METHOD FOR DETERMINING EVENTS IN A NETWORK

Abstract

A method for determining events in a network of consumption meters, in which an event is determined by grouping data wherein the event according to consumption meters which have detected this characterizing data. If groups of consumption meters match, this is determined as an event, the grouping being carried out according to the following criteriatemporal coincidence of the eventspatial concordance of the eventconsistency of the event type. The spatial concordance is determined by determining and ranking the Euclidean distance of each of a plurality of consumption meters to everyone of the of others of the plurality of consumption meters, and by assigning the n closest consumption meters that have detected this characterizing data to a group of consumption meters.

Claims

1. A method for determining events in a network of consumption meters, the method comprising the steps of: determining an event by grouping data characterizing the event according to consumption meters which have detected this characterizing data and, if groups of consumption meters match, this is determined as an event, the grouping being carried out according to the following criteria temporal coincidence of the event spatial concordance of the event consistency of the event type wherein the spatial concordance is determined by determining and ranking a Euclidean distance of each of a plurality of consumption meters to everyone of the of others of the plurality of consumption meters, and by assigning the n closest consumption meters that have acquired this characterizing data to a group of consumption meters.

2. The method according claim 1, wherein the grouping is carried out by first grouping those events which are of the same event type, and secondly grouping by those events which match in time and lastly grouping by determining the spatial concordance.

3. The method according claim 1, wherein a rank of distance of the n closest consumption meters which have detected an event is used for grouping and n being between 20 and 100.

4. The method according to claim 1, wherein for temporal coincidence of the event time detection is split into the time of the beginning of an event and into the time of duration or the end of an event.

5. The method according to claim 4, wherein the time of the beginning and/or the end of the event is considered as being the same if it does not differ by more than a time difference limit.

6. The method according to claim 4, wherein the duration of an event is considered as being the same event if the duration does not differ more than a second time difference limit.

7. The method according to claim 1, wherein the meters are electric power meters.

8. The method according to claim 1, wherein for consistency of the event type the type of event is undervoltage, overvoltage, deviation of the mains frequency or power outage.

9. The method according to claim 8, wherein a trigger level of undervoltage, overvoltage and/or deviation of the mains frequency is adjustable.

10. The method according to claim 1, wherein the Euclidean distance of each of all consumption meters of the network to everyone of the others of the consumption meters of the network is determined by using geographical data or address data of the meters via the internet.

11. The method according to claim 1, wherein if one or more events are determined in one single consumption meter only this consumption meter will be marked as being defective.

12. The method according to claim 1, wherein a first time difference limit is used for comparing events reported by the same consumption meter and a second time difference limit is used for comparing events reported from different consumption meters, the second time difference limit being smaller than the first time difference limit.

13. The method according to claim 1, wherein once an event detected by a consumption meter has been added to a group of events the reported data from the next closest consumption meter is evaluated according to the criteria temporal coincidence and consistence of the event type and if the criteria are fulfilled the event is added to the group of events.

14. A head end system which is data connected to a plurality of consumption meters configured to execute the method comprising the steps of: determining an event by grouping data characterizing the event according to consumption meters which have detected this characterizing data and, if groups of consumption meters match, this is determined as an event, the grouping being carried out according to the following criteria temporal coincidence of the event spatial concordance of the event consistency of the event type wherein the spatial concordance is determined by determining and ranking a Euclidean distance of each of a plurality of consumption meters to everyone of the of others of the plurality of consumption meters, and by assigning the n closest consumption meters that have acquired this characterizing data to a group of consumption meters.

15. The head end system according claim 14, wherein the head end system is configured to adjust the trigger level of the consumption meters.

16. The head end system according claim 14, wherein the head end system is configured to calculate and rank the Euclidian distance of each of the consumption meters to everyone of the consumption meters connected to this head end system once before determining a plurality of events and/or from time to time after having determined a plurality of events.

17. The method according claim 1, wherein the rank of distance of the n closest consumption meters which have detected an event is used for grouping and n being 50.

18. The method according claim 5, wherein the time of the beginning and/or the end of the event is less than 15 seconds in the beginning.

