JP2010213400A - Device for determining cause of grounding fault in power system in premises of customer - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To identify a cause of a grounding fault by discretizing grounding fault waveform signals and processing information. <P>SOLUTION: Sensors 2, 3 having such a frequency characteristic that a waveform is detected are applied to the input portion of a grounding fault cause determining device for accident waveforms. The device carries out signal processing using a computer based on amplitude, frequency characteristic, phase, and temporal characteristic information, and identifies a cause of a grounding fault in a short time. With respect to mathematical processing, time discrete signal transformation is carried out by Z transformation in addition to Fourier transformation. The signal is transformed into transaction data (form in which data processing is carried out) and cause determination processing is carried out based on a decision tree using a data mining technique. (The data mining technique is a technique in which a large amount of data is analyzed to find a correlation or a pattern). In ground-fault voltage and ground-fault current waveforms in an accident, a problem of aliasing is solved by using a sensor for detecting the frequency band of accident waveform. (Aliasing is a phenomenon in which any nonexistent waveform appears because a sampling theorem is not met). As a result, it is possible to determine a cause of a grounding fault in a short time and significantly enhance the accuracy of the determination. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to a device capable of determining a ground fault by adding a simple function to a conventional ground fault protection relay.

There are various causes of ground faults in the power system, but most of them are ground faults that occur on the customer's premises for mechanical crops. The total length of the power grid on the customer premises is generally long, but the usual way to identify the cause of the accident is to perform a visual inspection of the system and find the cause.
At the same time, depending on the scale of the ground fault, there is a possibility of a wide range of power outages. Electric power energy is a lifeline, and it is required to quickly recover from the accident. However, it took time to investigate a wide range of systems and identify the cause of the ground fault. Furthermore, if the contact is caused by a live part of a small animal, it may be difficult to identify the cause of the accident because the small animal that caused the accident may not stop at the location where the accident occurred.

In the customer premises, the power source is drawn from the power receiving location from the power company, branched from the power receiving point, and distributed on the premises, so that the ground fault voltage and ground fault current waveform at the power receiving point due to the cause of the ground fault Can be considered different in nature.
This is because, as proved by Ohm's law, the ground fault voltage and ground fault current at the time of the ground fault change due to the difference in impedance of the objects that make up the cause of the ground fault. It is well known that the ground fault current waveform (hereinafter, waveform) is not constant.

Currently, when a ground fault occurs, there is a device that cuts off the electric circuit only by the ground fault voltage or the amplitude of the ground fault current and prevents the influence of the fault to a minimum.
However, the cause has not been determined. Additional information processing technology is required to determine the cause of the accident. Furthermore, conventional voltage and current sensors in the commercial frequency band may be processed on the premise of an erroneous waveform due to waveform aliasing phenomenon (a phenomenon in which a nonexistent waveform appears because the sampling theorem is not satisfied).

Therefore, a sensor with a frequency characteristic that can detect the waveform is applied to the input part of the accident waveform, and signal processing is performed using a computer from the amplitude, frequency characteristic, phase, and time characteristic information to identify the cause of the ground fault. .
As mathematical processing, in addition to the conventional Fourier transform, time-discrete signal transform is performed by Z transform. The signal is replaced with transaction data (in a form that allows data processing), and cause determination processing is performed by a decision tree using a data mining technique (a technique for analyzing a large amount of data and searching for correlations and patterns).
In addition, in the ground fault voltage and ground fault current waveform at the time of an accident, the use of a sensor that detects the frequency band of the accident waveform can solve the aliasing problem and greatly improve the accuracy of determining the cause of the ground fault.

JP 4-127820A

  The problem to be solved is that the cause of the ground fault in the power system on the customer premises cannot be identified.

  In the present invention, the waveforms of the zero-phase voltage and the zero-phase current are converted into information according to the cause of the accident, and the cause of the ground fault can be determined using the decision tree by data mining.

If a ground fault occurs in the power system in which the ground fault cause determination device of the present invention is installed, the cause can be easily identified and the power supply can be restored in a short time.
In addition, by grasping the transition progress of the zero-phase voltage and zero-phase current, it also has an element to judge the sign of a ground fault in advance, which can contribute to large preventive maintenance.

It is a high voltage ground fault waveform acquisition circuit diagram. (Example 1) It is a block diagram of a signal processing flow. It is a low voltage ground fault waveform acquisition circuit diagram. (Example 2)

  This can be realized by changing the conventional commercial frequency band sensor that captures the amplitude of the zero-phase voltage and zero-phase current to a sensor with a frequency band that can reproduce the accident waveform, and adding information processing.

  The embodiment of FIG. 1 is a ground fault waveform capturing circuit diagram in a high voltage circuit. A current detection sensor 2 is arranged on the power system side 1, and a voltage detection sensor 4 is arranged on the customer building inner side 3.

  The signal input portion of the device of the present invention has a specification that can be realized by using a current detection sensor 2 having a frequency characteristic of about 20 kHz in addition to the generally used waveform acquisition circuit of FIG.

  FIG. 2 is a block diagram showing a signal processing flow. The accident waveform signal is processed as transaction data 7 via the signal processing unit 6 via the input part 5 and then processed as transaction data 7, and then collated by the data mining analysis 8 to obtain the output 9 of the cause of the ground fault.

That is, these accident waveform signals fetched from the input portion 5 are discretized and classified as transaction data 7 using a mathematical method such as Fourier transform. The cause of these data can be determined from the past data by a decision tree by using the data mining technique 8.

  The embodiment of FIG. 3 is a ground fault waveform capturing circuit diagram in a low voltage circuit. The current detection sensor 11 captures the current input from the power supply side 10 and outputs it to the demand equipment side 12 on the transformer secondary side.

  The current detection part has a specification that can be realized by changing the signal input part of the device of the present invention to the current detection sensor 11 having a frequency characteristic of about 20 kHz, as in the first embodiment.

  The identification of the cause of ground faults in the power system was extremely difficult as a method other than by visual confirmation, but by installing the device of the present invention, the cause can be determined with high accuracy and prevention in the power system on the customer premises. Contribute to conservation.

DESCRIPTION OF SYMBOLS 1 Electric power system side 2 Current detection sensor 3 Customer house inside 4 Voltage detection sensor 5 Accident waveform signal input part 6 Signal processing part 7 Transaction data 8 Data mining 9 Ground fault accident cause output 10 Power supply side 11 Current detection sensor 12 Demand equipment side

Claims (1)

  After changing the current input part of the ground fault detection circuit installed in the home electrical work to a current detection sensor with frequency characteristics that can reproduce the fault waveform, performing discrete acquisition processing, and performing Fourier transform A ground fault cause identification device that uses a decision tree to identify the cause of an accident by applying data mining techniques to transaction data (in a form that allows data processing).

Images