RU205145U1 - Arc suppression control device - Google Patents

Abstract

The utility model relates to electrical measuring equipment and can be used at electrical substations to control the setting of arc suppression reactors that compensate for capacitive earth fault currents in 6-35 kV networks. The device contains a differential amplifier, a computing module equipped with an analog-to-digital converter, and connected to the interface computing module, an oscillation exciter and a unit for interfacing with an arc suppression reactor inductance regulator, a memory unit, a period and averaging unit, a looping unit, an adder, a frequency determination unit and a control action calculation unit, and the output of the differential amplifier is connected to the input of the analog-to-digital converter, the output which is connected to the input of the memory block and the direct input of the adder, the output of the memory block is connected to the input of the block for determining the period and averaging, the output of the block for determining the period and averaging is connected to the input of the cyclic replay unit product, the output of the cyclic reproduction unit is connected to the subtractive input of the adder, the output of the adder is connected to the input of the frequency determination unit, the output of the frequency determination unit is connected to the input of the control action calculation unit, the output of the control action calculation unit is connected to the interface input of the computing module. 2 ill.

Description

The utility model relates to electrical measuring equipment and can be used at electrical substations to control the setting of arc suppression reactors that compensate for capacitive earth fault currents in 6-35 kV networks.

To compensate for capacitive earth fault currents in 6-35 kV networks, tunable arc suppression reactors (DGR) are used, which are included in the zero sequence circuit of the network (KNPS), for example, in the neutral of a power transformer. Capacitive currents are compensated when the natural frequency of the LCC, regulated by changing the inductance of the GDR, coincides with the frequency of the mains voltage (industrial frequency).

Known is a control device for setting the GDR, described in patent RU 2475915, published on 02/20/2013, IPC H02J 3/18 (2006.01), which generates control actions on the inductance regulator of the GDR to compensate for capacitive earth fault currents and contains a computing module equipped with an analog-to-digital converter , designed to digitize the bias voltage of the neutral (NCH), in which the GDR is included. The computational module of the device for separating the free component of the excited oscillations fixes the NCH and extracts the difference signal obtained by superimposing two sections of this voltage curve, recorded before and after the action of the excitation pulse.

The disadvantage of the described device is the low reliability of determining the natural frequency of the LCC, the dependence of the regulating action formed after each pulse of the reference current on the magnitude and phase of the LCC of the network and, as a consequence, the low accuracy and protracted nature of the compensation process for capacitive earth fault currents.

This disadvantage is due to the following. In the absence of an excitation pulse, the level of the monitored NSN signal is determined by the asymmetry of the mains phases, and the frequency coincides with the frequency of the mains voltage. After the action of the excitation pulse, the free component of the transient process is added to the NSN, for a short decay time of which the NSV remains practically unchanged in magnitude, frequency and phase. When the free component of the transient process is selected, which carries information about the natural frequency of the CNPS, the NSN appears as an interference, the elimination of which is directed by the procedure of superimposing signals (fragments of the controlled NSN), recorded before and after the action of the excitation pulse, respectively. In this case, the first signal contains the specified interference in its pure form, and the second - in sum with the free component of the resulting transient process.

When the fragments recorded according to the patent RU 2475915 are superimposed before and after the action of the excitation pulse, the HCH phases on the superimposed fragments are random and the elimination of the interfering HCH from the received signal during superposition (addition or subtraction) occurs only with a small probability. The signal obtained as a result of such an overlap will in most cases contain not only the free component of the transient process, but also noise, the magnitude of which will depend on the amplitude and random phase difference of the NCH in the fixed and superimposed signals. The natural frequency of the LCC and the regulating effect on the GDR based on such a signal are determined unreliable, the tuning process becomes unstable, and the earth fault current remains uncompensated.

The closest in technical essence to the proposed device (prototype) is the device described in the patent RU 147273 U1 "Device for controlling the setting of the arc-extinguishing reactor", published on October 27, 2014, IPC H02J 3/18 (2006.01). The device contains a computing module equipped with an analog-to-digital converter designed to digitize the bias voltage of the neutral, into which the arc suppression reactor is connected, and an oscillation exciter and an interface unit with the inductance regulator of the arc suppression reactor connected to the interface of the computing module, while the computing module is made with the possibility of delaying the digitized signal coming from the analog-to-digital converter for an integer number N of half-periods of industrial frequency, separating the free component of excited oscillations in the form of the sum or difference of the current and delayed digitized signals with an odd or even number N, respectively, determining the frequency of the isolated free component and generating control actions on regulator of the inductance of the arc suppression reactor, bringing the specified frequency closer to the industrial frequency.

