Neuro-fuzzy network architecture with on-line learning capabilities and corresponding control method
US6766310B1 · kind B1 · utility
Assignee
Inventors
Key dates
| Filing date | Nov 21, 2000 |
| Grant date | Jul 20, 2004 |
| Priority date | — |
| Expiry date | Jan 17, 2022 |
Classification
- Technology area (CPC G)Physics
- CPC primaryG06N3/043
- WIPO fieldComputer technology
- WIPO sectorElectrical engineering
Abstract
A neuro-fuzzy integrated architecture which permits on-line self-training includes at least one microcontroller of the fuzzy type dedicated to fuzzy rules computing and integrated monolithically on a semiconductor together with a non-volatile memory. Also included within the same integrated circuit are a microprocessor, a volatile memory unit, and an arbiter block linked to a bus interconnecting the fuzzy microcontroller, the microprocessor, and the volatile memory unit. The arbiter block controls access to the memory unit by the microprocessor or the fuzzy microcontroller. An additional fuzzy co-processor may be connected between the fuzzy microcontroller and the microprocessor for performing the fuzzy logic operations.
Source: USPTO / EPO open patent data. Objective bibliographic and citation counts.