Method for generating a fault indication signal

Communications: electrical – Condition responsive indicating system – Specific condition

Patent

Rate now

  [ 0.00 ] – not rated yet Voters 0   Comments 0

Details

340635, 361115, 324521, 395 22, G08B 2100

Patent

active

058545905

DESCRIPTION:

BRIEF SUMMARY
The invention concerns a method for generating a fault indication signal in the event of a fault in an electric power supply network being monitored, through a multilayer neural network arrangement which is taught, while being coached during simulated permanent and transient faults, so that it emits an output signal having a pre-defined value for permanent faults and an output signal having another pre-defined value for transient faults.
A prior art method of this type is described in Proceedings No. 368 of the Fifth International Conference on Developments in Power System Protection, York (Great Britain), Mar. 30/Apr. 1st, 1993. This prior art method is based on the recognition that, when a power circuit breaker is switched following a fault detected in the respective power supply network, signals with different frequency spectra can be detected. The different frequency spectra can be assigned to different faults in the power supply network being monitored, so that distinction can be made between the different faults regarding whether they are permanent or transient faults between a phase of the power supply network and the ground. To distinguish between permanent and transient faults between a phase and the ground, the signal that can be detected as the circuit breaker is switched is subjected to a fast Fourier transformation, and subsequently the digital values obtained are supplied to a neural network arrangement. The neural network arrangement is previously taught by simulation, using Rumelhart's backpropagation algorithm, in a suitable manner and, when used for generating a fault recognition signal, responds to a fault in a power supply network by producing a signal at its output whose magnitude allows one to distinguish between a permanent and a transient fault. An output signal with a pre-defined value would then signify a permanent fault, while an output signal with another transient value would signify a transient fault. The distinction between permanent and transient faults is important, because in the case of a transient fault the power company can reclose the circuit breaker after a certain waiting period, so that power is subsequently restored.
The prior art method can only be used for distinguishing between permanent and transient faults when these faults are faults occurring between a phase of the power supply network being monitored and the ground. The prior art method also requires a relatively long time for a fault indication signal to be generated, since the signals that can be detected as the circuit breaker switches must first be subjected to a fast Fourier transformation, which must be carried out during at least one cycle of the power supply network's voltage. The object of the invention is to provide a method for generating a fault indication signal allowing generation of a fault indication signal in a relatively short time even for very different cases of fault in the power supply network being monitored.
To achieve this object, in a process of the aforementioned type according to the invention, a neural network arrangement having a neural network for each phase of the power supply network being monitored is used for distinguishing between a short circuit due to arcing and a metal contact short-circuit, and each neural network has a behavior taught through simulated voltages of each phase, while being coached during arcing and metal contact short-circuits, so that an output signal with a pre-defined value appears in the case of a short circuit due to arcing and an output signal with another pre-defined value appears in the case of metal contact short-circuit; at the same time, sequentially sampled normalized values of the respective phase voltages of the power supply network are applied to different neurons of the input layer of each neural network.
An important advantage of the method of the invention is the fact that it allows a fault indication signal, whose magnitude allows distinction to be made between a metal contact short-circuit and a short circuit due to arcing, to be ge

REFERENCES:
patent: 5625751 (1997-04-01), Brandwajn et al.

LandOfFree

Say what you really think

Search LandOfFree.com for the USA inventors and patents. Rate them and share your experience with other people.

Rating

Method for generating a fault indication signal 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 for generating a fault indication signal, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method for generating a fault indication signal will most certainly appreciate the feedback.

Rate now

     

Profile ID: LFUS-PAI-O-1427333

  Search
All data on this website is collected from public sources. Our data reflects the most accurate information available at the time of publication.