阅读说明：本技术 用于确定用于电力系统的绝缘的介电质等效电路图的元件的方法和设备 (Method and apparatus for determining elements of a dielectric equivalent circuit diagram for insulation of an electrical power system ) 是由 朱利安·赖茨 埃克哈德·布洛克曼 于 2020-02-10 设计创作，主要内容包括：本发明涉及用于确定用于电力系统的绝缘的介电质等效电路图的元件的方法和测试装置以及绝缘监控设备。对于测量周期T<Sub>m</Sub>的持续时间存储系统的阶跃响应的所有数据点,并计算故障电阻R<Sub>f</Sub>和泄露电容C<Sub>e</Sub>的初始值C<Sub>e0</Sub>。在故障电阻R<Sub>f</Sub>和初始值C<Sub>e0</Sub>的此分析性确定后,通过使用近似算法的数值信号处理确定吸收元件R<Sub>a</Sub>和C<Sub>a</Sub>以及泄露电容C<Sub>e</Sub>的成分,近似算法连续地模拟记录的阶跃响应。为了模拟阶跃响应,分析地形成通过具有等效电路图元件R<Sub>f</Sub>,C<Sub>e</Sub>,R<Sub>a</Sub>,C<Sub>a</Sub>和测量电阻R<Sub>m</Sub>的等效电路图模型化的电力系统的传递函数G(s)并计算输出信号,该输出信号在激活电力系统时产生,且通过阶跃函数使用传递函数G(s)描述该输出信号。(The invention relates to a method and a test device for determining elements of a dielectric equivalent circuit diagram for the insulation of an electrical power system, and an insulation monitoring device. For a measurement period T m Stores all data points of the step response of the system and calculates the fault resistance R f And leakage capacitance C e Initial value of (C) e0 . At fault resistance R f And an initial value C e0 After this analytical determination, the absorption element R is determined by means of numerical signal processing using an approximation algorithm a And C a And a leakage capacitance C e The approximation algorithm continuously simulates the recorded step response. For simulating the step response, the element R is analytically formed by having an equivalent circuit diagram f ，C e ，R a ，C a And measuring the resistance R m And calculating an output signal that is generated when the power system is activated and that is described by a step function using the transfer function g(s).)

1. A method for determining elements of a dielectric equivalent circuit diagram for insulation of an electrical power system, the equivalent circuit diagram being formed by a fault resistance (R)_f) Leakage capacitance (C)_e) And a parallel connection of absorbing elements consisting of absorbing resistors (R)_a) And absorption capacitance (C)_a) The method comprising the method steps of:

applying a voltage step to a test voltage (u) of an electric power system as a voltage source₀(t)) the power system has a known measuring resistance (R) in series with the voltage source_m)；

Measured at the measuring resistance (R)_m) The voltage drop at (a) is taken as the step response (u) of the power system_a(t))；

Characterized in that the method further comprises:

during the measuring period (T)_m) Internally recorded step response (u)_a(T)), measuring the period (T)_m) Corresponds approximately to the step response (u)_a(t)) twice the settling stage;

step response from being in a settled state (u)_a(t)) calculating the fault resistance (R)_f)；

From step response (u) in the sedimentation phase_a(t)) determining a time constant (τ);

from time constant (τ), fault resistance (R)_f) And measuring the resistance (R)_m) Calculating leakage capacitance (C)_e) Initial value of (C)_e0)；

Forming by using equivalent circuit diagram elements (R)_f，C_e，R_a，C_a) And measuring the resistance (R)_m) Equivalent circuit diagram mode ofTransfer function (G (s)) of a modeled power system as a measured resistance (R)_m) Of (U)'_a(s)) and a test signal (U) of a voltage source as an input signal₀(s)) in a ratio;

used as test signal U₀Step function of(s) is simultaneously incorporated into the calculated fault resistance (R)_f) Leakage capacitance (C)_e) Is calculated as an initial value (C)_e0) And a known measurement resistance (R)_m) Calculating an output signal (U ') from a transfer function G(s)'_a(s))，

To minimize the calculated output signal (u ') converted in the time range'_a(t)) and the recorded step function of the measurement (u)_a(t)) via an approximation (R 'by an approximation algorithm'_a，C’_a，C’_e) Iteratively determining the absorption resistance (R)_a) And an absorption capacitor (C)_a) And leakage capacitance (C)_e) Each of (a).

