Detection of electric vehicle charging at premise

1. A computer implemented method performed by a computing device with at least one processor, the method comprising:

receiving unknown time series data of usage values of electricity consumption from a plurality of accounts, wherein the unknown time series data that corresponds to an account is unknown to have electric vehicle (EV) charge events;

for a given account of the plurality of accounts:

converting the unknown time series data of usage values into a data structure of time intervals with corresponding usage values;

encoding each of the time intervals with a symbol from a series of symbols representing a level of electricity consumption during the time interval;

wherein the encoding generates an encoded consumption pattern of symbols;

executing a machine learning classifier with the encoded consumption pattern of symbols, wherein the machine learning classifier is trained to identify a set of known electric vehicle (EV) charge motifs comprising a sequence of symbols that represents an EV charging event that is distinguished from a non-charge motif;

detecting, by the machine learning classifier, whether the encoded consumption pattern of symbols from the given account includes a string of high usage symbols that are similar to one of the set of known EV charge motifs; and

marking, by the machine learning classifier, the given account as having an electric vehicle charge event or as not having an electric vehicle charge event based at least on the detecting.

2. The computer implemented method of claim 1 , wherein the encoding further comprises:

applying an encoding scheme of letters that represent the level of electricity consumption from a low usage level to a high usage level relative to a range of the usage values in the unknown time series data.

3. The computer implemented method of claim 1 , wherein the encoding further comprises:

encoding each of the time intervals of electricity usage into one symbol from the series of symbols representing the level of electricity consumption in a given time interval, wherein a series of sequential time intervals becomes the encoded consumption pattern of symbols.

4. The computer implemented method of claim 1 , wherein the marking further comprises:

in response to detecting that the encoded consumption pattern of symbols matches one motif from the set of known EV charge motifs, identify the given account as having an electric vehicle; and

in response to detecting that the encoded consumption pattern of symbols does not match one motif from the set of known EV charge motifs, identify the given account as a non-electric vehicle owner.

5. The computer implemented method of claim 1 , wherein detecting whether the encoded consumption pattern of symbols includes the string of high usage symbols comprises:

comparing one or more charge features from one or more known EV charging events to corresponding charge features from the unknown time series data to determine if a match is found within a threshold; and

generating an output in accordance with a result of the match.

6. The computer implemented method of claim 1 , wherein detecting whether the encoded consumption pattern of symbols includes the string of high usage symbols comprises:

searching the encoded consumption pattern of symbols from the unknown time series data for a symbol pattern of high usage symbols over a sustained time period of time intervals;

identifying the symbol pattern of high usage symbols as a suspected EV charge event;

determining charge features from the usage values in the sustained time period of time intervals; and

determining, by the machine learning classifier, whether the symbol pattern and the charge features of the suspected EV charge event match any of the known EV charge motifs and corresponding known charge features of a known EV charge event.

7. The computer implemented method of claim 1 , wherein the method further comprises:

generating and transmitting an electronic message to a remote device associated with the account, wherein the electronic message includes instructions to change EV charging times to off-peak hours.

8. A computing system, comprising:

at least one memory;

at least one processor connected to the at least one memory;

an electric vehicle detection module stored on a non-transitory computer readable medium and including instructions that when executed by at least the processor cause the processor to:

receive unknown time series data of usage values of electricity consumption from a given account of a plurality of accounts, wherein the unknown time series data is unknown to have electric vehicle (EV) charge events;

convert the unknown time series data of usage values into a data structure of time intervals with corresponding usage values;

encode each of the time intervals with a symbol from a series of symbols representing a level of electricity consumption during the time interval;

wherein the encoding generates an encoded consumption pattern of symbols;

execute a machine learning classifier with the encoded consumption pattern of symbols, wherein the machine learning classifier is trained to identify a set of known electric vehicle (EV) charge motifs comprising a sequence of symbols that represents an EV charging event that is distinguished from a non-charge motif;

detect, by the machine learning classifier, whether the encoded consumption pattern of symbols from the given account includes a string of high usage symbols that are similar to one of the set of known EV charge motifs; and

mark, by the machine learning classifier, the given account as having an electric vehicle charge event or as not having an electric vehicle charge event based at least on the detecting.

9. The computing system of claim 8 , wherein the instructions for encoding are configured to cause the processor to:

apply an encoding scheme of letters that represent the level of electricity consumption from a low usage level to a high usage level relative to a range of the usage values in the unknown time series data.