19. The method according to claim 6, wherein the duration of an event is less than 30 seconds.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0035] In the drawings:

[0036] FIG. 1 is a schematic view of an electrical supply network;

[0037] FIG. 2a is a schematic view of a first example for ranking the distance;

[0038] FIG. 2b is a schematic view of a second example for ranking the distance with respect to another meter;

[0039] FIG. 3 is a graph showing three events of the same type with the differences in time and;

[0040] FIG. 4 is a flow process chart concerning the method for determining events in the network.

DESCRIPTION OF PREFERRED EMBODIMENTS

[0041] Referring to the drawings, in FIG. 1 an electrical supply network 1 is symbolized which shows a plenty of consumers in different houses 2 spread over town and country. Each house comprises at least one electric power meter 5 (not visible in FIG. 1) which is configured to detect electrical events in the network 1. Events are symbolized by circles 3 which are illustrated in solid, dotted, dot-dash-lines, which symbolize different events being detected. These events may differ in time, in type and in location.

[0042] These detected events are detected by electric power meters 5. These electric power meters 5 transmit their signals corresponding to the detected events wirelessly to a head end system 4 which is configured to evaluate these signals with the purpose of grouping in order to assign these events spatially, temporally and according to type. As this is an electrical network 1 the events may differ in the event type which may be overvoltage, undervoltage, mains frequency or power outage. The detection level of undervoltage and overvoltage is the same in all meters 5 which are installed in the houses 2. It corresponds to the standards and overvoltage/undervoltage is determined if the voltage deviates from the rated voltage by more than 15%. Undervoltage is determined if the voltage measured in a meter 5 installed in a house 2 is less than 85% of the nominal voltage. For grouping events they must match in type and in time. The time is split into the time of the beginning of an event and into the duration of the event. For the time of beginning there is a tolerance of about 60 seconds, with respect to the duration of the events there is a tolerance of about 120 seconds.

[0043] The tolerance in time is evaluated in the head end system 4 as well as the event type. However, the head end system 4 is configured to communicate with the meters 5 not only for receiving data but also for sending data. So the head end system 4 is able to adjust the detection level of undervoltage and overvoltage of a group or of all meters 5 connected to this head end system 4. This can be used to specify grouping for special events.

[0044] In FIG. 3 it is shown how three events A, B and C of the same type, for example undervoltage events, differ in time. The event A is detected by a first consumption meter starting at the time t1. Another consumption meter detects an event B which starts at t2. As can be seen events A and B have the same duration but different start times t1 and 2. The starting times t1 and t2 are sent by the consumption meters to the head end system which calculates the time difference between t2 and t1. A time difference limit of starting time tdl.sub.s is compared by the head end system to the time difference between t1 and t2 and as this difference is smaller than tdl.sub.s, events A and B are grouped which means they are regarded as events which coincide in time. From a third consumption meter the event C of the same type is detected and the starting time t3 and the duration is sent to the head end system. The head end system calculates the difference between the starting times t3 and t1, t2 and t3. The difference is smaller than tdl.sub.s. However, the duration of the event C is much longer than the duration of the events A and B (which are the same). The duration of the event C is too long, so the event C is not grouped with the events A and B. The duration of the event C is bigger than the time difference limit of duration tdl.sub.d. In practice there are two different time difference limits for an event, a time difference limit tdl.sub.s for the starting time and time difference limit tdl.sub.d for the duration. As there may be time fluctuations in the start time and in the stop time, the value for the time difference limit of the duration is usually bigger than the value for the time difference limit in the start time.

[0045] A very important factor for grouping however is the spatial concordance. This is done by ranking the Eucledian distance between one and all other meters. This ranking has to be done for each meter.

[0046] In practice the consistency of the event type is first checked, then the temporal coincidence of the event and at last the spatial concordance of the event.

[0047] In FIG. 2a and FIG. 2b there are symbolized eight meters 5 by small circles arranged at different distances from each other. In FIG. 2a the ranking of distances for the meter which is symbolized as a black circle is done with respect to the other seven meters 5 which are symbolized by circles. As can be seen in FIG. 2a this meter which is symbolized by a black circle has no distance to itself and is ranked to 0, the nearest meter 5 is ranked to 1, the second nearest to 2, the third nearest to 3 and so on up to 7 which has the longest distance to this (black) meter.

