14 Ф1 (suppressing high-frequency interference) to the differentiator 15, generates a pulse signal of the corresponding sign at the output of the latter. This signal is converted by the limiter 16 (which is, for example, a relay link with a dead zone) into a short-term signal of a fixed level (or into a two-bit logic signal, one of which is significant) and is fed to the inputs of the interrupt controller 31 and port 28.

в момент пробоя), а также посредством порта 28 П1 считывает и запоминает в ОЗУ значение сигнала h, то есть по сути дела знак мгновенного значения e(t_o+δ) напряжения e(t) смещения нейтрали сразу после пробоя. Помимо этого, в процессе выполнения данной процедуры блок 26 ПРЗУ изменяет сигнал z на выходе порта 29, вследствие чего блок 1 переходит в режим преобразования во временной интервал того мгновенного значения e(t_o+δ) напряжения e(t) смещения нейтрали, которое было зафиксировано сразу же после дугового пробоя. В указанном режиме аналоговый ключ 20 устанавливается в положение, показанное на фиг.5, и выходное напряжение интегратора (на элементах 18, 19 и 23) начинает изменяться с постоянной скоростью в сторону уменьшения по абсолютной величине (независимо от знака e(t_o+δ)).).In addition, an abrupt change in the neutral bias voltage e (t) is monitored by an inertial link with a small time constant (elements 18 to 22 in FIG. 5). The pulse at the input Ir2 of the interrupt controller 31 forces the latter to initiate a call to the corresponding interrupt processing procedure, during which the RAM block 26 reads and stores the contents of timer 35 in RAM (that is, it records information about the instantaneous phase wt _{o of the} moment t _{0 of the} arc breakdown), port status 28 (i.e. captures information about the sign of the derivative

at the time of breakdown), and also through port 28 P1 reads and stores in RAM the value of the signal h, that is, in fact, the sign of the instantaneous value e (t _o + δ) of the neutral bias voltage e (t) immediately after the breakdown. In addition, in the process of performing this procedure, the RAM block 26 changes the signal z at the output of port 29, as a result of which block 1 switches to the conversion mode in the time interval of that instantaneous value e (t _o + δ) of the neutral bias voltage e (t), which was fixed immediately after the arc breakdown. In this mode, the analog switch 20 is set to the position shown in Fig. 5, and the output voltage of the integrator (on elements 18, 19 and 23) starts to change at a constant speed in the direction of decrease in absolute value (regardless of the sign of e (t _o + δ )).).

After a certain time interval Δt (proportional to the instantaneous value e (t _o + δ), the neutral bias voltage e (t) immediately after the arc breakdown), the signal at the output of the specified integrator reaches zero, and then changes sign. At the moment of changing the sign of the signal at the output of the operational amplifier 18, it is violated between the h and g signals, respectively, as a result of which a logical "unit" appears at the output of the element 30 and the interrupt controller 31 initiates a call by the block 26 of the ROM to the interrupt processing routine at the input Ir3. This procedure reads and stores the contents of the timer 35 in the RAM unit 26, matches the signals h and g, and changes the signal z again, thereby returning the unit 17 to the inertial link mode with a small time constant. The device is prepared to receive similar information during the next arc breakdown (if it occurs). The described steps prepare the initial information for auto-tuning CES and CAS in accordance with the claims.

напряжения смещения нейтрали, имевшая место сразу после дугового пробоя, а по состоянию сигнала g в момент пробоя знак этого напряжения. По содержимому таймера 35 в момент t_o+δ определяется также мгновенная фаза ωt_o момента пробоя. На основании полученных данных блок 26 ПРЗУ рассчитывает по формуле (12) фазу β ЭДС, действующей между точкой ЗНЗ и нейтралью сети, которая является фазой первого опорного сигнала
b_r(t)=cos(ωt+β), (22)
а также фазу

второго опорного сигнала

(см. формулу изобретения). Для выполнения предусмотренных формулой изобретения операций синхронного детектирования блок 25 ПРЗУ вычисляет величины

которые используются далее этим же блоком для формирования управлений по КАС и по КЕС в соответствии с интегральным, пропорционально-интегральным, пропорционально-интегрально-дифференциальным или более сложными способами регулирования.Next, the block 26 of the RAM processes the received information. Based on the difference in the contents of the timer 35, read at the moment of changing the sign of the signal at the integrator output (interrupt at the input of Ir3) and read at the moment t _o + δ of the arc breakdown (interrupt at the input of Ir2), the absolute value is determined

neutral bias voltage, which took place immediately after the arc breakdown, and according to the state of the signal g at the time of breakdown, the sign of this voltage. The contents of the timer 35 at the time t _o + δ also determines the instantaneous phase ωt _{o of the} moment of breakdown. Based on the received data, the RPSU block 26 calculates, according to formula (12), the phase β EMF acting between the point of the short-circuit current and the neutral of the network, which is the phase of the first reference signal
b _r (t) = cos (ωt + β), (22)
as well as phase

second reference signal

(see the claims). To perform the synchronous detection operations provided for by the claims, the RAM unit 25 calculates

which are then used by the same unit to form controls for CAS and CES in accordance with integral, proportional-integral, proportional-integral-differential or more complex control methods.