2. Method according to claim 1, characterized in that the approximation algorithm is between four and five times the time constant (τ) and the measurement period (T)_m) Is minimized in the time section between the metering ends of the settling phase at approximation 3/5.

3. A method according to claim 1 or 2, characterized in that the approximation algorithm is performed according to the least squares method.

4. A method according to any of claims 1-3, characterized in that the application for determining dielectric properties of insulation in an ungrounded power supply system is combined with insulation monitoring.

5. A test apparatus for determining elements of a dielectric equivalent circuit diagram for insulation of an electrical power system, comprising:

signal processing device configured for performing the method according to the invention for determining elements of a dielectric equivalent circuit diagram for insulation of an electrical power system according to any of claims 1-4.

6. An insulation monitoring device for identifying an insulation resistance of an ungrounded power supply system, comprising:

signal processing device configured for carrying out the method according to the invention for determining elements of a dielectric equivalent circuit diagram for an insulated, ungrounded power supply system according to any of claims 1 to 4.

7. The insulation monitoring device of claim 6, further comprising:

for test voltages (u) with variable amplitude₀(t)) variable measuring resistance (R) for low noise detection_m) And/or a variable coupling impedance.

Technical Field

The invention relates to a method and an apparatus for determining elements of a dielectric equivalent circuit diagram for insulation of an electrical power system, the equivalent circuit diagram being a parallel connection consisting of a fault resistance, a leakage capacitance and an absorption element consisting of a series connection of an absorption resistance and an absorption capacitance.

Background

The power supply system has conductive elements that need to be separated from each other in some cases via an insulating material. In particular, when installing an ungrounded power supply system, it is necessary to monitor the insulation resistance. The complex insulation resistance (insulation impedance) is modeled by an equivalent circuit diagram of the power supply conductor, which consists of a parallel connection of the leakage capacitance and the ohmically measured fault resistance (real part of the insulation impedance). For monitoring the fault resistance, an insulation monitoring device is more preferably used. Measuring insulation resistance is particularly challenging in electrical devices with very large insulation levels greater than 500M Ω. This includes an undersea cable or a high voltage battery (HV battery).

Mainstream insulation monitoring devices that have been optimized for high impedance measurement ranges but are based on existing analytical techniques have drawbacks with respect to measurement accuracy. Equivalent circuit diagrams that are not sufficient for high impedance applications and do not take into account dielectric absorption by the insulating material are cited as sources of error.

IEEE Standard 43-2000 shows an equivalent circuit diagram for isolation, with the addition of parallel-switched RC-row elements and snubber resistor R in addition to the parallel connection of the leakage capacitors previously described_aAnd an absorption capacitor C_aCompositional absorbing elements have been added to the equivalent circuit diagram, which quantify the dielectric absorption behavior.

The dielectric properties of the insulation change over time and the environment and serve as an indication of premature identification of progressively deteriorating insulation. In addition to the hitherto known special measurement of the ohmic measured fault resistance, conventional quantification of the dielectric can be used in order to be able to detect the signs of aged insulation more quickly and to be able to evaluate their electrical state. Due to this, maintenance of the power system can be planned faster.

There is no technical solution to directly quantify the equivalent circuit diagram by dielectric observation of the insulation in the power system, especially in the active non-grounded power supply system.

