While lightweight, flexible materials in aircraft have favorable qualities, they can also lead to stability issues requiring control intervention to prevent flutter and mitigate the influence of gust disturbance. This study considers active aerostructural control using parametric data-driven reduced-order models (ROM). The observer-based linear quadratic Gaussian (LQG) control scheme is used along with an optimal sensor placement (OSP) method based on the genetic algorithm (GA). A common issue with parametric ROMs is poor state consistency and interpolatability. Therefore, the State-Consistency Enforcement AutoRegressive with eXogenous inputs (SCE-ARX) method is used to ensure ROMs and control gains can be interpolated between flight conditions, hence accommodating the entire flight parameter space. This study also aims to investigate the influence of controller interpolation on optimal sensor placement. Thus, the OSP algorithm will be carried out for multiple test cases with interpolated and noninterpolated ROMs. Results show that similar sensor locations and controller performance across a broad parameter space are achieved regardless of ROM interpolation. This shows that the SCE-ARX allows for efficient control across a broad flight envelope and development of ROM and controller database with significantly low effort.
Optimal Sensor Placement and Aerostructural Control Using State-Consistent Reduced-Order Models
AIAA Journal ; 1-17
01.06.2025
Aufsatz (Zeitschrift)
Elektronische Ressource
Englisch
Reduced-Order Model with an Artificial Neural Network for Aerostructural Design Optimization
Online Contents | 2013
|