[0048] This ranking has to be done for each meter 5. In FIG. 2b this is done in the same manner as described in FIG. 2a however this ranking is done with respect to another meter 5 on the right side of this Figure. This meter 5 in FIG. 2b symbolized by the black circle has to be ranked with respect to the other meters 5. The nearest neighbor is ranked with 1 and so on up to the farthest neighbor which is ranked with 7. As shown for two different meters 5 in FIG. 2a and in FIG. 2b this ranking has to be done for every meter 5 installed in the houses 2 belonging to the network 1 in FIG. 1. This ranked distance of each meter can be set up in a distance matrix like the below matrix 1. For simplicity this matrix concerns the distances of only four meters. These meters are numbered with 1, 2, 3, and 4. There is a distance of ten metric meters between the meter no. 1 and the meter no. 2. There is a distance of 15 metric meters between the meter no. 1 and the meter no. 3. There is a distance of 20 metric meters between the meter no. 1 and the meter no. 4. The distance between the meter no. 2 and the meter no. 3 is six metric meters and the distance between the meter no. 3 and the meter no. 4 is seven metric meters.

TABLE-US-00001 Matrix 1: Meter 1 2 3 4 1 0 10 15 20 2 10 0 6 12 3 15 6 0 7 4 20 12 7 0

[0049] According to the invention this Euclidian distance is replaced by the rank of distance as can be shown in the following matrix 2. A matrix like this has to be created including all meters 5 in the network 1. For determining the spatial concordance of the event this matrix 2 is used and the rank of the distance of the n closest meters 5 which have detected the event is used for grouping. This n is 50 for example and can be varied depending on the type of meters and the network.

TABLE-US-00002 Matrix 2: Meter 1 2 3 4 1 0 1 2 3 2 2 0 1 3 3 3 1 0 2 4 3 2 1 0

[0050] As can be seen from FIG. 2a and FIG. 2b this ranking is independent from the real Eucledian distances. For example in FIG. 2a, where n is set to five (five meters have to be ranked) the distances of the meters in this area 6 are much smaller than the distances in this area 7 in FIG. 2b having the same ranking.

[0051] In practice there can be one million or more consumption meters which have to be ranked. Even if the number of meters is high, by doing this ranking storage space and computing power are comparatively low.

[0052] FIG. 4 shows a flow chart concerning this grouping method. For starting this method for determining events the first step 10 is to determine the Eucledian distance between each of the meters 5 installed in houses 2 connected to the network 1. There may be 100 meters or 1.000 meters or more meters of which the distance to each other has to be determined and ranked to each other. Concerning these electrical power meters 5 the Eucledian distance usually is not known and it is the easiest way to take the address of the house 2 where the meter 5 is installed and to use the GPS data of this address for determining the distances between the meters 5.

[0053] After having determined the distances between all meters 5 the distances between these meters 5 have to be ranked in a second step 11. How this is done has been shown above in the example according to FIG. 2a and FIG. 2b and with the ranking matrix 1 and 2.

[0054] After having ranked the distances of the meters 5 in the second step 11 this ranking of all meters to each other is stored in the head end system 4 (step 12). The event detection then begins in a step 13 by receiving signals from meters 5 of events which have been detected. These signals are sent to the head end system 4 where they are stored in the central server which may be part of the head end system 4 or which may be distanced from this and where this grouping of events is calculated.

[0055] After having received these event data and stored them they have to be evaluated. In this embodiment according to FIG. 4 event data received from the meters and stored in the central server in step 13 are calculated with respect to the duration of the events (step 14). These event data are then stored in an event queue 15 and an identification number is assigned to each (set of) event data including the ID number which identifies the electricity meter. This ID number is fixed already at installation and this ID number is included in the wireless data telegrams when the meter sends data.

