An Overview of Modeling and Control of Large Flexible Space Structures

Abstract

This paper describes the latest trends and theoretical developments involved with the modeling and control of Large Flexible Space Structures (LFSS). The paper addresses first the basic problems, characteristics and difficulties inherent in modeling and control of LFSS. Major source of error is the stiffness and damping operators of the dynamic model. Extensions of Linear Quadratic Gaussian (LQGI theory as applied to LFSS are presented, including frequency-shaped cost functionals and perturbation methods. The minimum data/maximum entropy approach which uses a stochastic design model to overcome difficulties found in LQGbased methods is described. Latest trends in system theory including Hankel-Norm model reduction, Balanced Realization, and Singular-Value Analysis are used to determine reducedorder controllers and models. Ad hoc methods such as component cost analysis and modal cost analysis are discussed in context with the closed-loop reduction problem of controller order versus performance. Another method described earlier as the minimum data/maximum entropy approach also addresses controller order versus performance. The paper also attempts to identify those areas of control science and large scale systems that appear to have an important role in understanding and solving LFSS modeling aud control.