Robust Control for Aircraft with Reaction Jets Using Dynamic Inversion and
Fuzzy Neural Networks
Chang Yafei* Yuan Ruyi** Fan Guoliang*** Yi Jianqiang****
*Institute of Automation, Chinese Academy of Science, 100190, Beijing
(Tel: 86+13810688872; e-mail: yafei.chang@ia.ac.cn).
** Institute of Automation, Chinese Academy of Science, 100190, Beijing
(e-mail: ruyi.yuan@ia.ac.cn)
*** Institute of Automation, Chinese Academy of Science, 100190, Beijing
(e-mail: guoliang.fan@ia.ac.cn)
**** Institute of Automation, Chinese Academy of Science, 100190, Beijing
(e-mail: jianqiang.yi@ia.ac.cn)
Abstract: In this study, a robust control scheme based on dynamic inversion using fuzzy neural networks
(FNN) is designed to control a nonlinear aircraft driven by two different actuators (fins and reaction jets).
First, the dynamic model of the aircraft with reaction jets is briefly introduced. Then, the robust control
scheme is explained in detail, in which the dynamic inversion is adopted to solve the system nonlinearity
and the FNN cancels model uncertainties by on-line learning. Learning is accomplished by a simple
weight update rule derived from Lyapunov theory, thus assuring the stability of the closed loop system.
In addition, a control moment allocator is designed to coordinate the work of fins and reaction jets.
Finally, the control scheme is demonstrated in simulation. Its robustness is evaluated relative to a PID
control based on dynamic inversion; its rapid time response performance is compared with a
conventional aircraft with fins only, to certify the effectiveness of the control moment allocator.
Keywords: Aircraft with reaction jets, dynamic inversion, fuzzy neural network, Lyapunov stability,
control allocation, robustness.
1. INTRODUCTION
Aircraft with reaction jets has attracted many interests in a
wide variety of researches due to its excellent
maneuverability in the upper air. It is for the reason that
reaction jets could provide huge moment in a short time, and
will not be affected by the thin air in the upper (Barmes,
1998). However, the dynamics of an aircraft with reaction
jets at a high angle of attack is inherently nonlinear and may
vary rapidly with time. Furthermore, the dynamics is hardly
accurately obtained as a result of uncertainty of atmospheric
coefficients and atmospheric disturbances. Besides, the
system has two kinds of actuators with different control
characteristics, which should be coordinated well. These and
other concerns have prompted researchers to look for some
robust, nonlinear, and intelligent control schemes, and
coordinating strategies for the fins and reaction jets.
Some of researchers choose sliding mode control (SMC)
method (Zhao, 2009, Fan, 2008, and Weil, 1991). An
advantage of this method is its robustness to parameter
perturbations and external disturbances. However, the
discontinuous term in control input causes an undesirable
effect called chattering. High frequency switching may be
destructive for control devices, and may cause system
resonance. Dynamic inversion has been widely used to deal
with the system nonlinearity. However, it seriously depends
on the accuracy of the model. Thus, it can hardly afford the
desired robustness of the system. To compensate this
situation, some researchers combine several intelligent
methods (Michael, 2000, Chen, 2006, Xuan, 2011) such as
fuzzy logic, neural networks featuring universal
approximation with dynamic inversion. The fuzzy neural
network (FNN), as a strong modeling tool, which combine
the capability of fuzzy reasoning in handling uncertain
information with the capability of artificial neural networks
in on-line learning from process, has represented an
alternative method to deal with uncertainties of the control
system in recent years (Rahideh, 2007, Wang, 2012, and Wai,
2013). So, the paper concerns a dynamic inversion based
control scheme compensated by the FNN to deal with
nonlinearity and uncertainty of the aircraft system. Unlike
researches of Chen (2006), the control scheme in this paper
adopts two FNN compensators which can assure robustness
of the entire system. Besides, the FNN parameters are
updated via the Lyapunov stability constraints, which
guarantee the closed loop system stable.
As for coordination for fins and reaction jets, there are two
main strategies; one is command allocation (Sun, 2008 and
Zhan, 2009), and the other is control quantity allocation
(Ridgely, 2006, 2007, and Innocenti, 1996). In the command
allocation strategy, the angle of attack command is divided
into two parts; one is assigned to aerodynamic system and the
other to reaction jets system. The advantage of this strategy is
easy to design, but coupling of these two systems is not