Energy-based comparative analysis of optimal active control schemes for clustered tensegrity structures

Abstract

This study performs a series of numerical investigations of a novel energy-based control approach for effective vibration control of clustered tensegrity structures via different optimal active control algorithms. The comparative study among different control algorithms of clustered tensegrities are often challenging due to the geometrical non-linearity, complex loading conditions and assemblage uncertainties of structural components. In order to overcome these technical difficulties, an actuator input energy-based method is herein implemented to assess the optimal dynamic performances of clustered tensegrity structures via distinct optimal active control schemes. As a quantification tool, the structural displacement and elemental forces monitored from both the whole structure level and the elemental level were applied to assess control efficiency based on the same amount of actuator energy input. Specifically, the control efficiency comparisons are realized by setting identical energy input to actuated elements via linear-quadratic-Gaussian (LQG) and  ∞ algorithms. Different actuator placements of clustered cables and struts are considered and the control efficiency coefficients of the proposed method are examined through a spatial clustered tensegrity beam. The outcomes from the illustrative example indicate that the proposed method is efficient and reliable in comparative analyzing of different optimal active control schemes for clustered tensegrity structures, which implies the prospect of the investigated approach in analyzing and solving actual engineering problems. KEYWORDS active control algorithms, actuator placements, clustered tensegrity structures, control efficiency, energy-based control Struct Control Health Monit. 2018;25:e2215.wileyonlinelibrary.com/journal/stc

On the noncollocated control of structures with optimal static output feedback: Initial conditions dependence, sensors placement, and sensitivity analysis

“…18 Finally, LQG control is sensitive to the spillover phenomenon. 19 Given these considerations, the technique presented in this paper is different from the ones that considered dynamic output feedback, for example, Oliveira and Geromel 20 and Feng et al 21 These problems motivated the study of a control technique that still consists on minimizing the quadratic cost function but with the constraint of using only linear combinations of measured signals for feedback, 22,23 which is known as optimal static output feedback (OSOF) or partial state feedback. Despite simplification of hardware, this technique posed some theoretical challenges: The optimization problem is nonconvex, the optimal gain is dependent on system initial conditions, and the existence of a stabilizing static output gain is an open problem in control theory.…”

“…Finally, LQG control is sensitive to the spillover phenomenon . Given these considerations, the technique presented in this paper is different from the ones that considered dynamic output feedback, for example, Oliveira and Geromel and Feng et al …”

Deployment of a foldable tensegrity‐membrane system via configuration transitions using linear parameter‐varying control

“…In order to alleviate oscillations in system responses during deployment, closed‐loop controllers are required to control the dynamics of foldable tensegrity‐membrane systems. Note that robust linear controllers such as LQR controllers 25 and controllers 14,26,27 can be used to conduct active vibration control for flexible structural systems (e.g., tensegrity systems and tensegrity‐membrane systems) by regulating system dynamics around a given equilibrium. Since a foldable tensegrity‐membrane system may move along a trajectory that covers a large region within system equilibrium envelope during deployment, a simple, linear controller could not provide good control performance.…”

Data‐driven optimal predictive control of seismic induced vibrations in frame structures

“…However, this approach requires measurements of all the state variables, which in real applications is often impossible. Under certain hypotheses, Linear Quadratic Gaussian (LQG) control can be exploited, which allows to estimate the internal (unmeasured) variables through an observer, for example, a Kalman Filter. Nevertheless, also, the LQG has some drawbacks due to time delay between input and output, high sensitivity to spillover phenomenon, and no guarantees on stability in terms of robustness.…”