Electric Grid Load Prediction With Edge-Ready Autoregressive Modeling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for predicting electric energy consumption in electric grids face challenges due to limited computational and data storage resources in field devices, leading to poor reliability and accuracy compared to machine learning-based methods.
Innovation Solution
A method that acquires detection data, calendar data, and calculates training data to set a linear auto-regressive mathematical model, processing exogenous input values to predict electric energy consumption, suitable for implementation in Edge computing systems.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If machine learning-based prediction methods are used, then prediction accuracy is improved, but computational resource requirements increase
Solution Approach 1:
The patent extracts only the essential periodic components from historical consumption data using Fourier analysis, rather than processing complete raw datasets. This extraction of key periodic patterns enables accurate prediction while significantly reducing the computational burden on edge devices with limited resources.
Solution Approach 2:
The patent creates simplified mathematical models that replicate the essential periodic behavior of energy consumption patterns. By using autoregressive models with exogenous inputs based on extracted periodic components, the system achieves ML-level accuracy without requiring complex neural network architectures or large-scale computational resources.
2Device complexity
If linear regression-based prediction methods are used, then computational resource requirements are reduced, but prediction accuracy deteriorates
Solution Approach 1:
The patent transforms the input parameters by decomposing historical consumption data into periodic components using Fourier analysis. This parameter transformation enables a linear autoregressive model to capture complex periodic patterns that would otherwise require non-linear machine learning methods, thereby maintaining high accuracy with reduced computational complexity.
Solution Approach 2:
The patent introduces periodic component analysis as an intermediary processing step between raw data collection and prediction. By using Fourier analysis to extract periodic patterns as intermediate representations, the system bridges the gap between simple linear models and complex machine learning approaches, achieving both accuracy and efficiency.
3Reliability
If large amounts of data are processed for training, then prediction reliability is improved, but data storage requirements increase
Solution Approach 1:
The patent extracts only the essential periodic components from large historical datasets using Fourier analysis, storing only these extracted features rather than the complete raw data. This extraction approach maintains prediction reliability by preserving key patterns while dramatically reducing storage requirements for edge devices with limited capacity.
Data Source
AI summary
A method for predicting electric energy consumption in an electric grid, which employs a linear auto-regressive model to calculate prediction data related to the electric energy consumption in an electric grid. The prediction method ensures high level performances in terms of prediction accuracy and it can be easily implemented even when limited computational and data storage resources are available.