Abstract
From a sparse set of large-scale linear time-invariant dynamical models, a methodology to generate a low-order parameter-dependent and uncertain model, with guaranteed bounds on the approximation error, is first obtained using advanced approximation and interpolation techniques. Second, the stability of the aforementioned model, represented as a linear fractional representation and subject to actuator saturation and dynamical uncertainties, is addressed through the use of an irrational multiplier-based integral quadratic constraint approach. The effectiveness of the approach is assessed on a complex set of aeroservoelastic aircraft models used in an industrial framework for control design and validation purposes.
Highlights
M ANY techniques have been developed to model, control, and assess the stability and performance of dynamical systems
An optimal frequency-limited approximation algorithm is first applied followed by the creation of a frequencydependent mismatch bound, the interpolation and transformation into a linear fractional representation (LFR) structure is achieved, and the stability of the overall uncertain parameterdependent model is first assessed thanks to a μ analysis, and when subject to control input saturation, through a novel integral quadratic constraint (IQC) technique
A methodology that enables us to assess the stability of a set of controlled SIMO large-scale linear timeinvariant (LTI) dynamical models subject to input saturation has been presented
Summary
M ANY techniques have been developed to model, control, and assess the stability and performance of dynamical systems. The problem of assessing the stability of such a high-dimensional controlled system over the continuum of parametric variations, when the single control input u(t) is subject to saturations, is addressed here To this aim, as clarified in the rest of this paper and following Fig. 1 and Algorithm 1, a three-step methodology is proposed: 1) approximate the ns dynamical models and bound the mismatch error; 2) perform (inexact) interpolation of the reducedorder models with interpolation error bounds; and 3) assess the stability of the closed-loop model over both parametric variations and control input saturation limitations.. Quadratic constraint (IQC) analysis (step 3), respectively, provide sufficient stability conditions for the whole set of closed-loop models without and with saturation This represents the main contribution of this paper.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
