Machine learning application to predictive energy management
US11334035B2 · kind B2 · utility
Assignee
Inventors
Key dates
| Filing date | Dec 3, 2020 |
| Grant date | May 17, 2022 |
| Priority date | — |
| Expiry date | Dec 3, 2040 |
Classification
- Technology area (CPC G)Physics
- CPC primaryG05B2219/2639
- WIPO fieldControl
- WIPO sectorInstruments
Abstract
A system for automatically learning and adapting to the energy usage of an equipment operating according to a control input including at least one sensor for measuring an energy usage of the equipment an generating a baseline energy usage over time signature that is used to compare active energy usage measurements to so as to determine operational deviations. The system includes software that matches and compares equipment operation to established norms and can modify the functioning of the equipment when threshold deviations are detected. The system includes the ability to learn the functioning of the equipment and can adjust for dynamically changing conditions to avoid generation of false alerts or alarms while at the same time detecting longer term deviations that if left unchecked, could shorten the lifespan of the equipment and increase the costs associated with running the equipment.
Source: USPTO / EPO open patent data. Objective bibliographic and citation counts.