Electrical Meter Mode Switching for Device-Level Energy Disaggregation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing electricity monitoring systems struggle to accurately and efficiently disaggregate electrical usage data from multiple devices to provide real-time information on individual device consumption, requiring numerous expensive devices and significant manual effort.
Innovation Solution
A system comprising a power monitor that processes electrical signals from an electrical panel to disaggregate device usage, using machine learning classifiers and models to identify device events and states, and a search graph to determine device states and consumption, providing real-time information through a user interface.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If electricity monitors are installed for individual devices, then measurement precision of device usage is improved, but device complexity and installation cost increase significantly
Solution Approach 1:
The patent applies segmentation by dividing the aggregate electrical signal into individual device components through machine learning-based disaggregation. The system segments the complex mixed signal from multiple devices into separate device-level measurements, achieving precise individual device monitoring without physically installing separate monitors on each device.
Solution Approach 2:
The patent uses machine learning models as an intermediary between the aggregate electrical signal and individual device usage information. The ML models act as a mediator that processes the mixed signal and extracts device-specific consumption patterns, enabling precise measurement without direct device-level sensors.
2Ease of operation
If electricity monitors are installed at the electrical panel to monitor multiple devices, then ease of operation is improved, but measurement precision of individual device usage deteriorates
Solution Approach 1:
The patent introduces machine learning-based disaggregation as an intermediary processing layer that operates on the aggregate signal from the electrical panel monitor. This intermediary extracts device-level information from the mixed signal, maintaining the ease of single-point installation while achieving individual device measurement precision.
Solution Approach 2:
The patent replaces the mechanical approach of installing separate physical monitors on each device with a computational approach using machine learning algorithms. The ML system substitutes for multiple physical devices by processing the aggregate signal digitally to extract individual device usage information.
3Loss of information
If traditional disaggregation techniques are used, then device usage information can be extracted, but processing speed and accuracy deteriorate
Solution Approach 1:
The patent applies parameter changes by using machine learning models that have been trained to recognize specific electrical consumption patterns, harmonics, and temporal characteristics of different devices. The ML algorithms transform the analysis parameters from simple signal averaging to multi-dimensional pattern recognition, dramatically improving both speed and accuracy of disaggregation.
Solution Approach 2:
The patent replaces traditional signal processing methods with machine learning-based analysis. The ML system substitutes conventional disaggregation techniques with neural networks and classification algorithms that process electrical signals more efficiently, achieving faster and more accurate device identification and consumption measurement.
Data Source
AI summary
A system including an electrical meter and at least one server. The electrical meter: obtains a power monitoring signal by measuring an electrical property of a power line to a building; processes the power monitoring signal in a first mode of operation to determine first information about a first device in the building; transmits the first information to the at least one server; receives from the at least one server, an instruction to change to a second mode of operation; processes the power monitoring signal in the second mode of operation to determine second information about a second device in the building; and transmits the second information to the at least one server. The at least one server: receives the first information from the electrical meter, transmits the instruction to the electrical meter to change to the second mode of operation, and receives the second information.