ROBUST MULTIPLE MODEL ADAPTIVE CONTROL WITH FUZZY POSTERIOR PROBABILITY COMBINATION
Fatemeh Zare-Mirakabad and Mohammad H. Kazemi
Keywords
Multiple model adaptive control, adaptive control, robust control,fuzzy control
Abstract
This paper proposes a fuzzy posterior probability (FPP) combina-
tion in a robust multiple model adaptive control (MMAC) frame-
work, for linear systems subject to uncertain real parameters, either
constant or slowly time varying. The proposed control scheme sub-
stituted posterior probability evaluator in robust MMAC architec-
ture for FPP. A Takagi–Sugeno–Kang-type fuzzy rule-based system
is proposed to generate the weights for probabilistic weighting of
the local controls to form the global signal control. The local
controls are designed using robust mixed-µ synthesis of the plant
evaluated in a set of predefined values of uncertain parameters so
that the local stability and performance robustness are guaranteed.
Local Kalman filters are also designed to produce residual signals
which are utilized by the FPP. The proposed scheme is applied to
controller synthesis of a two-cart mass–spring–damper system. The
simulation results illustrate the advantages of the proposed FPP
against its conventional evaluator.
Important Links:
Go Back