HAL Id: hal-02549682
https://hal-enac.archives-ouvertes.fr/hal-02549682
Submitted on 5 Jan 2021
HAL is a multi-disciplinary open access
archive for the deposit and dissemination of sci-
entic research documents, whether they are pub-
lished or not. The documents may come from
teaching and research institutions in France or
abroad, or from public or private research centers.
L’archive ouverte pluridisciplinaire HAL, est
destinée au dépôt et à la diusion de documents
scientiques de niveau recherche, publiés ou non,
émanant des établissements d’enseignement et de
recherche français ou étrangers, des laboratoires
publics ou privés.
Towards a unied model-free control architecture for
tailsitter micro air vehicles: Flight simulation analysis
and experimental ights
Jacson Miguel Olszanecki Barth, Jean-Philippe Condomines, Murat Bronz,
Gautier Hattenberger, Jean-Marc Moschetta, Cédric Join, Michel Fliess
To cite this version:
Jacson Miguel Olszanecki Barth, Jean-Philippe Condomines, Murat Bronz, Gautier Hattenberger,
Jean-Marc Moschetta, et al.. Towards a unied model-free control architecture for tailsitter micro air
vehicles: Flight simulation analysis and experimental ights. AIAA Scitech 2020 Forum, Jan 2020,
Orlando, United States. �10.2514/6.2020-2075�. �hal-02549682�
Towards a Unified Model-Free Control Architecture for Tail
Sitter Micro Air Vehicles: Flight Simulation
Analysis and Experimental Flights
Jacson Miguel Olszanecki Barth
∗
, Jean-Philippe Condomines
†
, Murat Bronz
‡
, Gautier Hattenberger
§
ENAC, Université de Toulouse, 31055, France
Jean-Marc Moschetta
¶
Institut Supérieur de l’Aéronautique et de l’Espace, Toulouse, 31400, France
Cédric Join
‖
Université de Lorraine, Vandœuvre-lès-Nancy, 54506, France
Michel Fliess
∗∗
École polytechnique, Palaiseau, 91128, France
Hybrid Micro Air Vehicles (MAVs) combine the beneficial features of rotorcrafts with
fixed-wing configurations providing a complete flight envelope that includes vertical take-off,
hovering, transitioning flights, forward flight and vertical landing. Tail sitter MAVs belong to
a particular class of hybrid MAVs and its peculiar issue is related to the transitioning flight
phase where, for high incidence angles, fast changing of aerodynamic forces and moments
are observed which are difficult to model and control accurately. To overcome this issue, we
propose a control architecture with model-free control algorithms that is able to stabilize the
hybrid MAV’s attitude, velocity, and position without any modeling process. The proposed
control architecture consists basically on two steps : 1) The attitude control, in order to ensure
the hybrid MAV’s attitude orientation and stability during the entire flight envelope; 2) The
guidance system responsible to control both velocity and position. We validate the MFC archi-
tecture according to a comprehensive set of flight simulations and real flight experiments. Real
flight experiments shown an effective and promising control strategy solving the principal issue
of hybrid MAVs that is the formulation of accurate hybrid MAV’s dynamic equations to design
control laws. The obtained results provide a straightforward way to validate the methodolog-
ical principles presented in this article as well as certify the designed MFC parameters, and
establish a conclusion regarding MFC benefits in both theoretical and practical contexts.
I. Introduction
O
ver
the last decades, aerospace engineers have contributed to the design of different Micro Air Vehicle (MAV)
configurations proposing innovative solutions for complex flight missions in outdoor or indoor environments.