In an ungrounded power supply system, the dielectric properties of the insulating material are currently disregarded, monitoring only the effects of insulation deterioration due to continuously measuring the fault resistance. In a grounded network, periodic measurements of the polarization index are performed more significantly, however, disadvantageously, the network needs to be shut down. The polarization index is artificially large and provides information about the dielectric properties of the insulation, however it is not quantified.

Furthermore, methods of impedance spectroscopy are known, however, the methods are particularly constrained to laboratory testing in the field of material science and operate in the frequency domain. Furthermore, this method is not suitable for the extremely high impedance range and for active power systems.

Disclosure of Invention

The present object of the invention is therefore to propose a method and an arrangement for determining elements of an insulated dielectric equivalent circuit diagram for an electrical power system, which in particular takes into account an ungrounded power supply system during operation.

This object is achieved by a method for determining elements of a dielectric equivalent circuit diagram for insulation of an electrical power system. The equivalent circuit diagram is composed of a fault resistor R_fLeakage capacitor C_eAnd a parallel connection of absorbing elements consisting of an absorbing resistor R_aAnd an absorption capacitor C_aIs connected in series. The method comprises the following steps: applying a voltage step to a test voltage u as a voltage source of an electrical power system₀(t) the power system has a known measuring resistance R in series with the voltage source_m(ii) a Measuring at measuring resistance R_mThe voltage drop at (b) is taken as the step response u of the power system_a(t); during a measurement period T_mInternally recorded step response u_a(T), measuring period T_mCorresponds approximately to the step response u_a(t) twice the settling stage; from step response u in the settled state_a(t) calculating the Fault resistance R_f(ii) a From step response u in the sedimentation phase_a(t) determining the time constant τ; fromTime constant τ, fault resistance R_fAnd measuring the resistance R_mCalculating leakage capacitance C_eInitial value of (C)_e0(ii) a Forming by using an equivalent circuit pattern element R_f，C_e，R_a，C_aAnd measuring the resistance R_mAs a measured resistance R_mOutput signal U 'of'_a(s) and a test signal U as a voltage source for the input signal₀(s) ratio between; used as test signal U₀Step function of(s) is simultaneously incorporated into the calculated fault resistance R_fLeakage capacitor C_eIs calculated as an initial value C_e0And a known measurement resistance R_mCalculating an output signal U 'from a transfer function G(s)'_a(s); to minimize the calculated output signal u 'converted in the time range'_a(t) step function u with recorded measurements_a(t) by way of a deviation between the approximation algorithms via an approximation R'_a，C’_a，C’_eIteratively determining the absorption resistance R_aAnd an absorption capacitor C_aAnd leakage capacitance C_eEach of (a).

Test current u as a voltage source in applying a voltage step to an electrical power system₀After (t) (the power system has a measuring resistor R connected in series to the voltage source_m) In the measuring period T_mInternally measuring and recording the step response u_a(T), measuring period T_mIs approximately of step response u_a(t) twice the duration of the settling phase. In contrast to the methods known from the prior art, in which the measurement is interrupted as soon as the steady state is reached, with regard to a rapid detection of the insulation resistance, the measurement and recording takes place over a relatively long measurement period T_mAnd (4) the following steps.

Step response u_a(T) all data points during the measurement period T_mIs stored and subsequently subjected to digital signal processing. First, the average value is determined from the measured value in the measurement period T_mHalf of and measurement period T_mStep response u of the entire settlement between the ends of_a(t) analytically identifying the fault resistance R_f。

From step response u_a(t) the time constant τ is determined at the progression of the settling phase, the end of the settling phase corresponding approximately to five times the time constant τ. From time constant τ, fault resistance R_fAnd measuring the resistance R_mCalculating leakage capacitance C_eInitial value of (C)_e0。

At fault resistance R_fAnd leakage capacitance C_eInitial value of (C)_e0After this analytical determination of (2), the absorption element R is determined by means of numerical signal processing_aAnd C_aAnd a leakage capacitance C_eThe composition of (1).

