Electricity: measuring and testing – Fault detecting in electric circuits and of electric components – For fault location
Reexamination Certificate
2001-07-31
2002-11-19
Le, N. (Department: 2858)
Electricity: measuring and testing
Fault detecting in electric circuits and of electric components
For fault location
C324S535000, C324S537000, C324S122000
Reexamination Certificate
active
06483316
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to a method of diagnosing a partial discharge in a gas-insulated apparatus and a partial discharge diagnosing system for carrying out the same. More particularly, the present invention relates to a method of diagnosing a partial discharge in a gas-insulated apparatus by detecting a partial discharge signal, and subjecting the partial discharge signal to multiple stages of conditional branching for diagnosis, and a partial discharge diagnosing system for carrying out the method.
2. Description of the Related Art
A gas-insulted apparatus, such as a gas-insulated switchgear installed in a substation, produces a partial discharge when an abnormal insulation condition arises in the apparatus. The partial discharge generates electromagnetic waves therewith. The partial discharge, which can be regarded as a precursory indication of an accidental electric breakdown, must accurately be detected and diagnosis must be made on the partial discharge to decide whether or not there is any abnormal insulation condition to prevent the accidental electric breakdown of the gas-insulated apparatus.
Causes of abnormal insulation include the entrance of metallic foreign matters into the gas-insulated apparatus, defects in insulators, such as voids or cracks in spacers, faulty contact in conductors of shielding members and or the like.
If the partial discharge due to those causes of abnormal insulation progresses, the insulator is progressively deteriorated and there is the possibility that electric breakdown will occur when an overvoltage, such as a lightning, is applied to the gas-insulated apparatus. Therefore, a part where the partial discharge occurred, the level of the partial discharge must accurately be determined and the part where the partial discharge occurred must be inspected and repaired depending on the part and the level of the partial discharge occurred.
Detection of the partial discharge can be performed through the detection of an electromagnetic wave, a high-frequency current, a vibration or a sound that is generated along with the partial discharge, or through the analysis of a gas that is produced by the partial discharge.
A partial discharge that is produced in SF
6
gas contained in the gas-insulated apparatus exhibits a very sharp pulse discharge having a pulse width in the range of several nanoseconds to several tens nanoseconds. The electromagnetic wave generated along with the partial discharge has a frequency in a wide frequency band up to several gigahertz.
The gas-insulated apparatus is constructed by sealing components in a metal container and, similar to a waveguide, propagates electromagnetic waves efficiently. Since the components of the gas-insulated apparatus are contained in the metal container, it is relatively difficult for external electromagnetic noise to reach the component parts. Accordingly, electromagnetic waves generated along with the partial discharge can be detected in a high sensitivity by disposing an internal antenna in the metal container. A previously proposed method of finding a part where a partial discharge occurred receives electromagnetic waves of frequencies with a wide frequency band of several megahertz to several gigahertz by an internal antenna and analyzes the frequency of the electromagnetic waves by a spectrum analyzer or the like.
Another previously proposed diagnostic method receives an electromagnetic wave of a specific frequency that is not affected by external noise or electromagnetic waves of frequencies with a narrow frequency band, instead of receiving all the electromagnetic waves of frequencies in a wide frequency band.
As mentioned in JP-B No. Hei 7-50147, there have been proposed a method that analyzes detected electromagnetic waves for frequency and decides whether or not any abnormal condition has occurred on the basis of the levels of frequency of the electromagnetic waves, and a method that decides whether or not any abnormal condition has occurred on the basis of feature values computed from the patterns of waveforms. A known deciding means achieves pattern recognition by using a neural network or a fuzzy system.
For example, in a system employing a neural network, the neural network comprises an input layer, an intermediate layer and an output layer, a plurality of frequency data are given to the input layer, and diagnostic results are provided on the output layer.
It is difficult, however, for the known diagnostic method which detects electromagnetic waves generated by a partial discharge and diagnosis an abnormal insulating condition in the gas-insulated apparatus to accurately determine the occurrence of the partial discharge and the causes of the partial discharge.
The known method of detecting a partial discharge signal decides that a partial discharge has occurred when a signal of a level not lower than a predetermined level, having a frequency in a reference frequency band for decision is generated. It is difficult for this method to discriminate clearly a partial discharge signal from an external noise signal. If the external noise signal is produced steadily, such as a broadcast wave, the external noise signal is measured beforehand, and signals of frequencies outside a frequency band including the frequency of the external noise signal are measured or the external noise signal is removed by a noise signal removing process. However, if the external noise signal is generated unsteadily, such as a communication signal transmitted by a mobile telephone, noise from devices or an aerial corona noise, it is difficult to discriminate a partial discharge signal generated by an internal partial discharge from the external noise.
When the partial discharge signal is detected by an antenna disposed in the gas-insulated apparatus, the frequency characteristic of the partial discharge signal detected by the antenna is affected by the frequency characteristic of a sensor including the antenna, and the frequency characteristic of an electromagnetic wave propagating passage determined by the shape and construction of a tank included in the gas-insulated apparatus. Consequently, it is difficult to accurately determine whether the signal detected by the antenna is a partial discharge signal generated internally or externally, and specialist's diagnosis is necessary for the decision.
When the diagnostic method using a neural network or a fuzzy system for a general pattern recognizing method as a diagnostic algorithm, a mass of data is obtained from the detection signal, an excessively long learning time is required and learning does not converge. Besides, it is difficult to diagnose the cause of the partial discharge and to find a part where the partial discharge was produced, even if the learning converges because there are many types of partial discharge.
The neural network structure employed in the known back propagation method makes diagnosis immediately upon the reception of analytic data and provides results of analysis. However, since the input layer
41
and the output layer
43
each has a large number of nodes, making the internal configuration of the neural network structure complicated, there rise problems such that the convergence of the learning of known data is difficult, local minimum is liable to occur and an enormously long learning time is necessary. Furthermore, it is difficult to achieve diagnosis with a practically acceptable accuracy.
The present invention has been made in view of the foregoing problems and it is therefore an object of the present invention to provide a diagnostic system capable of detecting a partial discharge signal acting as a precursory signal indicating an accidental electric breakdown in a gas-insulated apparatus and of achieving accurate diagnosis through multiple conditional branching of the partial discharge signal, and a partial discharge diagnosing method to be carried out therein.
SUMMARY OF THE INVENTION
The present invention employs the following means to solve the foregoing problems
Endo Fumihiro
Kato Tatsuro
Hamdan Wasseem H.
Le N.
Mattingly Stanger & Malur, P.C.
LandOfFree
Method of diagnosing partial discharge in gas-insulated... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method of diagnosing partial discharge in gas-insulated..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method of diagnosing partial discharge in gas-insulated... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2986218