Recent advances in embedded systems which include sensors miniaturization and faster microprocessors allowing high
frequency processes for on board computing operations brought together high flight performance demands which imply,
for each flight mission, the assignment of an appropriated MAV configuration. For long endurance flight missions,
the use of fixed-wing configurations is suitable due to their optimized aerodynamic surfaces that, in contact with
mass of air in mouvement, generate lift force relieving the energy consumption. On the other hand, in terms of flight
∗
Ph.D. Candidate, ENAC, Université de Toulouse, e-mail:jacson-miguel.olszanecki-barth@enac.fr; AIAA Student Member
†
Assistant Professor, ENAC, Université de Toulouse, e-mail:jean-philippe.condomines@enac.fr
‡
Assistant Professor, ENAC, Université de Toulouse, Drones Research Group, e-mail:murat.bronz@enac.fr
§
Assistant Professor, ENAC, Université de Toulouse, e-mail:gautier.hattenberger@enac.fr
¶
Professor, ISAE-SUPAERO, Department of Aerodynamics, Energetics and Propulsion, e-mail:jean-marc.moschetta@isae-supaero.fr
‖
Professor, Université de Lorraine, CRAN (CNRS, UMR 7039) & AL.I.E.N (ALgèbre pour Identification Estimation Numérique), 54330
Vézelise, France, e-mail:cedric.join@univ-lorraine.fr
∗∗
Professor, École polytechnique, LIX (CNRS, UMR 7161) & AL.I.E.N (ALgèbre pour Identification Estimation Numérique), 54330 Vézelise,
France, e-mail:Michel.Fliess@polytechnique.edu
1
1
2
3
4
5
W
Fig. 1 Typical flight modes of Tail Sitter Micro Air Vehicles: 1 - Vertical take-off; 2 - Transitioning flight;
3 - Forward flight; 4 - Hovering flight; 5 - Vertical landing. The vector W represents the wind disturbances.
maneuverability, rotorcraft are preferred due to their hovering flight capabilities that enable vertical take-off and landing,
as well as stationary flights. However, their energetically expensive propulsion system is not viable for long endurance
flights. For missions that demand the combination of endurance and maneuverability features, structural aerodynamic
engineers developed the so-called hybrid MAVs that operate over a wide flight envelope including vertical take-off,
efficient forward flight, transitioning flights, hovering, and vertical landing according to Fig. 1. While these different
flight aptitudes enlarge its application range, aerodynamic optimization of the MAV cell must be led by aerodynamic
designers considering the challenges of each flight domain. Furthermore, the autopilot system must ensure the stability
and the tracking trajectories for the entire flight envelope considering the particularities of each flight domain and
also the interactions between them which results a higher degree of challenge and complexity also for the guidance,
navigation, and control community. Different hybrid MAV configurations can be found in the literature, such as tilt
rotors [
1
] or tilt wings [
2
], quadplanes [
3
], and tilt body or tail sitter [
4
]. These platforms have been designed in order to
solve the aerodynamic and mechanical limitations of each of them and the choice of the appropriated MAV configuration
varies according to the imposed flight mission specifications. For instance, maximum payload, the desired endurance,
the range and the inherent stability against the windy environment. Generally, tail sitter MAVs are designed and
optimized to perform an efficient forward flight, since this flight phase represents most of its mission. Various studies
have improved and assessed the aerodynamic properties of MAVs previously [
5
] [
6
]. However, the flap effectiveness
needs to be optimized in order to create sufficient pitch moment ensuring the control authority during the transitioning
flights. We focalize our research work in the design and control of tail sitter MAVs investigating the performance of this
peculiar MAV class for three reasons : 1) Tail sitter are more enduring in forward flight when we compare to others; 2)
The simple transition mechanisms, in relation to tilt rotors that need additional actuators to orientate the propeller in
order to perform the transitioning flight; 3) The challenge of attitude stabilization during hovering and transitioning
flights in windy conditions. Tail sitter are susceptible to wind disturbances during these flight phases, its stabilization
remains an attractive, motivating and challenging control research topic. Typically, its entire flight envelope can be
analysed in three distinct flight modes, namely, hovering flight, forward flight and transitioning flight. While hovering
and forward flights were well researched and can be studied using a linearized system around an equilibrium point
facilitating the impementation of classical linear control algorithms. The transition flight possesses some peculiarities
that includes fast changing of aerodynamic forces and moments with wing behaviours partially stalled. Based on such
aerodynamic effects, a reliable model that accurately represent the flight dynamics of a tail sitter MAVs over their
entire flight envelope remains expensive, time consuming and a difficult task. Due to these practical modeling issues,
some research works considered the transition flight as an undesirable and transient flight phase. However, we can
not neglect the fact that, transitioning phase needs to be continuously stabilized in order to ensure a smooth and safe
flight, especially against wind disturbances. Tail sitter MAV model is often considered by the control community as a
parameter-varying system, e.g. the change of aerodynamic coefficients according to its attitude orientation and the
environmental wind conditions. Consequently, design a control technique for autopilot systems that does not rely on
2
prior knowledge of tail sitter MAV model becomes an intuitive, inovative and, from the point of view of the authors, an
appropriate control methodology. Therefore, the development of a such controller that estimates the tail sitter MAV
dynamics and counteracts it, in real time, can be easily adaptable and implemented for different hybrid MAVs.
