electronics

Article
Speed-Gradient Adaptive Control for Parametrically Uncertain
UAVs in Formation
Alexander M. Popov 1 , Daniil G. Kostrygin 1 , Anatoly A. Shevchik 1 and Boris Andrievsky 1,2,3, *

1 Control Systems and Computer Technologies Department, Faculty of Information and Control Systems,
Baltic State Technical University “VOENMEH” Named after D.F. Ustinov, 190005 Saint Petersburg, Russia
2 Control of Complex Systems Laboratory, Institute of Problems in Mechanical Engineering,
Russian Academy of Sciences, 199178 Saint Petersburg, Russia
3 Applied Cybernetics Department, Faculty of Mathematics and Mechanics, Saint Petersburg University,
198504 Saint Petersburg, Russia
* Correspondence: b.andrievsky@spbu.ru; Tel.:+7-(812)-321-4776

Abstract: The paper is devoted to the problem of the decentralized control of unmanned aerial vehicle
(UAV) formation in the case of parametric uncertainty. A new version of the feedback linearization
approach is proposed and used for a point mass UAV model transformation. As result, a linear model
is obtained containing an unknown value of the UAV mass. Employing the speed-gradient design
method and the implicit reference model concept, a combined adaptive control law is proposed for a
single UAV, including the UAV’s mass estimation and adaptive tuning of the controller parameters.
The obtained new algorithms are then used to address the problem of consensus-based decentralized
control of the UAV formation. Rigorous stability conditions for control and identification are derived,
and simulation results are presented to demonstrate the quality of the closed-loop control system for
various conditions.

Citation: Popov, A.M.; Kostrygin,

Keywords: robust adaptive control; distributed formation control; fixed-wing UAV; parameter
D.G.; Shevchik A.A.; Andrievsky, B. estimation; speed gradient
Speed-Gradient Adaptive Control for
Parametrically Uncertain UAVs in
Formation. Electronics 2022, 11, 4187.
https://doi.org/10.3390/ 1. Introduction
electronics11244187 Adaptation methods are widely used for various tasks involved in controlling the
Academic Editors: Olivier Sename formation flight of unmanned aerial vehicles, such as mission formulation, target deter-
and Soheib Fergani mination, trajectory formation, and ensuring movement along it, as well as control of the
Unnamnned Aerial Vechicles’ (UAVs) angular position under the conditions of parametric
Received: 19 November 2022
uncertainty and characteristics of acting disturbances. Below, we consider a number of
Accepted: 12 December 2022
related publications from recent years.
Published: 14 December 2022
Du et al. [1] dealt with three-dimensional (3D) unmanned aerial vehicle (UAV) path
Publisher’s Note: MDPI stays neutral planning in a complex environment with the aim of avoiding obstacles and finding the best
with regard to jurisdictional claims in path to reach the target position, proposing an improved chimpanzee optimization algo-
published maps and institutional affil- rithm (IChOA) based on an adaptive weight somersault foraging strategy. Sefati et al. [2]
iations. aimed to expand UAV group control concepts, including channel selection and load balanc-
ing, to improve UAV power consumption and reduce latency. Trujillo et al. [3] considered
the problem of distributed control of the position of a group of UAVs to achieve the desired
type of formation and avoid collisions using an adaptive convex combination of two control
Copyright: © 2022 by the authors.
laws: formation control and obstacle avoidance. Two control protocols were proposed to
Licensee MDPI, Basel, Switzerland.
guarantee convergence to the desired formation in a finite or given time. [4] considered
This article is an open access article
distributed under the terms and
the problem of collision avoidance for a swarm of UAVs and presented the results of ex-
conditions of the Creative Commons
perimental studies for a group of quadrocopters. They introduced a group-level detection
Attribution (CC BY) license (https://
and adaptation mechanism for detecting potential collisions between different groups
creativecommons.org/licenses/by/ of UAVs and restructuring the group into subgroups for better collision and deadlock
4.0/). avoidance. Muslimov and Munasypov [5] presented a multi-agent approach to adaptive

