Detecting non-technical losses in electrical networks based on multi-layered statistical techniques from smart meter data

A computer program product, the computer program product comprising a computer readable storage medium having program instructions embodied therewith, the program instructions executable by a computing device to cause the computing device to:

• compute a consumption estimation for each of multiple consumers associated with an electrical distribution network based on a plurality of items of input data, wherein said computing is carried out by at least one computing device communicatively linked to (i) a plurality of smart meters monitoring electrical usage of the multiple consumers within the electrical distribution network and (ii) one or more additional data sources;
• determine a difference between (i) the consumption estimation for each of the multiple consumers and (ii) actual consumption for each of the multiple consumers;
• cluster the multiple consumers into one or more clusters based on a consumption pattern associated with each of the multiple consumers at a given point in time;
• determine a level of deviation of (i) the consumption pattern associated with each of the multiple consumers at the given point in time from (ii) a consumption pattern representative of the cluster to which each of the multiple consumers belongs;
• cluster the multiple consumers into two or more clusters based on a consumption pattern associated with each of the multiple consumers during a first interval of time;
• cluster the multiple consumers into the two or more clusters based on a consumption pattern associated with each of the multiple consumers during a second interval of time;
• determine, for each of the multiple consumers, a level of evolution from (i) a first of the two of more clusters during the first interval of time to (ii) a second of the two or more clusters during the second interval of time;
• identify one or more of the multiple consumers associated with a given type of loss within the electrical distribution network based on (i) the determined difference, (ii) the determined level of deviation, and (iii) the determined level of evolution;
• and
• implement a feedback mechanism, wherein the feedback mechanism comprises inspection of at least one of (i) the one or more of the multiple consumers associated with a non-technical loss within the electrical distribution network, (ii) the determined difference, (iii) the determined level of deviation, and (iv) the determined level of evolution.

A system comprising:

• a memory;
• and
• at least one processor coupled to the memory and configured for:
• computing a consumption estimation for each of multiple consumers associated with an electrical distribution network based on a plurality of items of input data, wherein said computing is carried out by at least one computing device communicatively linked to (i) a plurality of smart meters monitoring electrical usage of the multiple consumers within the electrical distribution network and (ii) one or more additional data sources;
• determining a difference between (i) the consumption estimation for each of the multiple consumers and (ii) actual consumption for each of the multiple consumers;
• clustering the multiple consumers into one or more clusters based on a consumption pattern associated with each of the multiple consumers at a given point in time;
• determining a level of deviation of (i) the consumption pattern associated with each of the multiple consumers at the given point in time from (ii) a consumption pattern representative of the cluster to which each of the multiple consumers belongs;
• clustering the multiple consumers into two or more clusters based on a consumption pattern associated with each of the multiple consumers during a first interval of time;
• clustering the multiple consumers into the two or more clusters based on a consumption pattern associated with each of the multiple consumers during a second interval of time;
• determining, for each of the multiple consumers, a level of evolution from (i) a first of the two of more clusters during the first interval of time to (ii) a second of the two or more clusters during the second interval of time;
• identifying one or more of the multiple consumers associated with a given type of loss within the electrical distribution network based on (i) the determined difference, (ii) the determined level of deviation, and (iii) the determined level of evolution;
• and
• implementing a feedback mechanism, wherein the feedback mechanism comprises inspection of at least one of (i) the one or more of the multiple consumers associated with a non-technical loss within the electrical distribution network, (ii) the determined difference, (iii) the determined level of deviation, and (iv) the determined level of evolution.