For this purpose, an approximation algorithm is used, using a known measuring resistor R_mAnd the equivalent circuit element fault resistance R which is analyzed and identified_fAnd leakage capacitance C_eInitial value of (C)_e0The approximation algorithm continuously simulates the actual measurement signal, i.e. the recorded step response u_a(t) of (d). Absorption resistance R_aAnd absorption capacitor C_aForms an approximation R 'learned from the value domain'_a，C’_aAnd C'_e。

To simulate a step response u_a(t) analytically forming the element R by having an equivalent circuit diagram_f，C_e，R_a，C_aAnd measuring the resistance R_mThe transfer function G(s) of the electric power system modeled by the equivalent circuit diagram of (1), wherein the transfer function G(s) represents the measured resistance R_mOutput signal U 'of'_a(s) and a test signal U as a voltage source for the input signal₀Ratio of(s):

G(s))＝U’_a(s))/U₀(s)

the dependency on the variable s describes the laplace transform (written in uppercase letters) for a time range quantity (written in lowercase letters). The initial quantities represent iteratively variable quantities in the sense of an approximate solution.

Computing an output signal U 'by a transfer function G(s)'_a(s)＝U₀(s) G(s), and the output signal is generated when the power system is activated, and passes as the test signal U₀(s) describing the output signal via a transfer function g(s). At the incorporation of calculated fault resistance R_fLeakage capacitor C_eIs calculated as an initial value C_e0And a known measuring resistance R_mIs calculated to minimize the time range (u'_a(t)) converted calculated output signal U'_a(s) and measured recorded step response u_a(t) by way of a deviation between the approximation algorithms via an approximation R'_a，C’_aAnd C'_eIteratively determining the leakage resistance R_aLeakage capacitor C_aAnd leakage capacitance C_eEach of (a).

If the deviation falls below a certain threshold, the approximation process may be terminated and element R of the equivalent circuit diagram may be replaced_a，C_a，C_eCan be considered via approximation R'_a，C’_aAnd C'_eDetermined sufficiently accurately.

The approximation algorithm advantageously multiplies four and five times the time constant τ by the measurement period T_mIs minimized in the time section between the metering ends of the settling phase at approximation 3/5.

The approximation algorithm focuses on the signal section between the point in time at which the time constant τ is approximately four to five times reached and the end of the metering of the sedimentation process, however an overlap of the sections is explicitly possible.

To receive a faster approximation of the step response recorded via the data points, the approximation algorithm is limited to minimizing the deviation and masking the step response u for the middle segment of the signal progression_a(t) start and end sections. Depending on this application necessity, this intermediate section may be reduced or lengthened, for example, to approximate the section by reducing the "focus".

In another embodiment, the approximation algorithm is performed according to a least squares method.

Drawings

FIG. 1 shows a dielectric equivalent circuit for insulation;

FIG. 2 by measuring the resistance R_m(step response u)_a(t)) shows the progression of the voltage;

FIG. 3 shows the steps of numerical signal processing by an approximation algorithm;

FIG. 4 shows a test apparatus according to the invention in an ungrounded power supply system; and

fig. 5 shows an insulation monitoring device with a signal processing device according to the invention.

Detailed Description

A dielectric equivalent circuit diagram for insulation in a test circuit is shown in fig. 1. Between a conductor L and a ground PE of, for example, an ungrounded power supply system, a fault resistance R_fAnd leakage capacitance C_eIs effective, fault resistance R_fAnd leakage capacitance C_eTogether forming a complex-valued insulation resistance (insulation resistance) of the non-grounded power supply system. To describe the conductor L and the groundDielectric absorption behavior of insulation between PEs by means of a resistor R switched in parallel with the insulation resistance_aAnd an absorption capacitor C_aThe series connection of (a) and (b) enhances the equivalent circuit diagram.