Electronics 2022, 11, 4187. https://doi.org/10.3390/electronics11244187 https://www.mdpi.com/journal/electronics

Electronics 2022, 11, 4187 2 of 33

UAV control while tracking a moving ground target. It was proposed to do this by having
the UAV swarm move along a circular trajectory centered on the target. A fuzzy logic-based
Model-Reference Adaptive Control (MRAC) scheme was used to cope with the uncertainty
of the UAV dynamics. The approach was studied for various scenarios using full nonlinear
UAV models. A strategy for searching for survivors in a dynamically changing flood zone
using a group of UAVs was proposed in [6]. A robust adaptive controller was proposed
for its implementation, the feasibility of which was verified by simulation in the presence
of time-varying uncertainties.
Robust consensus tracking problems for a group of cloud-connected UAVs with a
leader–follower formation structure in the presence of uncertainty were considered by
Islam et al. [7]. The convergence performed by the Lyapunov method shows that the
consensus protocol can make the states of UAV followers asymptotically track the state of
the leader.
A group of agents can form a swarm using the extended Cooker–Smale (C–S) model [8].
Song et al. [9] analyzed various levels of inactivity as the degree of control efficiency for
several UAVs in the flocking algorithm. In [9], a heuristic approach of adaptive inactivity
was proposed, which adaptively changes the inactivity level of selected agents in accor-
dance with their position and direction relative to the flock center. Zhou and Chen [10]
used a semi-global consensus leader-following approach in which the control inputs of the
leader and slave agents were bounded. Distributed static and adaptive control protocols
were proposed, and adaptive updating served to avoid the use of global communication
network information. The results of [10] were confirmed by both computer simulations
and real flight tests employing the nanoquadrotor Crazyflie 2.1.
Ahmed et al. [11] noted that the placement of UAVs is critical when designing a
network to collect data from sensors in intelligent environments, as it affects the cost,
reliability, energy consumption, efficiency, and network delay. The problem was considered
one of increasing the efficiency of the system by maximizing the number of serviceable
sensors while using the minimum number of UAVs. Chen et al. [12] studied the multi-user
UAV-enabled mobile edge computing system suffering from a jamming attack by a UAV.
In this paper, reinforcement learning was applied to develop an offloading strategy to
alleviate the effect caused by attack suppression that was able to comply with latency and
power consumption constraints. Mahmood et al. [13] considered improving the quality of
network service through the transfer of extensive computing to the mobile edge cloud and
through the deployment of a UAV integrated with intelligent reflective surfaces. To achieve
an effective solution to the formulated complex problem, the original optimization problem
was divided into subtasks using the block coordinate method. The paper in [14] is devoted
to the integration of satellite and terrestrial networks. A joint optimization design for a non-
orthogonal multiple access-based satellite–terrestrial integrated network was proposed
and studied in [14], and it was shown that the suggested algorithm can improve the
computational efficiency of the iterative algorithm in comparison with habitual randomly
generated initial points.
For networked UAVs, Güzey and Güzey [15] developed a search and tracking scheme
using an adaptive hybrid grouping controller. For search, antennas were placed on at least
four UAVs while keeping the system “+” in the XY-plane (east–west/north–south). UAVs
on the abscissa axis were controlled by the difference in the signal powers of the eastern
and western groups, and along the y-axis by the difference in the respective signal powers
of the northern and southern groups. A backstepping controller was derived using the
Lyapunov stability criteria to allow the UAV to fly in a “+” shape at a fixed altitude. Based
on the example of a quadrotor troop formation, [16] proposed a number of approaches for
assigning UAV targets using time optimization and gain maximization in two stages. In the
first stage, urgently desired targets are selected from the detected target group using the
proposed objective function. In the second, appropriate UAVs are assigned for the selected
desired targets, taking into account coverage factors, adaptive limitations, and constraints.
Electronics 2022, 11, 4187 3 of 33

