Copyright © IFAC Robust Control Design Milan, Italy, 2003 ELSEVIER IFAC PUBLICATIONS www.elsevier.com/locate/ifac SOLVING WEIGHTED MIXED SENSITIVITY H _ PROBLEM BY DECENTRALIZED CONTROL FEEDBACK VIA MODIFYING WEIGHTING FUNCTIONS AND USING DESCRIPTOR SYSTEM REPRESENTATION Batool Labibi", Ali Khaki Sedigh', Parviz Jabedar Maralani", and Boris Lohmann'" K.N Toosi University of Technology, Tehran, Iran •• University ofTehran. Tehran, Iran ••• University of Bremen, Bremen, Germany labibi@eeld.knlu.ac.ir. sedigh@eetd.kntu.ac.ir, pjabedar@ut.ac.ir, bl@iat.uni-bremen.de Abstract: This paper considers the problem of achieving stability and certain H_ performance for a large-scale system by a decentralised control feedback law. For a given large-scale system an equivalent descriptor system in input-output decentralised form is defined. For solving the performance problem which is formulated as the standard weighted mixed sensitivity H _ problem, modification of the original weighting functions is proposed. Some sufficient conditions are proposed when satisfied the overall stability and performance of the large-scale system is guaranteed. Copyright © 2003 IFAC Keywords: Large-scale system, H _ control, Decentralised control, Weighted mixed sensitivity H _ problem, Weighting function, Input decentralised form, Input-output decentralised form, Descriptor system. I. INTRODUCTION The problem of designing decentralised control for large-scale interconnected systems has attracted a great amount of interest in recent years. One reason is that the interconnected system can be decomposed into several lower-order subsystems and therefore the design and implementation of each subsystem can proceed independently (Jamshidi, 1997; Siljak, 1991 ). Mixed sensItIvIty is the name given to transfer function shaping problems in which the sensitivity 19 function is shaped along with one or more other closed-loop transfer functions such as KS or the complementary sensitivity function (Maciejowski, 1989; Skogestad, 1996). In this paper a method for solving the mixed sensitivity H _ problem by a decentralised control is proposed. It is shown how by appropriately modifying the weighting functions in original mixed sensitivity problem the overall stability and performance can be achieved by a decentralised feedback control.