[0056] Those event data of the event queue 15 are processed one after another starting with a first event data which is ranked to number 1 and which is identified with respect to the type and the time. This event data is stored in an event cluster 16. Then the second event data of the event queue 15 is processed. In step 17 it is checked if this event data comes from the same meter as the first event data or from another meter. If this event comes from a different meter than the first event data in step 18 the distance rank with respect to the first event data is checked. It is further checked in step 19 if this second event data is within the temporal threshold (i.e. time difference limits) with respect to the first event data. At last in step 20 it is checked if those events data are of the same type. If all these three steps 18, 19 and 20 have a positive result, the event data is added to a corresponding cluster 21. If one of these steps 18 to 20 is not positive this event data is added to the event queue 15. This event cluster 16 is the result of grouping of event data and each cluster concerns the same event which has been detected in the network of consumption meters.

[0057] So all the events data of the event queue 15 are processed in this manner and grouped into clusters which are arranged in the event cluster 16. This processing of events data compares each event data with each other events data which has been grouped in the event cluster 16. This fact is important with respect to step 18. As each event data is compared with each event data of an event cluster 16, it is not necessary that this rank of distance fits to each event data in an event cluster 16 is has only to fit in this set distance rank of one of the event data in this event cluster 16. This processing creates bigger event clusters and makes sure that all event data which originate from the same event are grouped in the same event cluster.

[0058] If an event data is identified as an event from the same meter in step 17, it is further checked if this is the same type of event (step 22) and it is further checked in step 23 if this event data corresponds with respect to the time limits. If both results of step 22 and 23 are positive this event data is considered to be same event data and the meter which has detected these events data is marked as not working correctly. If one or both results of these steps 22 and 23 are negative, the event data is checked in the steps 18, 19, 20 belonging to another cluster in the event cluster 16. These check boxes 18, 19 and 20 are executed for each event data in the event cluster 16. If the event data does not belong to any of the clusters in the event cluster 16, it is removed from the event queue 15.

[0059] Further details of this method for determining events in the network are disclosed in the flow process chart according FIG. 4 where these details are described in detail: 24 consider single even, 25 consider separate event, leave in event queue, 26 look at event from queue, 27 look at event from cluster, 28 any events left in selection?, 29 is the event being searched from an established cluster?, 30 the event does not belong to a cluster. Remove from event queue, 31 no new events to add to cluster, 32 uninvestigated events left in cluster?, 33 clustered events are investigated to see if their neighbor belong to the clusters too, 34 cluster finished, 35 events left in event queue?, 36 all events are scanned and all cluster built. Program ends until receiving new events.

[0060] While specific embodiments of the invention have been shown and described in detail to illustrate the application of the principles of the invention, it will be understood that the invention may be embodied otherwise without departing from such principles.

LIST OF NUMERALS

[0061] 1 network [0062] 2 houses with meters inside [0063] 3 circles [0064] 4 head end system [0065] meter [0066] 6 area [0067] 7 area [0068] tdl time difference limit [0069] tdl.sub.s time difference limit in start time [0070] tdl.sub.d time difference limit in duration [0071] 10 first step, calculating [0072] 11 second step, ranking [0073] 12 step, storing [0074] 13 step, receiving event data, storing [0075] 14 step, calculating duration [0076] 15 event queue [0077] 16 cluster queue [0078] 17 step, checking where the event comes from [0079] 18 step, checking the rank [0080] 19 step, checking the time [0081] 20 step, checking the type [0082] 21 step, add the event to a cluster [0083] 22 step, checking type [0084] 23 step, checking the time [0085] 24 step, consider single event [0086] 25 step, consider separate event, leave in event queue [0087] 26 step, look at event from queue [0088] 27 step, look at event from cluster [0089] 28 step, any events left in selection [0090] 29 step, is the event being searched from an established cluster [0091] 30 step, the event does not belong to a cluster. Remove from event queue [0092] 31 step, no new events to add to cluster [0093] 32 step, uninvestigated events left in cluster [0094] 33 step, clustered events are investigated to see if their neighbor belong to the clusters too [0095] 34 step, cluster finished [0096] 35 step, events left in event queue [0097] 36 step, all events are scanned and all cluster built. Program ends until receiving new events