Data processing: measuring – calibrating – or testing – Measurement system in a specific environment – Electrical signal parameter measurement system
Reexamination Certificate
2001-06-18
2003-01-21
Hoff, Marc S. (Department: 2857)
Data processing: measuring, calibrating, or testing
Measurement system in a specific environment
Electrical signal parameter measurement system
C714S004110
Reexamination Certificate
active
06510393
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention generally relates to methods for detection of faults which occur in an electrical unit, such as a switching device. In particular, the present invention relates to methods for detection of transient faults which cannot be verified.
2. Brief Discussion of the Related Art
The method of the present invention detects faults, in particular transient faults which cannot be verified, which occur in a switching device. The method uses a fault counter which, when a fault occurs, counts by a specific first counting value in a first counting direction. At specific times in each case, the fault counter counts by a second counting value which is greater or less than the first counting value, or is equal to the first counting value, in a second counting direction, which is the opposite of the first counting direction, in situations where the relevant unit has operated without any faults within a predetermined time interval. An alarm signal is triggered when the relevant fault counter reaches or exceeds a defined counting threshold value in the one counting direction. The fault counter of the present invention is sometimes referred to as a leaky bucket counter.
What is meant by transient faults are faults which cannot be verified, that is to say, faults which occur sporadically and cannot be associated with specific causes.
In one practical embodiment, the occurrence of such faults leads to counting by two in each case in the first counting direction, and counting by one in each case in the second counting direction at the defined times. This means that faults which occur only sporadically do not lead to the immediate triggering of an alarm, while those faults which occur frequently do result in the relevant alarm being triggered. If, on the other hand, the relevant faults no longer occur, the fault counter will count in said second counting direction until a specific counter position is reached, at which no further actions need be taken any longer.
It is thus evident that an alarm signal is triggered only on reaching or exceeding the defined counting threshold value in said first counting direction, and this then leads to fault localization and fault rectification measures, although this is sometimes regarded as being too late.
A method for detection of faults which occur in at least one electrical unit, in particular a switching device, is given in WO 93/20527, in which the procedure described above is used. In this case, there are three monitoring parameters which play a major role, namely a disturbance step, the level of a lower limit (floor) and the level of an upper limit (ceiling). Said disturbance step in this case represents that variable by which an existing fault counter counts up in steps. No actions are taken until the fault counter reaches the upper limit value (ceiling C).
Furthermore, a method and a circuit arrangement for determination of critical faults is given in DE 43 02 908, which can occur in a system such as a communications system. For this purpose, fault signals are sorted on the basis of predetermined criteria, with each fault event being assigned a threshold value (number of fault events per unit time) associated with a time interval, such that a fault is classified and signaled as being critical when this threshold value is exceeded. If the number of fault events occurring within a time interval is less than the threshold value, the value which is stored in the event counter at the start of the next time interval is defined taking account of the number of fault events determined during the most recent time interval. This means that serious faults whose frequency is below the threshold value are detected.
An indirectly controlled switching system is discussed in DE 28 03 002, in which devices which use plausibility checks are used for identification of faults, and in which sporadically occurring faults are signaled when an adjustable threshold is exceeded.
GB 1 502 415 discusses quality monitoring relating to connection attempts to be carried out in a telephone switching device. Two counters are provided for this purpose, and are used to count connection attempts. One counter, which is referred to as the one's counter, in this case counts up to 100 and emits a counting pulse to the other counter once it has counted 100 connection attempts, with this other counter having ten counting stages and being referred to as the hundred's counter. These two counters can thus be used to count 1000 connection attempts. Furthermore, two fault counters are provided, using which faulty connection attempts are counted. The one fault counter, which is referred to as the mean-value fault counter, can count, for example, up to 25. This fault counter is reset to zero when the hundred's counter reaches its zero count position, that is to say when 1000 connection attempts have been counted. The other fault counter, which is referred to as the peak-value fault counter, can count, for example, up to 40 and is reset to zero when the one's counter reaches its zero count position, that is to say when 100 connection attempts have been counted. The mean-value fault counter thus emits an output signal if 2.5% of the connection attempts were faulty, and the peak-value fault counter emits an output signal if 40% of the connection attempts were faulty. In the former case, only an early warning about the existence of faulty connection attempts is emitted, which do not require any urgent countermeasures. However, in the second case, a serious warning is emitted, which requires counter-measures to be taken urgently.
This known prior art provides for the introduction of two warnings, but there is no provision whatsoever for the telephone switching device to be switched off upon reaching a specific fault threshold.
In contrast, the present invention provides a way in which a certain amount of early warning can be obtained firstly, in a relatively simple manner and in the course of detection of faults using a method of the type mentioned initially, without the respectively faulty unit, or the device containing it, having to be switched off immediately, and, secondly, how the time interval can be lengthened in a relatively simple manner once such an early warning has been emitted, during which time interval, if required, faults which occur can be detected once again.
According to the present invention, a first alarm signal is triggered on reaching or exceeding a further counting threshold value, which is below said defined counting threshold value and which signals only the occurrence of faults in the relevant unit, in that a second alarm signal is triggered, in conjunction with the respective unit being switched off, only on reaching or exceeding the one said counting threshold value which is higher than the former, and in that on reaching only said further counting threshold value and the faults subsequently not being evident, or if the count falls below the relevant further counting threshold value due to faults not being evident, the time interval between any two of said times at which said fault counter counts in the second counting direction is lengthened.
The present invention offers the advantage that the detection of faults which occur in an electrical unit, in particular a switching device, by the fault counter can be used in a particularly simple manner to provide an early warning, so that this makes it possible to monitor the relevant unit and, if necessary, to take particular actions at an early stage. Therefore, it is possible to identify the relevant faults and/or to identify the reasons for the occurrence of the relevant faults. This is particularly advantageous in the case of transient faults which cannot be verified and which occur only sporadically, and which cannot be identified by the conventional fault testing routines which are normally provided in electrical units. Relevant observation can be extended over a relatively long active period of the fault counter for detecting any faults wh
Fish & Richardson P.C.
Hoff Marc S.
Raymond Edward
Siemens Atkiengesellschaft
LandOfFree
Method for detecting faults in a switching device 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 detecting faults in a switching device, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method for detecting faults in a switching device will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3000745