Abstract
This article addresses fixed-time passivity (FIXTP) for multi-weighted coupled quaternion-valued neural networks (CQVNNs), where the FIXTP does not require that the output and input are real-valued. By means of the quaternion-valued sign function and some related properties, several sufficient conditions are given to guarantee the FIXTP of CQVNNs. Moreover, considering that passivity is a useful tool to tackle the synchronization for multi-weighted coupled neural networks, the relationship between FIXTP and fixed-time synchronization (FIXTS) for CQVNNs is also revealed, which extends the application scope of the passivity. In addition, some FIXTS criteria for CQVNNs are derived on basis of FIXTP, which has not been done by others. Finally, a simulation example is given to check the validity of the obtained results.
| Original language | English |
|---|---|
| Article number | 127289 |
| Journal | Neurocomputing |
| Volume | 574 |
| DOIs | |
| State | Published - 14 Mar 2024 |
| Externally published | Yes |
Keywords
- Coupled quaternion-valued neural networks (CQVNNs)
- Fixed-time passivity (FIXTP)
- Fixed-time synchronization (FIXTS)
- Multi-weights
Fingerprint
Dive into the research topics of 'Fixed-time passivity of multi-weighted coupled quaternion-valued neural networks'. Together they form a unique fingerprint.Cite this
- APA
- Author
- BIBTEX
- Harvard
- Standard
- RIS
- Vancouver