Derivation of uncertainty models of multivariable plants as well as problems arising in the determination of robust controllers on the basis of such models are considered. A new approach to the determination of probabilistic uncertainty bounds for multivariable plants based on black box identification is proposed. These bounds are implemented in an uncertain model in the form of upper Linear Fractional Transformation. An example for the design of robust controller of a two input/two output system using such a model is presented. Appropriate software in MATLAB®/Simulink® environment is developed which is embedded in a digital processor. Results from the simulation of the closed-loop system and experimental results obtained by using the robust controller are given.
Key words: identification of multivariable plants, robust control, embedded control systems
Topic: ENGINEERING SCIENCES