The disadvantages of the prototype are the high error in determining the natural frequency of the CNPC and, as a consequence, the low accuracy of the resonant tuning of the CNPC due to the possible incomplete compensation of the periodic interference, as well as due to the amplification of random noise present in the input signal of the NSN.

раз по сравнению с исходным. Вследствие этого собственная частота КНПС может определяться некорректно, а резонансная настройка КНПС будет неточной.These disadvantages are due to the following. Firstly, the shapes of the positive and negative half-periods of the NCH may differ, which in the case of adding the current and delayed digitized signals with an odd number of half-periods N will lead to incomplete compensation of the NCH and, as a consequence, to distortion of the shape of the investigated free component of the transient process and to an increase in the error or to the complete impossibility of determining the natural frequency of the CNPC. Secondly, the current and delayed signals in real conditions contain random components due to noise and interference. When these signals are added or subtracted, the variances of their random components are added, and the standard deviation (RMSD), which determines the random component of the error in the values of the transient process under study, increases by

times compared to the original. As a result, the natural frequency of the LOC may be determined incorrectly, and the resonant tuning of the LOC will be inaccurate.

The task of the proposed technical solution is to improve the accuracy of the resonance tuning of the KNPS. The technical result consists in increasing the accuracy of measuring the natural frequency of the CNPC by reducing the sensitivity of the device to random and periodic interference.

The task is achieved by the fact that a storage unit is additionally introduced into the control device of the arc suppression reactor, which contains a differential amplifier, a computing module equipped with an analog-to-digital converter (ADC), and an oscillation exciter and an interface unit with an inductance regulator of the arc suppression reactor connected to the interface of the computing module, block for determining the period and averaging, block for cyclic playback, adder, block for determining the frequency and block for calculating the control action, and the output of the differential amplifier is connected to the input of the ADC, the output of the ADC is connected to the input of the memory block and the direct input of the adder, the output of the memory block is connected to the input of the determination block period and averaging, the output of the period determination and averaging unit is connected to the input of the cyclic playback unit, the output of the cyclic playback unit is connected to the subtractive input of the adder, the adder output is connected to the input of the frequency determination unit s, the output of the unit for determining the frequency is connected to the input of the unit for calculating the control action, the output of the unit for calculating the control action is connected to the input of the interface of the computing module.

FIG. 1 shows a functional diagram of the device, and FIG. 2 - time diagrams explaining its operation, and in FIG. 2a shows the waveform at the ADC output, FIG. 2b - the shape of the synthesized compensation signal at the output of the loop reproducing unit, and FIG. 2c - the shape of the signal at the output of the adder, corresponding to free oscillations in the CNPC.

In the diagram of FIG. 1 shows a computing module 1 equipped with an ADC 2. An oscillation exciter 4 and an interface unit 5 are connected to the interface 3 of the computing module 1. In addition, FIG. 1 shows a tunable GGR 6 through which the neutral of the power transformer 7 connected to the electrical network is grounded, and the distributed capacitances 8 of the phases of this network to the ground. DGR 6 is equipped with a signal winding 9 and an inductance regulator 10 (for example, a DGR plunger drive).

The output of the exciter 4 is connected to the signal winding 9. The input of the ADC 2, intended for digitizing the bias voltage of the neutral, grounded through the GDR 6, can be connected, by analogy with the prototype, to the winding "open delta" of the zero sequence voltage measuring transformer or, in the frequent case, shown on fig. 1, - to the signal winding 9 through a differential amplifier 11. The output of the interface unit 5 is connected to the input of the inductance regulator 10.

The output of the ADC 2 is connected to the input of the memory unit 12 and the direct input of the adder 13, the output of the memory unit 12 is connected to the input of the period determination and averaging unit 14, the output of the period determination and averaging unit 14 is connected to the input of the cyclic playback unit 15, the output of the cyclic playback unit 15 is connected to the subtractive input of the adder 13, the output of the adder 13 is connected to the input of the frequency determination unit 16, the output of the frequency determination unit 16 is connected to the input of the control action calculation unit 17, the output of the control action calculation unit 17 is connected to the input of the interface 3 of the computing module 1.

The device works as follows.

длительностью T0 путем усреднения формы фрагментов x_i(t).ADC 2 digitizes the signal supplied to its input from the signal winding 9 through the amplifier 11 and is proportional to the neutral bias voltage (NCH). At the beginning of the next working cycle, starting from the time t ₁ and during the time interval T ₁ (Fig. 2), the storage unit 12 stores the samples of the signal x (t), which are received from the ADC 2, corresponding to the NCH before the excitation of free oscillations in the LSC. The interval T ₁ is chosen in such a way that K signal periods fit into it, taking into account the possible deviation of the mains frequency from the nominal value (K is an integer). After the expiration of time T _{1, the} block for determining the period and averaging 14 determines the period T0 of the stored signal and divides the stored signal x (t) into K fragments x _i (t) of duration T0 each (Fig.2a), after which it forms a compensation fragment

duration T0 by averaging the shape of the fragments x _i (t).

