The quantification of large SNR for MLE of ARARMAX models
Research output: Contribution to conference › Paper › peer-review
Electronic versions
DOI
Maximum likelihood estimation(MLE) is widely applied in system identification because it is consistent and has excellent convergence properties. However gradient based optimization of likelihood function might end up in local convergence. It is known that for ARMAX and ARARX models, providing a large enough Signal-to-Noise-Ratio(SNR) will avoid the potential local convergence. We show the same condition can be extended to ARARMAX models in this paper. To ease the application of this condition, the exact value of such SNR needs to be quantified. Here we realize the quantification by constrained optimization.
| Original language | Unknown |
|---|---|
| Pages | 5108-5113 |
| Number of pages | 6 |
| DOIs | |
| Publication status | Published - 29 Jan 2010 |
| Externally published | Yes |
| Event | Proceedings of the 48h IEEE Conference on Decision and Control (CDC) held jointly with 2009 28th Chinese Control Conference - Shanghai, China Duration: 15 Dec 2009 → 18 Dec 2009 |
Conference
| Conference | Proceedings of the 48h IEEE Conference on Decision and Control (CDC) held jointly with 2009 28th Chinese Control Conference |
|---|---|
| Country/Territory | China |
| City | Shanghai |
| Period | 15/12/09 → 18/12/09 |