Control augmentation systems based on Incremental Nonlinear Dynamic Inversion (INDI) are able to provide high-performance nonlinear control without a holistic model. Considering an angular rate control law for a fixed-wing aircraft, only a control effectiveness (CE) model and angular acceleration measurement feedback is required. Despite enhanced robustness against parametric model mismatches due to reduced model dependency, the performance of INDI-based control laws can still vary due to inaccurate CE models. This paper confirms that longitudinal centre of gravity (CG) shifts and CE uncertainty result in varying handling qualities and stability (HQ\&S) characteristics. An adaptive solution using Least-Mean-Square (LMS) based parameter estimation is investigated to address these variations. The results demonstrate that online CE model correction result in reduced HQ\&S variation. However, it was found that some flight conditions together with adverse CG shifts could lead to violation of the time-scale separation assumption that underlies the adaptive control law design. As this assumption is inherent to the INDI control design itself as well, online CE model correction is only partly able to resolve the resulting performance variations.@en

The recently introduced twisted and coiled polymer muscle is an inexpensive and lightweight compliant actuator. Incorporation of themuscle in applications that rely on feedback creates the need for deflection and force sensing. In this paper, we explore a sensing principle that does not require any bulky or expensive additional hardware: Self-sensing via electrical impedance. To this end, we characterize the relation between electrical impedance on the one hand, and deflection, force, and temperature on the other hand for the Joule-heated version of this muscle. Investigation of the theoretical relations provides potential fit functions that are verified experimentally. Using these fit functions results in an average estimation error of 0.8%, 7.6%, and 0.5%, respectively, for estimating deflection, force, and temperature. This indicates the suitability of this self-sensing principle in the Joule-heated twisted and coiled polymer muscle.

@en