%A I. Abbasi
%A S.S.A. Ali
%A R. Ibrahim
%A S.H. Adil
%A M. Ovinis
%I Institute of Electrical and Electronics Engineers Inc.
%T U-model based depth control of underwater glider
%J 2015 International Conference on Information and Communication Technologies, ICICT 2015
%L scholars7040
%O cited By 1; Conference of International Conference on Information and Communication Technologies, ICICT 2015 ; Conference Date: 12 December 2015 Through 13 December 2015; Conference Code:121662
%R 10.1109/ICICT.2015.7469492
%D 2016
%K Adaptive control systems; Controllers; Intelligent control; Inverse problems; Model predictive control; Newton-Raphson method; Robotics; Stabilization, Control techniques; Controller implementation; Internal model control; Polynomial expression; RBFNN; Reject disturbances; U-model; Underwater robotics, Aircraft control
%X Controlling underwater glider brings a unique challenge for control engineers and academic researcher. The dynamics, nonlinearity and multivariable nature of glider plus the highly turbulent underwater disturbances has disregarded many effective control techniques. The Internal Model Control (IMC) structure, where the controller implementation includes explicit model of the plant has shown to be very effective for control of stable plants in process industries. The inherent capabilities of IMC to perform robustly and reject disturbances make it attractive to be used in applications like underwater robotics. In this paper adaptive IMC based on U-model is investigated for controlling depth of underwater Glider. U-model is an adaptive modeling framework that models system based on current control term only. This implies that the control law can be synthesized simplistically using Internal Model Control (IMC). Hence inverse of polynomial expression based on u(t-1) is computed using Newton-Raphson method. The effectiveness of U-model based IMC is illustrated with aid of simulation for depth control of glider. Furthermore, performance of proposed controller is compared with PID. Results show that U-model methodology performs better than PID in terms of settling time. © 2015 IEEE.