на вычитающем входе сумматора 13 необходимое количество раз (фиг. 2б). Фактически компенсационный сигнал на выходе блока 15 является задержанным на М периодов и усредненным по форме сигналом x(t), сохраненным в запоминающем блоке 12 до воздействия возмущающего импульса. На прямой вход сумматора 13 с выхода АЦП 2 поступает оцифрованный сигнал q(t), изображенный на фиг. 2а и являющийся суммой затухающей свободной составляющей переходного процесса y{t), возбуждающего импульса p(t) и периодической составляющей x'(t), причем из-за небольшой величины разности t₂ - t₁ частоту и форму x'(t) можно считать идентичными частоте и форме компенсационного сигнала, формируемого блоком 15. В результате сигнал на выходе сумматора (фиг. 2в) содержит только возбуждающий импульс p(t) и свободную составляющую y(t). Получив этот сигнал, блок определения частоты 16 отбрасывает его начальную часть, в которой присутствует короткий возбуждающий импульс, и вычисляет частоту свободной составляющей переходного процесса y(t). Блок вычисления управляющего воздействия 17 сравнивает эту частоту с частотой сети f₀=1/T₀ и на основании полученной разности вычисляет величину управляющего воздействия, которое необходимо оказать на ДГР 6 с целью приближения собственной частоты КНПС к частоте сети на заданную величину. Это воздействие через интерфейс 3 и блок сопряжения 5 передается на регулятор индуктивности 10 ДГР 6.Then, at the nearest time moment t ₂ = t ₁ + M × T ₀ , where M is an integer, the computing unit 1 through the interface unit 3 acts on the oscillator 4, which issues a short excitation pulse to the signal winding 9 of the GDR 6, causing a decaying transient process in KNPS. At the same time, the loop reproducing unit 15 begins to generate a compensation signal, repeating the compensation fragment

at the subtractive input of the adder 13 the required number of times (Fig. 2b). In fact, the compensation signal at the output of unit 15 is delayed for M periods and a waveform-averaged signal x (t) stored in the memory unit 12 before the disturbance pulse. The direct input of the adder 13 from the output of the ADC 2 receives a digitized signal q (t), shown in FIG. 2a and is the sum of the damped free component of the transient process y (t), the exciting pulse p (t), and the periodic component x '(t), and due to the small value of the difference t ₂ - t _{1, the} frequency and shape x' (t) can be considered identical to the frequency and shape of the compensation signal generated by unit 15. As a result, the signal at the output of the adder (Fig. 2c) contains only the exciting pulse p (t) and the free component y (t). Having received this signal, the unit for determining the frequency 16 discards its initial part, in which there is a short exciting pulse, and calculates the frequency of the free component of the transient process y (t). The control action calculator 17 compares this frequency with the network frequency f ₀ = 1 / T ₀ and, based on the obtained difference, calculates the value of the control action that must be exerted on the GDR 6 in order to approximate the natural frequency of the KNPS to the network frequency by a given value. This influence is transmitted through the interface 3 and the interface unit 5 to the inductance regulator 10 of the DGR 6.

The described working cycle is performed with the required frequency or at arbitrary moments of time at the command of the control program of the computing module 1, ensuring the timely adjustment of the GDR.

The use of the compensation signal delay for an integer number of periods provides, in contrast to the prototype, the operability of the device with asymmetry of the positive and negative HCH half-waves.

вместо задержанного исходного сигнала x(t) позволяет снизить уровень шума в выделенной затухающей свободной составляющей переходного процесса y(t), обеспечивая тем самым возможность анализа этой составляющей в области малых значений, что, в свою очередь, повышает точность измерения частоты свободных колебаний за счет увеличения интервала наблюдения. Более точное измерение указанной частоты позволяет повысить точность резонансной настройки ДГР, обеспечивая более полную компенсацию емкостных токов.Using a Synthesized Compensation Signal

instead of the delayed initial signal x (t), it allows to reduce the noise level in the selected decaying free component of the transient process y (t), thereby providing the possibility of analyzing this component in the region of small values, which, in turn, increases the accuracy of measuring the frequency of free oscillations due to increasing the observation interval. A more accurate measurement of the specified frequency makes it possible to increase the accuracy of the resonant tuning of the GDR, providing a more complete compensation of capacitive currents.