A. Control literature review
Different control strategies have been designed for hybrid MAVs, we present some of them in the following with
particular emphasis in the controllers developed for the tail-sitter class. For practical reasons, classical linear controllers
designed using PID techniques have been used in some works [
7
] [
8
]. Although simple to tune without the knowledge of
the controlled system dynamics, PID controllers are known by the lack of robustness against significant wind disturbances.
Autopilot systems develop from optimal control theory, have been researched. For instance, the Linear Quadratic
Regulator (LQR) which was designed and applied for a tail-sitter MAV previously modelized and identified from wind
tunnel campaign [
9
]. The performance of model-based controllers differs primarily in the fidelity with which the plant
is modeled and the accuracy of the identified model parameters. Hence, classical model-based control techniques seem
to be neither optimal for hybrid MAVs nor easily transposable for a new platform. Gain scheduling methods employing
different control algorithms with both linear [
10
] and nonlinear approaches [
11
], have been developed to stabilize
hybrid MAVs at different pitch angle orientations within the transitioning flight. Gain scheduling techniques allow easy
understanding and simple implementation of the control gains that cover the entire flight envelope of hybrid MAVs.
However, the principal disadvantage of this control method, found in the literature [
12
], is the expensive computational
cost for operations in real time. In the same way, an attitude controller based on optimal control algorithms was proposed
by [
13
], different control solutions for a set of attitude errors were precomputed and stored in a lookup table. According
to the current flight conditions and for each autopilot system update, the desired control informations are obtained by
reading the predefined values from the table. Experimental flights proven that this control approach enable the hybrid
MAV to recover from a significant range of attitude errors. Further analysis is needed to determine if this proposed
control strategy can be effective and easily adaptable for different hybrid MAVs. Adaptive control techniques which
account for uncertainties present in the hybrid MAV model were developed by [
14
] [
15
]. However, instability problems
with adaptive control methods can still exist with regard to unmodeled dynamics or inaccurate models used in the
adaptation criterion of controller’s gains. Different research topics applying nonlinear control techniques on hybrid
MAVs, such as backstepping [
16
], NDI [
17
] and INDI [
18
], appears to be positively researched in the literature. The
INDI approach, which is a control that depends less on the model, was experimentally flight tested providing excellent
performance against wind disturbances. This controller requires the identification of the system actuator behaviour in
order to estimate its control effectiveness. As the actuators effectiveness vary according to the flight phase, e.g hovering
or forward flight, a gain scheduling method was implemented to fit the actuator effectiveness under the respective flight
domain. Some theoretical research analyzed the performance of nonlinear feedback control on axisymmetric aerial
vehicles [
19
] proposing an extended control solution to a larger set of generic aerodynamic models [
20
] which could
include hybrid MAVs. Additionally, a variety of nonlinear control strategies based on Lyapunov’s stability concepts
have been designed to hybrid MAVs [4] [21].
B. Links with Model-Free Control algorithm
Although most of the controls described in the literature, are designed according to a modelling process, we can
mention some particular techniques where the controller does not rely on modelling. For instance, the model-free
control approach proposed by [
22
] that have been successfully used in different concrete case-studies varying from
wastewater denitrification [
23
], nanopositioning of piezoelectric systems [
24
] up to inflammation resolution in biomedical
applications [
25
], see also its references for additional case-study examples and supplementary information. Some
research works based on model-free control techniques has been led to patents, such as [
26
] [
27
]. In the aerospace
field, this control approach has been little applied [
28
] [
29
] and, except for our previous work, it has never been applied
on hybrid MAVs which is an additional motivation for the development of our research project. The advantage of
the control methodology proposed in this paper is the capability to estimate the hybrid MAV dynamics, without a
prior knowledge of its parameters, only from its output and input-control signal measurements. Thus, any disturbance
that may affect the flight performance are measured and the MFC algorithms are able to estimate and counteract the
undesirable dynamics in order to continuously stabilize the hybrid MAV for arbitrary attitude orientations.
