Abstract
A robust system identification algorithm is presented which makes use of linear system identification algorithms, such as Eigensystem Realization Algorithm, Observer/Kalman Identification, etc, and an orthogonal polynomial-based artificial neural network. Adaptive learning laws are derived by a thorough Lyapunov analysis to adjust different parameters of the neural network based model. The learning algorithm proposed in this paper is inspired by recent developments in adaptive control. The algorithm presented here is validated by analysis and simulation of examples based mainly on space applications. A detailed comparative study is performed to show the performance of the proposed algorithm with respect to some existing identification algorithms, specifically the Eigensystem Realization Algorithm.
Original language | English (US) |
---|---|
Article number | AAS 05-163 |
Pages (from-to) | 983-1002 |
Number of pages | 20 |
Journal | Advances in the Astronautical Sciences |
Volume | 120 |
Issue number | I |
State | Published - 2005 |
Event | AAS/AIAA Space Flight Mechaics Meeting - Spaceflight Mechanics 2005 - Copper Mountain, CO, United States Duration: Jan 23 2005 → Jan 27 2005 |
All Science Journal Classification (ASJC) codes
- Aerospace Engineering
- Space and Planetary Science