Detection of electric power diversion

1. A method, comprising:

obtaining time-stamped voltage and current measurement data measured by an electrical meter;

obtaining a time-series of voltage-changes and current-changes based on the time-stamped voltage and current measurement data, wherein:

the time-series of voltage-changes and current-changes are determined based on time-stamped voltage and current measurement data measured by the electrical meter;

the time-series of voltage-changes and current-changes comprises a plurality of data points resulting from a plurality of load changes measured by the electrical meter; and

the time-series of voltage-changes and current-changes comprises change-data indicating current changes and voltage changes resulting from appliances that are turned on or off;

performing a regression analysis of the voltage-changes against the current-changes; and

outputting an indication of whether the electrical meter was bypassed.

2. The method of claim 1 , wherein a statistically significant number of on/off events involving electrical loads measured by the electrical meter comprise at least one appliance turned-on event and at least one appliance turned-off event.

3. The method of claim 1 , wherein:

the regression analysis indicates a degree of impedance correlation; and

determining if the electrical meter was bypassed is based at least in part on the degree of impedance correlation.

4. The method of claim 1 , wherein the time-series of voltage-changes and current-changes is based on intervals of 15 minutes or less.

5. The method of claim 1 , additionally comprising:

obtaining additional time-series of voltage-changes and current-changes over a period of days and at different times of day; and

flagging the electrical meter as potentially bypassed responsive to calculated correlations that differ by a statistically significant amount.

6. The method of claim 1 , wherein measurements to create the time-series of voltage-changes and current-changes begin at least in part in response to a load change at the electrical meter.

7. The method of claim 1 , additionally comprising:

obtaining an additional time-series of voltage-changes and current-changes over a different time period than was used to obtain the additional time-series of voltage-changes and current-changes;

performing an additional regression analysis on the additional time-series of voltage-changes and current-changes; and

flagging the electrical meter to indicate a bypass responsive to statistically different values between the regression analysis and the additional regression analysis.

8. The method of claim 1 , wherein:

the electrical meter performs ongoing regression analyses; and

the electrical meter transmits at least one of an impedance value and a correlation value obtained from the regression analyses.

9. The method of claim 1 , further comprising:

determining that a statistically significant number of on/off events were not seen within a threshold period of time;

determining that power was used during the threshold period of time; and

flagging the electrical meter as potentially bypassed in response to the determining that power was used during the threshold period of time.

10. The method of claim 1 further comprising:

obtaining a time-series of voltage-changes and current-changes periodically over multiple time periods; and

determining if there is consistency between correlations associated with each of the multiple time periods.

11. The method of claim 1 , wherein determining whether the electrical meter was bypassed is additionally based on factors comprising:

an R 2 value of time-series of voltage-changes and current-changes of each meter connected to a transformer; and

a correlation of R 2 values of meters connected to the transformer.

12. A meter, comprising:

a processor;

a memory, in communication with the processor;

a metrology unit, in communication with the processor, to provide time-stamped voltage measurements and time-stamped current measurements over time;

a time-series generator, defined in the memory and executed by the processor, to generate a time-series of voltage-changes and current-changes based at least in part on the time-stamped voltage measurements and the time-stamped current measurements over time, wherein the time-series of voltage-changes and current-changes comprises a plurality of data points resulting from a plurality of load changes measured by the metrology unit, and wherein the time-series of voltage-changes and current-changes comprises change-data indicating current changes and voltage changes resulting from appliances that are turned on or off;

an analyzer, defined in the memory and executed by the processor, to perform an analysis on voltage-change data against current-change data, and to perform actions comprising:

performing a regression analysis of the time-series of voltage-changes and current-changes;

comparing output of the regression analysis to a regression analysis of at least one meter connected to a same transformer as the meter; and

outputting, based at least in part on the comparing, an indication of whether the meter was bypassed.

13. The meter as recited in claim 12 , wherein the analyzer performs a regression analysis function.

14. The meter as recited in claim 12 , wherein the analyzer performs a least squares regression of the voltage-change data against the current-change data of the time series of voltage-changes and current-changes.

15. The meter as recited in claim 12 , wherein the time-series generator begins generation of the time-series of voltage-changes and current-changes in response to a load change measured by the metrology unit.

16. The meter as recited in claim 12 , wherein a notification module transmits notification of an impedance calculation or the regression analysis consistent with a bypass of the meter.

17. The meter as recited in claim 12 , wherein the time-series generator is programmed to determine whether a statistically significant number of on/off events indicate a valid correlation between voltage-changes and current-changes.

18. An electrical meter, comprising:

one or more processors;

a metrology unit in communication with the one or more processors;

memory, in communication with the one or more processors, the memory storing processor-executable instructions that, when executed by the one or more processors, cause the electrical meter to perform acts comprising:

measuring voltage and current at the metrology unit;

creating time-stamped voltage and current measurement data using data output from the metrology unit;

creating a time-series of voltage-changes and current-changes based on the time-stamped voltage and current measurement data measured by the metrology unit, wherein the time-series of voltage-changes and current-changes comprise a plurality of data points, and wherein the time-series of voltage-changes and current-changes comprises change-data indicating current changes and voltage changes resulting from appliances that are turned on or off;

performing a regression analysis of the time-series of voltage-changes and current-changes;

determining if the electrical meter was bypassed based at least in part on the regression analysis; and

outputting an indication of whether the electrical meter was bypassed.

19. The electrical meter of claim 1 , wherein determining whether the electrical meter was bypassed is additionally based on factors comprising:

an R 2 value of the time-series of voltage-changes and current-changes of each meter connected to a transformer; and

a correlation of R 2 values of meters connected to the transformer.