-
Notifications
You must be signed in to change notification settings - Fork 10
Home
This project was part of the Global Innovation Exchange, as part of TECHIN 514: Hardware/Software Lab 1. Over the course of one ten-week quarter, students built prototypes that mixed hardware (e.g. embedded systems and sensors) and software (data collection) to be used in machine learning (ML) models. This class was broken into three milestones to present progress.
It was built using the EmonPi open-source energy monitor to enable scalable disaggregation research. The experimental setup was built using a power strip as a proof of concept for a single phase of a residential house. By building this atop an extensible open-source project, we hope to share this work with the energy disaggregation community.
Will Buchanan, Louis Quicksell, Ricky Powell
Provide a proof of concept residential energy disaggregation feedback mechanism to provide a breakdown of appliance-specific consumption information in realtime, in a nonintrusive manner (e.g. no extra wiring/electrical work) at a low cost. Such a device would involve current and voltage sensors, which would then break down the unique signature of an appliance.
No low-cost device exists to inform power consumption in real-time. Existing plug-level devices (such as the KillaWatt) only measure consumption of individual appliances, and require a plug at each outlet. Nonintrusive Aggregated Load Monitoring (NIALM) algorithms have not been used to their full potential value, reducing the impact of Smart Metering deployment. There is a need for real-time disaggregation monitoring for whole-home power consumption, as studies have shown average reduction in electricity consumption up to 11% (J. Lynam, K. Nitta, T. Saijo, N. Tarui (2014). Why does real-time information reduce energy consumption?)
In the intended use case, the device would identify appliances that were plugged-in in real time. Thus the model would have to identify the appliance nearly instantaneously. The detail of the classification (not simply the type of appliance, but make, model, etc) would make for a compelling experience, however this was de-prioritized due to time constraints. Machine Learning models will be applied, and the results sent to the user.