Back to results
US 8,805,998 B2Granted
Automatic matching of sources to loads
AssigneeEaton Corporation
InventorLuis R. Pereira
Life of this patent
Estimating the term…
Abstract
A method of matching sources to loads includes receiving load activity data and source activity data corresponding to samples of activity over a time period at a plurality of loads and a plurality of sources, respectively. The method of matching sources to loads further includes scoring similarity between the load activity and the source activity based on the load activity data and the source activity data, and matching loads to sources based on the scoring.
Claims17
A computer implemented method comprising:
- a computer obtaining a first load activity data corresponding to activity at a first load over a time period;
- a computer obtaining a second load activity data corresponding to activity at a second load over the time period;
- a computer obtaining a first source activity data corresponding to activity at a first source over the time period;
- a computer obtaining a second source activity data corresponding to activity at a second source over the time period;
- a computer calculating, using a correlation coefficient, an association cost metric between:
- the first load activity data and the first source activity data,
- the first load activity data and the second source activity data,
- the second load activity data and the first source activity data, and
- the second load activity data and the second source activity data;
- a computer using a lower bounding function for indexing time series to eliminate at least one pair, comprising one of the first load activity data and the second load activity data and one of the first source activity data and the second source activity data, that is not a best match;
- and
- a computer matching the first load to one of the first source and the second source and matching the second load to one of the first source and the second source according to an optimization criteria applied to the calculated association cost metrics, wherein a higher similarity corresponds to a lower association cost metric and the optimization criteria corresponds to minimizing a sum of the association cost metrics.
9 more elements
A computer implemented method comprising:
- a computer receiving load activity data corresponding to sampled activity over a time period at a plurality of loads;
- a computer receiving source activity data corresponding to sampled activity over the time period at a plurality of sources;
- a computer calculating, using a correlation coefficient, association cost metrics within sample pairs of the load activity data and the source activity data;
- a computer using a lower bounding function for indexing time series to eliminate at least one pair, comprising one of the plurality of loads and one of the plurality of sources, that is not a best match;
- and
- a computer assigning each load to at least one source based on the association cost metrics according to an optimization criteria, where a higher similarity corresponds to a lower value for the association cost metrics and the optimization criteria corresponds to minimization of a sum of the values of the association cost metrics.
3 more elements
A computer implemented method comprising:
- a computer receiving load activity data corresponding to samples of activity over a time period at a plurality of loads;
- a computer receiving source activity data corresponding to samples of activity over the time period at a plurality of sources;
- a computer calculating, using a correlation coefficient, association cost metrics between the load activity and the source activity based on the load activity data and the source activity data;
- a computer using a lower bounding function for indexing time series to eliminate at least one pair, comprising one of the plurality of loads and one of the plurality of sources, that is not a best match;
- and
- a computer assigning each load from the plurality of loads to at least one source from the plurality of sources based on the calculated association cost metrics, such that the sum of the association cost metrics is minimized, where a higher similarity corresponds to a lower association cost metrics.
3 more elements
Dependencies are inferred from the claim text.