Robustness Qualification and Improvement of Multivariable Regulators, with Application to Airplane Control

摘要

The robustness of a dynamic system multivariable control law with respect to external disturbances, structural perturbations due to modeling and linearization errors, and dynamic perturbations in the feedback loop was analyzed. This proportional-plus-integral type control law minimizes the mathematical expectation of a quadratic performance index. Stability margins (especially the phase margin) of linear quadratic regulators are generalized. An algorithm that adjusts the cost penalty matrices iteratively to increase the system robustness is provided.