Elahe Saeedi - Young Researchers and Elites Club Lenjan Branch, Islamic Azad University Isfahan, Iran
Bahram Karimi - Department of Electrical Engineering Malek ashtar University of Technology Shahinshahr, Isfahan,Iran
Mostafa Pourbehi - Iran Khodro Tam Instrument Expert, Isfahan, Iran

In this paper, a decentralized adaptive controller with using multilayer neural network is proposed for a non affine nonlinear aircraft cabinpressure system with unknown parameter and unknown interactions. The ultimate, bounded and the stability of the closed loop system isguaranteed by this controller. Through a lyapunov stability analysis. In addition each subsystem is able to adaptively compensate ofinteractions. Due to the functional approximation capabilities of multilayer neural networks, the dynamics for each subsystem are notrequired to be linear in a set of unknown coefficient, as is typically required in decentralized adaptive schemes. In particular, it is shownthat the tracking error converges asymptotically to zero. At the end, to show the effectiveness of the proposed a decentralized adaptivecontroller with using multilayer neural network is illustrated by considering an aircraft cabin pressure system with unknown interactionsand the simulation results verify the merits of the proposed controller for this system.

Adaptive control, aircraft cabin pressure, neural network