согласно выражениям, соответственно, (16) и (18), которые затем осредняются и используются этим же блоком 26 для коррекции величины β описанным выше способом (зависящим от выполнения условия (19), (20)). Полученная величина b, как и ранее используется блоком 26 ПРЗУ при выполнении операций синхронного детектирования в соответствии с выражениями (24), (25), результаты которых, в свою очередь, используются при формировании управлений по КАС и по КЕС.When subsequent breakdowns occur, the device acts in a similar way, but the information collected during the processing of interrupt processing at the inputs Ir2, Ir3 of the control unit 31 is processed somewhat differently. This means that according to the calculated (as before) values of ωt _i and e (ωt _i ), for each of the subsequent breakdowns, the block 26 of the RAM determines β ′ and

according to expressions, respectively (16) and (18), which are then averaged and used by the same block 26 to correct the value of β in the manner described above (depending on the fulfillment of condition (19), (20)). The obtained value of b, as previously used by the block 26 of the RAM when performing synchronous detection operations in accordance with expressions (24), (25), the results of which, in turn, are used in the formation of controls for CAS and CES.

 As the CAS and CES fine-tuning are achieved, the arc breakdowns at the ZNZ site cease, they interrupt the calls to the interrupt processing procedure at the input Ir2 of the interrupt controller 31, the CAS and CES control take zero values and the device further supports the achieved fully compensated neutral mode.

 For the transfer to the executive bodies 37 IOR DGA and 38 IOA of the active component compensator, CAS and CES controls received in digital block 26 from the PRZU, control formation blocks are used, respectively, 36 BFUR and 35 BFUA. The principles of operation of the blocks are determined by the applied methods of CAS and CES, as well as the principles of action of the executive bodies 37 and 38. So, for example, if the executive body 37 IOR DGA (by analogy with the device that implements the prototype method) is a thyristor key in the open triangle circuit of the secondary windings Bauch transformer, together with galvanic isolation circuits and thyristor blocking sensors, current block 36 BFUR is a programmable timer that determines the time delay between the moment the thyristor is locked key and the moment of filing on it the next unlocking pulse. If the executive body 37 of the IOR is a controlled rectifier of an arc suppression reactor with magnetization, then the BFUR block 36 is a programmable timer with synchronization and galvanic isolation circuits, which determines the opening angle of the thyristors of the controlled rectifier. If the executive body 37 of the IOR is the drive of the arcing reactor plunger, then the BFUR block 36 is an output (output) port with galvanic isolation circuits, the signals of which control the actuators of the reactor drive. Further, if the executive body 38 of the IOA of the compensator 12, the active component is (by analogy with the device that implements the prototype method) a thyristor key in the circuit of an additional inductor, then the construction of block 35 BFUA is similar to the construction of block 36 BFUR for the same case. The signal ρ “mode” necessary for the functioning of such an equalizer of the active component, which determines the state of its switching devices, is generated by the PRZU block 26 and transmitted through the output port 29 of P2. If the IOA block 38 is a single-phase dependent inverter, by analogy with (Obabkov V.K., Tseluevsky, Yu.N. A method for compensating single-phase currents in a three-phase network with an extinguishing reactor. Aut. St. USSR N 1264263, class N 02 N 9 / 08), then the construction of block 35 BFUA is similar to the construction of block 36 BFUR described above for the case of an arcing reactor with magnetization.

 It should be noted that in conditions of fine-tuning the CAS and CES, a change in the ZZZ parameters, including restoration of the dielectric strength of the insulation, does not affect the first harmonics of the voltage and current in the network. Therefore, it is obvious that in order to recognize the fact of repairing the damage and restoring the normal operation of the network, a number of actions are required. In this case, you can use any known method of recognition and restoration of the normal mode of operation of the network, for example, proposed in the prototype. The necessary signals for this are generated by block 25 DBSS. The same block is used to form the missing input signals in the case when it is assumed that the inventive method will be combined with any other (known) methods in the auto-compensation device (for example, with the prototype method).

Claims (1)

A method for automatically adjusting the compensation of capacitive and active components during an arc earth fault, according to which the neutral bias voltage is measured and control signals are generated to compensate the active and capacitive components of the earth fault currents, characterized in that the EMF phase value is determined, differentiated, limited and synchronously detected by two mutually orthogonal reference signals, the phase of the first of which is set equal to the phase of the EMF, the phase of the second is set equal to different from the above the first phase of the EMF by 90 ^o and the first result obtained is used to generate a control signal to compensate for the active component of the earth fault current, and the second to generate a control signal to compensate for the capacitive component of the earth fault current.

Images