Application Classification
- michael.stuber
- Jonah Petri
- richard.lam
The descriptions below represent a continuum. Applications may shift along the continuum under certain conditions. For example, a monitoring application may be low intensity much of the time, until an event occurs, at which point it becomes medium intensity for the duration of the event.
Low Intensity
Runs in regularly scheduled intervals
Mainly passive, no significant state
Minimal Fixed cost (e.g. low memory footprint, minimal CPU consumption)\
Medium Intensity
Runs based on triggers
Control Applications
Continuous Analysis of high resolution data (e.g. 1 second data)
Waveform capture based on basic triggers
High Intensity
Runs in Real Time
Large memory or processing power requirements
Continuous Analysis of high resolution data (e.g. 128 samples/cycle, up to 1 MHZ)
Known Desired Applications:
Application | Intensity | Requires Real Time? | Resource Requirements | Comments |
|---|---|---|---|---|
High bill alarm | Low | No |
|
|
Data Reporting | Low | No |
|
|
Daily Usage Summary (Disaggregation Lite) | Low | No |
|
|
Outage Reporting | Low* | Yes |
| *Low intensity but requires real time monitoring |
Enhanced Outage Analysis (Fault Location) |
|
|
|
|
Transformer Loading |
|
|
|
|
Transactive Energy | Low (unless more analysis needed) |
|
|
|
Power Disaggregation |
|
|
|
|
EV Detection |
|
|
|
|
Control Application | Medium | Trigger |
|
|
Smart Inverter Interfacing |
|
|
|
|
EV Charging Controls (avoids Panel Upgrades / Transformer Overloads) |
|
|
|
|
Demand Response Protocol Gateway (communicate with in-home devices to have them modulate usage) |
|
|
|
|
Grid Connectivity Model Detection |
|
|
|
|
Fault Identification (Loose Neutral, Incipient Fault etc.) | Medium | Trigger |
|
|
Waveform Capture | Medium | Trigger |
|
|
Continuous Waveform Analysis | High | Yes |
|
|
Machine Learning | High | No |
|
|