Mapping of Frequency Response Performance Specifications Into Parameter Space

Abstract

A procedure is presented that enables frequency response performance specifications (e.g. magnitude, phase, sensitivity, H∞, etc.) to be mapped directly into parameter space. This procedure significantly enhances the versatility of parameter space techniques by facilitating the design and analysis for systems with both structured and unstructured uncertainty. For analysis, the procedure can be used to immediately show at which operating conditions the system provides the desired performance. For control design, the simplicity and transparency attributed to parameter space techniques are retained, while ensuring that the additional frequency domain specifications are also satisfied. This new procedure is demonstrated by designing a flutter suppression controller for NASA Langley’s Benchmark Active Controls Technology (BACT) wing flutter model. The control design example incorporates both eigenvalue specifications and frequency response magnitude specifications.