Original Article
Adaptive decoupling control of hypersonic
vehicle using fuzzy-neural network
observer
Chen Bai, Jian Chen, Zhang Ren, Qingdong Li and Zihao Xiong
Abstract
An adaptive decoupling control approach using fuzzy-neural network (FNN) observer for a class of MIMO
nonlinear systems with parameter uncertainties is presented in this article. First, a decoupling controller is constructed
based on the decentralized control theory. Furthermore, the system coupling terms and uncertainties are
estimated by the FNN observer and added into the control law for compensation. The FNN approximate-matrix
update law and the control law guarantee that the tracking errors of the system states, the observer states and
the approximate matrix are all uniformly ultimately bounded within a region that can be kept arbitrarily small.
Secondly, a model for the hypersonic vehicle is given and an attitude controller is designed using the decoupling control
approach. Finally, simulations are carried out on the hypersonic vehicle to demonstrate the effectiveness of the proposed
method.
Keywords
Hypersonic vehicle, adaptive decoupling control, fuzzy-neural network observer, MIMO nonlinear systems, parameter
uncertainties
Date received: 4 November 2014; accepted: 18 August 2015
Introduction
Hypersonic vehicle refers to the vehicle whose
Mach number is greater than 5. It is a typical plane-
symmetric vehicle which requires high Angle Of
Attack (AOA) to obtain enough lift force. The
sideslip angle is usually constrained to zero to avoid
or attenuate aerodynamic heating effects. Lateral
maneuver motion of hypersonic vehicles is always
Bank-To-Turn (BTT), which results in that the bank
angle varies dramatically.
1,2
Owing to all of the above
factors, there exist coupling effects among motions of
the vehicle, especially between yaw motion and roll
motion. The sideslip angle might be very large because
of the coupling during the BTT motion, which
leads to the danger to the vehicle. For tracking
guidance command accurately, the flight control
system has to overcome the coupling effects.
Thus, it is necessary to study the decoupling con-
trol problem of a hypersonic vehicle which is a
Multiple Input Multiple Output (MIMO) nonlinear
system. In addition, aerodynamic parameters,
environmental parameters and body structure param-
eters have large uncertainties during the entire
flight envelope, so the designed flight control system
should have strong robustness against those
uncertainties.
3,4
Although the traditional gain-scheduled control
approach has been employed extensively on practical
projects of flight control, it needs a large amount of
calculation and cannot be used to deal with coupling
effects since it does not utilize output information
from other channels. A controller in Zeng et al.
5
has
been proposed based on nonlinear dynamic inversion
to achieve decoupling of a MIMO system, but the
approach can only be applied to the system of
which the model is inaccurate. Thus, the method in
Zeng et al.
5
cannot be used to deal with the system
with uncertainties. The controllers in Zhou et al.
6
and
Wang et al.
7
have been proposed utilizing sliding
mode control theory, but the frequent switching of
the controllers might induce high frequency oscilla-
tion which threats the vehicle’s safety. Furthermore,
the robust control theory has been applied in the lit-
eratures.
8–11
For a BTT missile, a gain-scheduled
autopilot is designed by a H-infinity loop-shaping
Proc IMechE Part G:
J Aerospace Engineering
2016, Vol. 230(7) 1216–1223
! IMechE 2015
Reprints and permissions:
sagepub.co.uk/journalsPermissions.nav
DOI: 10.1177/0954410015606165
uk.sagepub.com/jaero
School of Automation Science and Electrical Engineering, Beihang
University, Beijing, China
Corresponding author:
Jian Chen, School of Automation Science and Electrical Engineering,
Beihang University, Beijing 100191, China.
Email: chenjian@buaa.edu.cn