| Autor: |
Menicucci, David, Caudell, Thomas P., Menicucci, Anthony |
| Předmět: |
|
| Zdroj: |
SOLAR Conference Proceedings; 2013, p1-8, 8p |
| Abstrakt: |
As a result of burgeoning installations of solar hot water systems (SHW), utilities have become concerned about the reliability of these systems and their real impact to the grid. If the systems fail or underperform, the utility must supply the shortfall of energy. Determining this information has been problematic for various reasons, one of which is the cost of field monitoring. A new idea has surfaced to use Adaptive Resonance Theory (ART) neural networks to analyze hourly whole-house energy data that are being collected by Smart Meters. ART is a computerized selforganizing learning network that mimics processing in biological neural systems. We applied ART to whole-house records of energy use that were recorded by Smart Meters. The results show that ART clearly identifies the date of installation of a SHW system. Similarly, ART was able to identify a simulated failure. Using these techniques we can measure the actual energy and power demand reduction from SHW systems without monitoring them directly. [ABSTRACT FROM AUTHOR] |
| Databáze: |
Complementary Index |
| Externí odkaz: |
|
