Monitoring Apparatus for Automatic Feature Amount Extraction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing home energy management systems face challenges in efficiently creating and storing reference data for estimating the operating states of electrical devices, as they require customized data for each household or store, which is time-consuming and costly, and existing techniques lack efficient methods for updating this data due to fluctuations in device types and numbers.
Innovation Solution
A monitoring apparatus and method that acquires time-series data of electrical devices, extracts feature amounts from waveform data, and stores them in a feature amount storage unit when specific conditions are met, allowing for the automatic generation and updating of reference data for estimating device operating states.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If feature amounts are manually measured and stored for each electrical device in each household, then estimation accuracy of device operating states is improved, but the time and cost for creating customized reference data increases significantly
Solution Approach 1:
The monitoring apparatus automatically performs measurement and storage of feature amounts without requiring manual intervention. The system self-updates the database by detecting device operation states and automatically extracting feature amounts from power consumption data, eliminating the need for technicians to visit each household for data collection
Solution Approach 2:
The system pre-stores feature amounts in the database before they are needed for estimation. By continuously monitoring and automatically storing feature amounts as devices are operated, the system prepares reference data in advance, making it readily available when estimation is required without last-minute manual collection
2Measurement precision
If dedicated power consumption measurement devices are installed in each electrical device, then power consumption of individual devices can be precisely measured, but the system cost increases due to the large number of sensors required
Solution Approach 1:
A single monitoring apparatus installed at the household level performs the function of multiple dedicated measurement devices. By monitoring total power consumption and using feature amount databases to decompose and estimate individual device contributions, one universal device replaces what would otherwise require numerous specialized sensors in each appliance
Solution Approach 2:
The monitoring apparatus acts as an intermediary between the power supply and individual devices. Instead of placing sensors directly in each device, the system measures aggregate power consumption at the household level and uses the feature amount database as a mediator to infer individual device states, avoiding the need for physical sensors in each appliance
3Measurement precision
If feature amount databases are customized for each household, then estimation accuracy is improved, but the ease of system deployment and updates deteriorates due to the need for on-site measurements
Solution Approach 1:
The system automatically creates and updates household-specific feature amount databases without requiring technician intervention. The monitoring apparatus detects device operations, extracts feature amounts, and stores them in the database autonomously, making deployment as simple as connecting power measurement equipment without scheduled on-site measurement sessions
Solution Approach 2:
The system begins building the customized database immediately upon deployment by continuously monitoring and storing feature amounts as devices are naturally operated. This preliminary automatic data collection eliminates the need for separate on-site measurement campaigns, allowing the database to be prepared in advance and updated automatically over time
Data Source
AI summary
Provided is a monitoring apparatus including an acquisition unit that acquires first time-series data which is time-series data of a measured value and/or a feature amount regarding an electrical device group, and a registration and updating unit that, when a difference in values between a pre-state-change data value which is any one of a data value at a first point in time in the first time-series data and a statistic of a plurality of data values from a point in time earlier than the first point in time by a predetermined period of time to the first point in time and a post-state-change data value which is any one of a data value at a second point in time later than the first point in time and a statistic of a plurality of data values from the second point in time to a point in time later than the second point in time by a predetermined period of time satisfies a predetermined condition, stores a feature amount extracted from any waveform data of a total current consumption, a total power consumption, and a total input voltage of the electrical device group corresponding to the pre-state-change data value and the post-state-change data value which satisfy the predetermined condition, in a feature amount storage unit.