H2-Optimal Control with Generalized State-Space Models for use in Control-Structure Optimization

Control-Structure optimization is concerned with the design of closed-loop structure control systems where the cost associated with an optimal design reflects structure costs (e.g., mass) as well as costs associated with the closed-loop performance and control effort. The use of generalized state-space models in these optimization problems avoids the need to explicitly invert the associated mass matrix and preserves the sparsity in the system model, providing ing some promise for efficient, reliable numerical algorithms. In this paper formulae are presented for computation of the control cost and its gradient with respect to the plant parameters.