Method and system for electricity consumption prediction

Patent metadata

Field	Value
Publication number	US-2025087998-A1
Application number	US-202418805294-A
Country	US
Kind code	A1
Filing date	Aug 14, 2024
Priority date	Sep 7, 2023
Publication date	Mar 13, 2025
Grant date	—

Abstract

Existing electricity consumption prediction approaches depend largely on data-based models, which may be statistical techniques or more precisely time-series predictions that predict based on the auto-regressive nature of the load curve with a few external variables at best, such as calendar events and ambient temperature. While these models are effective to a certain degree for the overall grid level requirements, they may not be able to predict disruptive changes that may happen over longer periods of time, such as the demographic shifts, etc. Method and system disclosed herein predict the electricity consumption by taking into consideration various parameters associated with such disruptive changes, and then predict the electricity consumption in a target area by aggregating the electricity consumption predicted at the agent level.

First claim

What is claimed is:

1 . A processor implemented method of electricity consumption prediction, comprising:
receiving, via one or more hardware processors, a requirement data comprising a) a region of interest, b) a target time period; predicting, via the one or more hardware processors, one or more distributions of interest for a plurality of agents from a population of the region of interest for the target time period, using a plurality of data models, wherein the one or more distributions of interest comprises of a spatial and temporal distribution of each of the plurality of agents and associated current values; grouping, via the one or more hardware processors, each of the plurality of agents to one of a) an individual agent category, and b) a collective agent category, characterized by one or more activities being performed by each of the plurality of agents, by processing the spatial and temporal distribution of each of the plurality of agents and associated current values; predicting, via the one or more hardware processors, the electricity consumption for the target time period, for each of the plurality of agents in the individual agent category and the collective agent category, based on a) the one or more distributions of interest predicted for each of the plurality of agents, b) information on one or more electrical devices and associated parameter values for the one or more distributions of interest, and c) a historical data with respect to one or more factors affecting the electricity consumption for the one or more distributions of interest and for the associated parameter values for different combinations of the one or more electrical devices; and aggregating, via the one or more hardware processors, the predicted electricity consumption of the plurality of agents, to determine an electricity consumption at the region of interest.

2 . The processor-implemented method of claim 1 , wherein one or more of the plurality of agents undergo transition between two or more agent categories during the target time period, wherein by following the transition between the two or more agent categories, associated one or more activities and one or more electrical devices are identified.

3 . The processor-implemented method of claim 1 , wherein the plurality of data models comprises of a residential consumer model, a transport consumer model, an industrial consumer model, an institutional consumer model, an agricultural consumer model, and an environmental model, and wherein the plurality of data models are physics-driven, data-driven and hybrid models.

4 . The processor-implemented method of claim 3 , wherein the plurality of data models are re-tuned if a measured accuracy of the electricity consumption prediction is below a threshold of accuracy.

5 . The processor-implemented method of claim 1 , wherein the one or more factors affecting the electricity consumption, forming the historical data, and associated current data, comprises a) a weather data, b) a demographic and economic data, c) an energy price data, d) a calendar data, e) a technology data, f) an administrative and policy decision data, g) a land use data, and h) a usage and behaviour data.

6 . The processor-implemented method of claim 1 , wherein the plurality of data models is trained using the historical data as training data, comprising:
pre-processing the historical data to obtain a pre-processed historical data; dividing the historical data to a training data set and a testing dataset; training each of the plurality of data models using the training data set to capture one or more distribution parameters associated with an energy consumption pattern of the historical data at each of a plurality of time instances; and testing each of the plurality of data models using the testing dataset, to obtain an associated confidence score, wherein each of the plurality of data models is retuned till the associated confidence score is at least matching a threshold of confidence score.

7 . The processor-implemented method of claim 1 , wherein the current values in the spatial distribution of each of the plurality of agents are associated with one or more of a current location of the agent, a collective group, an agent role, the one or more activities, and one or more electrical device characteristics.

8 . The processor-implemented method of claim 1 , wherein the distribution of interest comprises of a behavioral distribution of a plurality of roles associated with each of the plurality of agents, wherein each of the plurality of roles has an associated set of probable activities, and wherein the one or more electrical devices are associated with the set of probable activities.

9 . A system for electricity consumption prediction, comprising:
one or more hardware processors; a communication interface; and a memory storing a plurality of instructions, wherein the plurality of instructions causes the one or more hardware processors to:
receive a requirement data comprising a) a region of interest, b) a target time period;
predict one or more distributions of interest for a plurality of agents from a population of the region of interest for the target time period, using a plurality of data models, wherein the one or more distributions of interest comprises of a spatial and temporal distribution of each of the plurality of agents and associated current values;
group each of the plurality of agents to one of a) an individual agent category, and b) a collective agent category, characterized by one or more activities, by processing the spatial and temporal distribution of each of the plurality of agents and associated current values;
predict the electricity consumption for the target time period, for each of the plurality of agents in the individual agent category and the collective agent category, based on a) the one or more distributions of interest predicted for each of the plurality of agents, b) information on one or more electrical devices and associated parameter values for the one or more distributions of interest, and c) a historical data with respect to one or more factors affecting the electricity consumption for the one or more distributions of interest and for the associated parameter values for different combinations of the one or more electrical devices; and
aggregate the predicted electricity consumption of the plurality of agents, to determine an electricity consumption at the region of interest.

10 . The system of claim 9 , wherein one or more of the plurality of agents undergo transition between two or more agent categories during the target time period, and wherein the one or more hardware processors are configured to identify associated one or more activities and one or more electrical devices by following the transition between the two or more agent categories.

11 . The system of claim 9 , wherein the plurality of data models comprises of a residential consumer model, a transport consumer model, an industrial consumer model, an institutional consumer model, an agricultural consumer model, and an environmental model, and wherein the plurality of data models are physics-driven, data-driven and hybrid models.

12 . The system of claim 11 , wherein the one or more hardware processors are configured to re-tune the plurality of data models if a measured accuracy of the electricity consumption prediction is below a threshold of accuracy.

13 . The system a claimed in claim 9 , wherein the one or more factors affecting the electricity consumption, forming the historical data and associated current data, comprises a) a weather data, b) a demographic and economic data, c

Assignees

• Tata Consultancy Services Ltd

Inventors

• Subramanian Sivakumar
• Gujar Pranav
• Nistala Sri Harsha
• Shinde Gopal
• Agrawal Roshni
• Runkana Venkataramana
• Kulkarni Vinay

Classifications

• H02J2103/30

  Simulating, planning, modelling, reliability check or computer assisted design [CAD] of electric power networks · CPC title

• G06Q10/06

  Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling · CPC title

• G06Q10/04

  Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem" (market predictions or forecasting for commercial activities G06Q30/0202) · CPC title

• H02J3/003 Primary

  Load forecast, e.g. methods or systems for forecasting future load demand · CPC title

• G06Q50/06 Primary

  Energy or water supply · CPC title