Adjustment of Design Parameters for Improved Feedback Properties in the Linear Quadratic Regulator.

摘要

Classical analysis and design methods for single input-single output (SISO) systems, such as gain and phase margins, do not generalize easily to MIMO systems. Recently, the singular values of the return difference and inverse Nyquist matrices have proven useful in analyzing multiple input-multiple output (MIMO) systems. The linear quadratic formulation is useful for the design of MIMO controllers. A disadvantage of this design method is that all the design specifications must be incorporated into a quadratic cost functional. This thesis contains a systematic method for adjusting the quadratic cost to manipulate the singular value functionals and the feedback properties and thus achieve the design requirements. (Author)