3
II. Related Work
Hybrid MAVs have now reached a level of maturity such that the problem of improving their autonomous flight
capabilities is now becoming a major concern. While the different flight aptitudes of hybrid MAVs enlarge their
application range, autopilot systems must ensure the stability and the tracking trajectories of all flight domains which
results in a higher degree of challenge and complexity for the guidance, navigation, and control community. Often, the
control community considers the hybrid MAV model as a parameter-varying system, e.g. the change of aerodynamic
coefficients according to the hybrid MAV attitude and the environmental wind conditions. Consequently, design a
control technique for autopilot systems that does not rely on prior knowledge of the hybrid MAV model becomes an
intuitive, innovative and, from the point of view of the authors, an appropriate control methodology. Therefore, the
development of a such controller that estimates the hybrid MAV dynamics and counteracts it, in real time, can be
easily adaptable and implemented for different hybrid MAVs. The reviewed control approaches for hybrid MAVs draw
attention to the following points :
•
Control systems are usually designed from a linearized model of the hybrid MAV behaviour. Nonlinear dynamics
that include aerodynamic effects, such as propeller-wing interaction and stall phenomena, are not correctly
represented in the linearized model around equilibrium points of the hybrid MAV.
•
The entire flight envelope of hybrid MAVs, in terms of control design, is usually addressed by considering two
different flight phases: one for hovering and one for forward flight. After the control design of each flight phase
tackling their respective dynamics, the transitioning phase stability is assured by gain scheduling techniques or by
switching between these two control designs.
•
Model-based control approaches require an identification of aerodynamic forces and moments acting in the system
in order to properly design the controller. This identification, especially for high incidence angles, remains a
difficult, expensive and time consuming process.
This work focuses on the development of a new control architecture for hybrid MAVs composed of model-free
control algorithms. We propose a control strategy that ensures the stability of the system without switching or gain
scheduling methods contributing to the development of a unified control architecture. The present work covers different
steps of flight dynamics field including control design, simulation flight analysis, algorithm implementation up to
experimental flight tests. In terms of flight simulation, a good understanding of aerodynamic forces and moments that
act in the system is required in order to define a realistic hybrid MAV flight simulator. Unfortunately, accurate and
realistic hybrid MAV model remains a very complex task without certain simplifications. Thus, in the following section,
we present a simplified tail sitter MAV model with its aerodynamic assumptions. The obtained tail sitter MAV model is
used to establish a flight simulator in order to test the proposed control approach before of its implementation in real
flight experiments. However, we emphasize that the tail sitter MAV dynamics are unknown to the control and they are
not used to design the controller.
III. Simplified Tail Sitter MAV Model
We present an analytic continuous singularity-free formulation of aerodynamic forces
F
a
b
∈ R
3
and moments
M
a
b
∈ R
3
acting in a wing over a complete 360
◦
angle of attack, based on previous work proposed by [
30
]. The wing
with a surface
S
, is immersed in an incompressible and inviscid airflow with air density
ρ
. The free-stream velocity is
composed by the linear element
v
∞
∈ R
3
and the angular component defined by
ω
∞
∈ R
3
which, in the absence of
wind, is equal to the hybrid MAV angular velocity
ω
b
∈ R
3
. This formulation of aerodynamic forces and moments is
given by :
*
,
F
a
b
M
a
b
+
-
= −
1
2
ρSηCΦ(η
b
)Cη
b
(1)
where
η =
q
v
2
∞
+ µ c
2
ω
2
∞
, with µ ∈ R > 0 (2)
and
η
b
=
*
,
v
∞
ω
∞
+
-
(3)
4
![Fig. 19 Pitch angle result for t ∈ [35; 75] seconds during outdoor flight experiment.](/figures/fig-19-pitch-angle-result-for-t-35-75-seconds-during-outdoor-t9xxp0x7.webp)
