A H/μ solution for microvibration mitigation in satellites: A case study

Preda, Valentin; Cieslak, Jérôme; Henry, David; Bennani, Samir; Falcoz, Alexandre

doi:10.1016/j.jsv.2017.03.015

Cited by 13 publications

(21 citation statements)

References 42 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In order to account for both sensors and actuators, the structural model G (Ω) given by needs to be augmented with the model y = N y s − n of the sensing unit, which is composed by four tri‐axis force cells placed underneath the passive isolators and connected to the base structure. The model N is given into a LFT form so that

boldN = F_{u} false(bold {boldN}^{bold+}, Δ_{n} false)

; see the work of Preda et al for the complete description of N + and Δ n . This model captures the influence of sensor placement uncertainty and measurement noise n ; the LFT model

boldH = F_{u} false(H_{pma}, Δ_{pma} false)

of the actuators, which is composed by six PMAs mounted on top of the active plate on set of three cubes with three actuators perpendicular to the plate and three in the tangential direction (see Figure for an illustration).…”

Section: System Modelingmentioning

confidence: 99%

“…This model captures the influence of sensor placement uncertainty and measurement noise n ; the LFT model

boldH = F_{u} false(H_{pma}, Δ_{pma} false)

F_{u} false(H_{pma}, Δ_{pma} false)

.…”

Section: System Modelingmentioning

confidence: 99%

“…The system

\overset{K}{true} false(normalΩ false)

of order n K forms the LPV part of the overall controller K (Ω) and the size n Ω of the scheduling block

normalΩ {boldI}_{n_{Ω}}

determines the maximum rank of the matrices A 1 , B 1 , C 1 , D 1 in the affine description (36). The fixed matrix

K_{i n} = X_{y} N^{+} \in R^{2 \times 12}

has the role of a sensor fusion block containing the pseudoinverse

N^{+} \in R^{6 \times 12}

of the nominal sensor model N (see the work of Preda et al for details) and the selection matrix

X_{y} = ([], \begin{matrix} centerarray \\ array {bold0}_{2 x 3} & array {boldI}_{2} & array {bold0}_{2 x 1} \end{matrix})

. The reduced vector

y_{k} = K_{i n} boldy \in R^{2}

will therefore represent, in the absence of sensor noise and placement uncertainty, a reconstruction of the torques

([], \begin{matrix} centerarray \\ array T_{s u} & array T_{s v} \end{matrix})

in the vector y s given in .…”

Section: Performance Objectives and Controller Designmentioning

confidence: 99%

“…The goal, which is to establish stability and performance, bounds to drive the closed‐loop system to the limits of achievable performance. This paper builds upon the previous work reported in other works and introduces the following key contributions in modeling, performance specification, robust controller synthesis, and performance analysis: a new disturbance model that is suitable for

H_{\infty}

synthesis/analysis for the perturbations produced by a reaction wheel spinning at a particular rate; the uncertain and parameter varying model that is derived in this paper contains the dynamics of the satellite with its flexible attachment, the ACS, the microvibration mitigation unit including the reaction wheel. The resulting end‐to‐end model considers the coupling between all these various subcomponents; the control objectives are defined both in terms of pointing performance of the optical instrument and are guaranteed for the particular mission profile; a procedure is introduced to synthesize a low‐order robust linear parameter varying (LPV) controller for different levels of system uncertainty and the closed loop is afterwards analyzed for time‐varying parameters, saturations and time‐varying delays using μ and IQC analysis. …”

Section: Introductionmentioning

confidence: 99%

“…In line with these goals, the aim of this paper is to design a microvibration mitigation system for generic observation satellites equipped by flexible appendages, reaction wheels and an ACS, with the main objective to improve the pointing performance of the optical instrument, the final goal being the improvement of the image quality. The proposed microvibrations strategy is based on the previous work reported in other works . The isolation architecture relies on passive elastomer isolators to mitigate high‐frequency harmonics while a complementary active control strategy targets the low‐frequency perturbations.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Robust microvibration mitigation and pointing performance analysis for high stability spacecraft

Preda

Cieslak

Henry

et al. 2018

Intl J Robust & Nonlinear

Self Cite

View full text Add to dashboard Cite

Summary This paper deals with the development of a mixed active‐passive microvibration mitigation solution capable of attenuating the transmitted vibrations generated by reaction wheels to a satellite structure. A dedicated simulation environment, provided by the European Space Agency and Airbus Defence and Space industries, serves as a support for testing the proposed solution at satellite level. This paper covers modeling, control system design, and worst‐case analysis for a typical satellite observation mission that requires high pointing stability. Combined with a novel disturbance model for the reaction wheel perturbations, the pointing performance and stability requirements are reformulated as bounds on the worst‐case L2 system gains. Subsequently, the active microvibration controller is tuned to manage the conflicting design goals and optimize different trade‐offs between robustness and performance. Finally, robust stability margins and worst‐case performance bounds with respect to various system uncertainties, time‐varying reaction wheel spin rates, actuator saturation, and time delays are obtained using the structured singular value, integral quadratic constraints, and time‐domain nonlinear simulations.

show abstract

boldN = F_{u} false(bold {boldN}^{bold+}, Δ_{n} false)

; see the work of Preda et al for the complete description of N + and Δ n . This model captures the influence of sensor placement uncertainty and measurement noise n ; the LFT model