10. The computing system of claim 8 , wherein the instructions for encoding are configured to cause the processor to:

encode each of the time intervals of electricity usage into one symbol from the series of symbols representing the level of electricity consumption in a given time interval, wherein a series of sequential time intervals becomes the encoded consumption pattern of symbols.

11. The computing system of claim 8 , wherein the instructions for marking further comprise instructions configured to cause the processor to:

in response to detecting that the encoded consumption pattern of symbols matches one motif from the set of known EV charge motifs, identify the given account as having an electric vehicle; and

in response to detecting that the encoded consumption pattern of symbols does not match one motif from the set of known EV charge motifs, identify the given account as a non-electric vehicle owner.

12. The computing system of claim 8 , wherein the instructions for detecting whether the encoded consumption pattern of symbols includes the string of high usage symbols further comprise instructions configured to cause the processor to:

compare one or more charge features from one or more known EV charging events to corresponding charge features from the unknown time series data to determine if a match is found within a threshold; and

generate an output in accordance with a result of the match.

13. The computing system of claim 8 , wherein the instructions for detecting whether the encoded consumption pattern of symbols includes the string of high usage symbols further comprise instructions configured to cause the processor to:

search the encoded consumption pattern of symbols from the unknown time series data for a symbol pattern of high usage symbols over a sustained time period of time intervals;

identify the symbol pattern of high usage symbols as a suspected EV charge event;

determine charge features from the usage values in the sustained time period of time intervals; and

determine, by the machine learning classifier, whether the symbol pattern and the charge features of the suspected EV charge event match any of the known EV charge motifs and corresponding known charge features of a known EV charge event.

14. The computing system of claim 8 , wherein the instructions further comprising instructions to cause the processor to:

identify locations of EV owners in an electric grid based on the plurality of accounts that are marked as having the electric vehicle charge event; and

identify EV hot spots in the electric grid that represent a load increase in an area that has multiple EV owners.

15. The computing system of claim 8 , wherein the instructions further comprising instructions to cause the processor to, for an account that is marked as having the electric vehicle charge event:

generate and transmit an electronic message to a remote device associated with the account, wherein the electronic message includes instructions to change EV charging times to off-peak hours.

16. A non-transitory computer-readable medium storing computer-executable instructions that when executed by at least a processor of a computer cause the computer to:

receive unknown time series data of usage values of electricity consumption from given account from a plurality of accounts, wherein the unknown time series data is unknown to have electric vehicle (EV) charge events;

convert the unknown time series data of usage values into a data structure of time intervals with corresponding usage values;

encode each of the time intervals with a symbol from a series of symbols representing a level of electricity consumption during the time interval;

wherein the encoding generates an encoded consumption pattern of symbols;

detect whether the encoded consumption pattern of symbols from the given account includes a string of high usage symbols that are similar to a known EV charge motif that represents a known EV charging event that is distinguished from a non-charge motif; and

mark the given account as having an electric vehicle charge event or as not having an electric vehicle charge event based at least on the detecting.

17. The non-transitory computer-readable medium of claim 16 , wherein the instructions for encoding are configured to cause the processor to:

apply an encoding scheme of letters that represent the level of electricity consumption from a low usage level to a high usage level relative to a range of the usage values in the unknown time series data.

18. The non-transitory computer-readable medium of claim 16 , wherein the instructions for encoding are configured to cause the processor to:

encode each of the time intervals of electricity usage into one symbol from the series of symbols representing the level of electricity consumption in a given time interval, wherein a series of sequential time intervals becomes the encoded consumption pattern of symbols.

19. The non-transitory computer-readable medium of claim 16 , wherein the instructions for marking further comprise instructions configured to cause the processor to:

in response to detecting that the encoded consumption pattern of symbols matches one motif from the set of known EV charge motifs, identify the given account as having an electric vehicle;

in response to detecting that the encoded consumption pattern of symbols does not match one motif from the set of known EV charge motifs, identify the given account as a non-electric vehicle owner; and

repeat the converting, encoding, detecting, and marking for other accounts from the plurality of accounts.

20. The non-transitory computer-readable medium of claim 16 , wherein the instructions for detecting whether the encoded consumption pattern of symbols includes the string of high usage symbols further comprise instructions configured to cause the processor to:

search the encoded consumption pattern of symbols from the unknown time series data for a symbol pattern of high usage symbols over a sustained time period of time intervals;

identify the symbol pattern of high usage symbols as a suspected EV charge event;

determine charge features from the usage values in the sustained time period of time intervals; and

determine whether the symbol pattern and the charge features of the suspected EV charge event match any of the known EV charge motifs and corresponding known charge features of a known EV charge event.