Structured MIMO [formula omitted] design for dual-stage inertial stabilization: Case study for HIFOO and Hinfstruct solvers

Abstract

The paper reports on application of two available numerical solvers for structured and fixed-order controller design to a realistic laboratory MIMO electromechanical system. Namely, an experimental platform consisting of two coaxial motorized gimbals that stabilize an angular motion of an optical payload around a single axis is considered. Inertial angular rate of the payload is measured using a MEMS gyro and the mechanically constrained misalignment between the two gimbals is measured with an incremental encoder. This represents a simplified testbed for the more practically useful multi-axis line-of-sight inertial stabilization systems. The paper defends the choice of the control system structure and argues that the recent Matlab-based computational MIMO control design procedures which are capable of enforcing some structural constraints upon the controller transfer function matrix—HIFOO and Hinfstruct—constitute efficient and practical design tools. They only lose little in the optimality of the full unrestricted MIMO controllers computed by standard H∞ optimization, and yet they produce controllers which are easy to implement and fine-tune within a standard motion control infrastructure based on PID feedback and feedforward terms. Numerical simulations and laboratory experiments were used to validate the findings.