To determine the equivalent circuit diagram quantity R_f，C_e，R_aAnd C_aThe dielectric equivalent circuit diagram is integrated with a test voltage u₀(t) and measuring the resistance R_mIn the test circuit in which the step function is applied as the test voltage u₀(t) in the case of measuring the resistance R_mTo measure the step response u_a(t)。

In FIG. 2, the resistance R is measured_mThe progression of the voltage at (a). When applying a step function as the test voltage u₀(t) voltage progression and step response u_a(t) are the same. During a measurement period T_mInternally recorded step response u_a(T), measuring period T_mMay be divided into three partial sections A, B and C which in turn may overlap.

The time segment a extends from a time t equal to 0 to approximately a time 5 τ, at which the voltage amplitude has a step function height value U₀τ forms the time constant that indexes the voltage progression. The second time segment B extends approximately from four to five times the time constant to the measuring period T_mApproximately 3/5 of. The third time segment C comprises a slave measurement period T_mTo approximately half of the measurement period T_mThe duration of the end of (d). From representing a sinking state and in which the capacitance C is leaked_eAnd an absorption capacitor C_aThe time segment C, which can be considered as an open circuit, is measured as the resistance R_mWhen known, the fault resistance R is preferably calculated by determining an average value_f. Assuming that the settling process is nearly terminated at five times the time constant τ, the time constant τ is thus identified and the resistance R is measured using a pair of measuring resistors_mAnd a previously determined fault resistance R_fRecognition of leakage of capacitance C_eDetermining an initial value C_e0。

Calculating the fault resistance R analytically from the time range C_fAnd analytically determines the leakage capacitance C from the time segment A_eOfInitial value C_e0Then, in step response U_aThe absorption resistance R is determined numerically in the middle time section B of (t)_aAnd an absorption capacitor C_aAnd more accurately determine the leakage capacitance C via an approximation algorithm_e。

Fig. 3 shows a step of digital signal processing, which iteratively identifies the absorption resistance R in a numerical manner by means of an approximation method_aAnd an absorption capacitor C_aAnd a leakage capacitance C_eApproximation R 'of'_a，C’_a，C’_e。

Assuming an absorption resistance R_aWide range of values (1k Ω … 100G Ω), absorption capacitance C_aWide range of values (1pF … 100mF) and leakage capacitance C_e＝(0.8…1.2)*C_e0Initial values are first established for these three quantities. Naturally, empirical values from a particular device may also be resorted to.

By further incorporating measuring resistance R_mKnown value of (d) and identified fault resistance R_fThe transfer function g(s) is calculated. Transfer function multiplied by laplace transform step function U₀(s) (step height U)₀) And generates a signal output U'_a(s). The output signal U'_a(s) generating a step response u converted in the time domain and captured in reality according to a least squares method_a(t) compared output signal u'_a(t) of (d). By continuously varying the absorption resistance R_aAnd an absorption capacitor C_aAnd a leakage capacitance C_eApproximation R 'of'_a，C’_a，C’_eIteratively calculating a new transfer function G(s) until output signal u 'converted in the time range when compared'_a(t) and recorded step response u_a(t) the sum of the squared errors falls below a set threshold. Thus, the absorption resistance R is determined sufficiently accurately_aAnd an absorption capacitor C_aAnd a leakage capacitance C_eThe value of (c).

Fig. 4 shows a test device 10 according to the invention in an ungrounded power supply system 2. The test device 10 is connected between the active conductor L and the ground PE of the power supply system 2 and comprises a signal processing device 12 according to the invention configured for carrying out the method according to the invention.

In fig. 5, the insulation monitoring device 20 is shown in the ungrounded power supply system 2. The standardized insulation monitoring device 20 is connected between the active conductor L and the ground PE of the power supply system 2 and additionally comprises a signal processing device 12 according to the invention for carrying out the method according to the invention for determining elements of a dielectric equivalent circuit diagram for insulation.