The noise reduction when using the averaged compensation signal is explained as follows. The NSN signal x (t) arriving at the ADC input before the excitation of free oscillations in the CNPC can be represented as

x (t) = x _d (t) + Δ (t),

where x _d (t) is the deterministic component of the power frequency corresponding to the NCH, from the moment of time t ₁ ;

Δ (t) is a random component corresponding to noise and interference (mathematical expectation M [Δ (t)] = 0).

The function of the transient process q (t) after the excitation of free oscillations can be written as

q (t) = x ' _d (t) + * A (t) + y _d (t),

where x ' _d (t) is the deterministic component of the power frequency,

the corresponding HCH, from time t ₂ ;

y _d (t) is a deterministic component corresponding to free oscillations in the CNPC.

Since the NSN does not change significantly during the execution of the working cycle, we replace x ' _d (t) by x _d (t).

In the case of excluding the component x (t) from the function q (t) by simply subtracting the signal x (t) delayed by an integer number of periods, as is done in the prototype, the resulting function y ₁ (t) will be equal to

. Тогда дисперсия D₁ результирующей помехи Δ₁(t) будет равна 2D[Δ(t)], а ее СКО, определяющее случайную составляющую погрешности значений y₁(t):where Δ ₁ (t) is the result of the addition of the random components of the functions x (t) and q (t). Let the interference Δ (t) in the original signal have a variance D [Δ (t)] and, accordingly, the standard deviation

... Then the variance D _{1 of the} resulting interference Δ ₁ (t) will be equal to 2D [Δ (t)], and its standard deviation, which determines the random component of the error of the values y ₁ (t):

, то есть в

раз больше исходного. С целью уменьшения случайной составляющей результирующей погрешности в предлагаемом устройстве используется синтезированный компенсационный сигнал, полученный путем усреднения формы сигнала x(t) за несколько периодов. При этом

, that is, in

times larger than the original. In order to reduce the random component of the resulting error, the proposed device uses a synthesized compensation signal obtained by averaging the waveform x (t) over several periods. Wherein

,

,

- усредненная реализация функции x(t) за K периодов;Where

- averaged implementation of the function x (t) over K periods;

t - time, 0≤t <T ₀ ;

T ₀ is the period of the function x (t);

K is the number of averaged periods;

Δ ₂ (t) is a random component of the averaged signal.

, дисперсия случайной составляющей результата D3 составит D[Δ(t)]+D[Δ(t)]/K, а ее СКО

. Таким образом, усреднение формы сигнала x(t) за K периодов обеспечит уменьшение СКО случайной составляющей погрешности результата в

раз.The variance D _{2 of the} function Δ ₂ (t) will be equal to D [Δ (t)] / K. Then, when subtracting from q (t) the compensation signal synthesized from repeating fragments

, the variance of the random component of the result D3 will be D [Δ (t)] + D [Δ (t)] / K, and its standard deviation

... Thus, averaging the waveform x (t) over K periods will provide a decrease in the standard deviation of the random component of the error of the result in

time.

The typical length of the observed free component of the transient process in the KNPS is 5-6 periods of power frequency, therefore, in comparison with the prototype, the random error will be reduced by about 1.3 times. In contrast to the prototype, with stable signal parameters x (t), the number of averaged fragments K can be increased, providing a further decrease in the random error and, consequently, a further increase in the accuracy of the resonance tuning of the GDR.

Thus, due to the combination of the operation of the described blocks and their connections, the operability of the device is achieved in a wider range of characteristics of the NSN, as well as an increase in the accuracy of the resonant tuning of the GDR.

Claims (1)

An arc suppression reactor control device containing a differential amplifier, a computing module equipped with an analog-to-digital converter designed to digitize the neutral bias voltage, which includes an arc suppression reactor, and an oscillation exciter and an interface unit with an arc suppression reactor inductance regulator connected to the interface of the computing module. that a memory unit, a period determination and averaging unit, a cyclic playback unit, an adder, a frequency determination unit and a control action calculation unit are introduced, and the output of the differential amplifier is connected to the input of an analog-to-digital converter, the output of which is connected to the input of the memory unit and the direct input of the adder , the output of the storage unit is connected to the input of the period determination and averaging unit, the output of the period determination and averaging unit is connected to the input of the cyclic playback unit, the output of the cyclic playback unit is connected is connected to the subtractive input of the adder, the adder output is connected to the input of the frequency determination unit, the output of the frequency determination unit is connected to the input of the control action calculation unit, the output of the control action calculation unit is connected to the interface input of the computing module.

Images