SBIR-STTR Award

This research shall focus on two improvements to the current state of modeling and shape control of large space structures: (I) the use of frequency domain-spectral factorization methods for the development of shape control laws: and (II) the use of advanced methods, including homogenization, or the reduction of complexity of the models of large space structures with regular lattice structure. The approach will be to adapt some proven algorithms of Davis and stinger for optimal gain computation by spectral factorization to a generic model of a large, flexible structure. In the second part of the work we shall assess relevance of homogenization asymptotic analysis to the simplification of models of some typical large, lattice structures planned for space applications.

An integrated software system will be developed to make the powerful analytical design techniques for control and modeling developed in Phase I useful to design engineers. This will be done by using a commercially available symbolic manipulation language in conjunction with a relational data base manager, a flexible "communications file" structure, and special purpose (fortran) programs for numerical evaluations. The system will be capable of evaluating distributed parameter models from standard structural mechanical specification, of combining elementary rigid and flexible components together to make a "hybrid" system, and of computing (symbolically and numerically) and evaluating stabilizing control laws for the hybrid system, including nonlinear laws. The system will be ported on a commercial (sun) graphics work station of modest cost and a supervisor interface will be provided to facilitate user interaction (at the symbolic, graphical, or numerical level) with the system. A demonstration of the system on the "NASA challenge system (scole)" will be carried out to assess the effectiveness of the system.