TY - JOUR
T1 - Online parameter identification and optimal input design using perturbed nonlinear programming
AU - Subedi, S.
AU - Hosseini, B.
AU - Diepolder, J.
AU - Holzapfel, F.
N1 - Publisher Copyright:
© Published under licence by IOP Publishing Ltd.
PY - 2023
Y1 - 2023
N2 - Flight tests are performed to gather data for flight vehicle parameter estimation. If an a priori model is available, the inputs during the runtime of the flight test experiment can be designed, such that the information content in the gathered data is maximized. These optimal inputs, however, are based on an inaccurate system model, and the improvement of the system model is the ultimate goal of the system identification and therewith, the flight tests. There is an implicit relationship between the optimal inputs and the system model. In this paper, we investigate a scheme to update the optimal inputs during the flight using an online optimization technique, while performing online parameter estimation. An extended Kalman filter is used to compute new parameter values during the maneuver runtime. Concurrent to that, an interior point NLP solver computes the optimal maneuver updates, as new parameters become available in each iteration. The method is validated via a numerical example.
AB - Flight tests are performed to gather data for flight vehicle parameter estimation. If an a priori model is available, the inputs during the runtime of the flight test experiment can be designed, such that the information content in the gathered data is maximized. These optimal inputs, however, are based on an inaccurate system model, and the improvement of the system model is the ultimate goal of the system identification and therewith, the flight tests. There is an implicit relationship between the optimal inputs and the system model. In this paper, we investigate a scheme to update the optimal inputs during the flight using an online optimization technique, while performing online parameter estimation. An extended Kalman filter is used to compute new parameter values during the maneuver runtime. Concurrent to that, an interior point NLP solver computes the optimal maneuver updates, as new parameters become available in each iteration. The method is validated via a numerical example.
UR - http://www.scopus.com/inward/record.url?scp=85164254935&partnerID=8YFLogxK
U2 - 10.1088/1742-6596/2514/1/012020
DO - 10.1088/1742-6596/2514/1/012020
M3 - Conference article
AN - SCOPUS:85164254935
SN - 1742-6588
VL - 2514
JO - Journal of Physics: Conference Series
JF - Journal of Physics: Conference Series
IS - 1
M1 - 012020
T2 - 2nd International Workshop on Mathematical Modeling and Scientific Computing: Focus on Complex Processes and Systems, MMSC 2022
Y2 - 4 October 2022 through 7 October 2022
ER -