Privacy-preserving configuration of location-specific electrical load models
Methods, systems, and apparatus, including medium-encoded computer program products, for configuring location-specific electrical load models while preserving privacy. A first machine learning model configured to predict electrical load curves of an electrical utility grid can be obtained from a server. Load values associated with a particular region of the electrical utility grid can be obtained. The load values can be applied as calibration input to the first machine learning model to produce first adjustment parameters for the first machine learning model. The first adjustment parameters can be provided to the server.
1 . A computer implemented method implemented on one or more computing devices, the method comprising:
obtaining, by a first server and from a second server, a first machine learning model configured to predict electrical load curves of an electrical utility grid, wherein the first server is associated with a particular region of the electrical utility grid;
obtaining, by the first server and from a plurality of electric meters within the particular region, load values associated with the particular region of the electrical utility grid;
applying, by the first server, the load values as calibration input to the first machine learning model to produce first adjustment parameters for the first machine learning model; and
providing, to the second server, the first adjustment parameters without disclosing the load values to the second server, thereby, preserving privacy of customers associated with the load values;
configuring, by the first server, the first machine learning model according to the first adjustment parameters; and
processing an input using the first machine learning model to produce a predicted electrical load curve at the particular region of the electrical utility grid.
2 . The computer implemented method of claim 1 , further comprising:
providing the electrical load curve.
3 . The computer implemented method of claim 1 , further comprising:
obtaining, from the second server, second adjustment parameters; and
configuring, by the first server, the first machine learning model according to the second adjustment parameters.
4 . The computer implemented method of claim 1 , further comprising:
obtaining, from the second server, a second machine learning model configured, at least in part, according to the first adjustment parameters.
5 . The computer implemented method of claim 1 , wherein the first adjustment parameters include one or more gradients.
6 . The computer implemented method of claim 1 , wherein obtaining, by the first server and from a plurality of electric meters within the particular region, load values associated with the particular region of the electrical utility grid comprises: receiving data from a meter over a network.
7 . The computer implemented method of claim 6 , wherein the meter is an advanced metering infrastructure (AMI) meter.
8 . The computer implemented method of claim 7 , wherein at least a subset of the load values are provided by an automated meter reading (AMR) system.
9 . The computer implemented method of claim 1 , further comprising obtaining load values associated with the particular region of the electrical utility grid by reading data stored on a storage device or on a storage system.
10 . The computer implemented method of claim 1 , wherein the first machine learning model is trained to predict generalized electrical load curves for a generic electrical utility grid.
11 . The computer implemented method of claim 1 wherein the number of applied load values used as calibration input to the first machine learning model does not exceed ten million.
12 . A system comprising one or more computers and one or more storage devices storing instructions that when executed by the one or more computers cause the one or more computers to perform operations comprising:
obtaining, by a first server and from a second server, a first machine learning model configured to predict electrical load curves of an electrical utility grid, wherein the first server is associated with a particular region of the electrical utility grid;
obtaining, by the first server and from a plurality of electric meters within the particular region, load values associated with the particular region of the electrical utility grid;
applying, by the first server, the load values as calibration input to the first machine learning model to produce first adjustment parameters for the first machine learning model; and
providing, to the second server, the first adjustment parameters without disclosing the load values to the second server, thereby, preserving privacy of customers associated with the load values;
configuring, by the first server, the first machine learning model according to the first adjustment parameters; and
processing an input using the first machine learning model to produce a predicted electrical load curve at the particular region of the electrical utility grid.
13 . The system of claim 12 , wherein the operations further comprise:
providing the electrical load curve.
14 . The system of claim 12 , wherein the operations further comprise:
obtaining, from the second server, second adjustment parameters; and
configuring, by the first server, the first machine learning model according to the second adjustment parameters.
15 . The system of claim 12 , wherein the operations further comprise:
obtaining, from the second server, a second machine learning model configured, at least in part, according to the first adjustment parameters.
16 . The system of claim 12 , wherein the first adjustment parameters include one or more gradients.
17 . The system of claim 12 , wherein obtaining, by the first server and from a plurality of electric meters within the particular region, load values associated with the particular region of the electrical utility grid comprises: receiving data from a meter over a network.
18 . The system of claim 17 , wherein the meter is an advanced metering infrastructure (AMI) meter.
19 . The system of claim 12 , further comprising obtaining load values associated with the particular region of the electrical utility grid by reading data stored on a storage device or on a storage system.
20 . One or more non-transitory computer-readable storage media storing instructions that when executed by one or more computers cause the one or more computers to perform operations comprising:
obtaining, by a first server and from a second server, a first machine learning model configured to predict electrical load curves of an electrical utility grid, wherein the first server is associated with a particular region of the electrical utility grid;
obtaining, by the first server and from a plurality of electric meters within the particular region, load values associated with the particular region of the electrical utility grid;
applying, by the first server, the load values as calibration input to the first machine learning model to produce first adjustment parameters for the first machine learning model; and
providing, to the second server, the first adjustment parameters without disclosing the load values to the second server, thereby, preserving privacy of customers associated with the load values;
configuring, by the first server, the first machine learning model according to the first adjustment parameters; and
processing an input using the first machine learning model to produce a predicted electrical load curve at the particular region of the electrical utility grid.