MULTIVARIABLE CONTROLLER FOR REDUNDANTLY ACTUATED SYSTEMS - A NUMERICAL DESIGN APPROACH FOR OPEN-LOOP DATA

A numerical approach for the synthesis of multivariable controllers from plant input-output data is presented, The method is applicable whenever a system is open-loop stable, redundantly actuated, and when its transfer function matrix is of full rank for all s. The method has two major advantages. First, model-matching design is achieved without explicit open-loop zero cancellation. Second, explicit knowledge of an open-loop model is not needed for the design.