The problem of cooperative control of a UAV group with finite-time orientation

under conditions of external disturbances and parametric uncertainty was considered
by Han et al. [17]. Using a quaternion model of UAV dynamics, a non-singular sliding
mode terminal surface was constructed for the event-triggered controller. An adaptive
neural network was used to estimate disturbances. Simulation examples were presented
illustrating the efficacy of the proposed control algorithm. Wu et al. [18] considered the
problem of control over a rigid formation with fixed time for a class of nonlinear multi-
agent systems with indefinite dynamics and external disturbances. To compensate for
the uncertainty of the system dynamics and disturbances for a fixed time, an adaptive
observer was introduced for the state and disturbances. Next, a fixed-time formation control
strategy was proposed. The authors proved that the designed controller implements the
boundedness of closed-system errors with a fixed time independently of the initial state
based on stiffness theory and the Lyapunov approach. Results based on simulations
implemented on the UE4 simulation platform confirmed the control of both a single agent
and a UAV formation.
A distributed adaptive fault-tolerant formation control for heterogeneous multiagent
systems with faults in the communication channel was discussed in [19], where a group of
heterogeneous multi-agent systems consisting of UAVs and unmanned ground vehicles
(UGVs) was considered taking into account parametric uncertainties and failures in com-
munication channels. An adaptive fault-tolerant formation control protocol was developed
that uses information about local states for each slave UAV and UGV with different gains,
meaning that all slave vehicles track the dynamic trajectory of the virtual leader and simul-
taneously provide the specified formation in case of communication channel failures or
actions due to external disturbances.
Popov et al. [20] considered the problem of forming and maintaining the formation
of small satellites in near-earth projected circular orbits. In [21], a two-stage algorithm for
group UAV guidance was proposed. In the first stage, a consensus algorithm was employed
for decentralized multi-agent UAV control. In the second, each UAV was independently
guided to the target using a modified proportional 3D guidance law.
Gamagedara and Lee [22] presented a geometric adaptive position control system for
a quadrotor UAV. The thrust direction, which is critical for position tracking, was controlled
irrespective of the yaw direction. Adaptation elements were introduced into the control
loop to mitigate the influence of the disturbances. The efficacy of the proposed control
system was illustrated both numerically and experimentally by indoor and outdoor flights.
Chatterjee and Dutta [23] examined the messaging process for coordinating decisions
within a fully autonomous UAV group in the overall task of Plume Wrapping, which is
used to determine the shape and size of airborne hazardous materials. This is a practical
problem involving a real scenario. An algorithm was proposed to ensure the correctness of
operation and resistance to variable delays when receiving messages.
In [24], a fractional order and trajectory control system was proposed for several
quadrotors in formation using Super Twisting Sliding Mode Control (STSMC) technology.
The Lyapunov function method was employed to synthesize controllers able to compensate
for the influence of parametric uncertainties and wind gusts and provide justify the stability
of the control system. Three types of controllers were considered: a fixed-gain STSMC, an
Adaptive Super Twisting Sliding Mode Control (ASTSMC), and a Fractional-order Adaptive
Super Twisting Sliding Mode Control (FASTSMC). These were tested by UAV swarm
simulation. From the authors’ simulation results, the FASTSMC method demonstrated
better robustness compared to the fixed-gain STSMC and integer-order ASTSMC.
For an underactuated quadrotor, Liang et al. [25] proposed an adaptive robust hier-
archical control strategy. Based on the geometric approach, the orientation error in [25]
was directly defined in the tangent space of the rotation group. To eliminate the influence
of uncertainties, it was proposed to use a detailed dynamic model to represent the main
aerodynamic and mechanical parameters of the quadrotor, and this was used to orga-
nize direct/indirect adaptive control. The presented simulation results showed the good
Electronics 2022, 11, 4187 4 of 33

performance of the system based on the proposed method in terms of tracking accuracy
and robustness.
In a recent paper by Zhi et al. [26], the formation control of multiple fixed-wing
unmanned aircraft vehicles (UAVs) based on a virtual leader structure and subject to
unknown uncertainties and disturbances was studied. A fully distributed model reference
robust adaptive controller (DMRRAC) was proposed, with the aim of swiftly constructing
the formation in complex situations. This approach was able to dramatically reduce reliance
on UAV models and global info. Meanwhile, an adaptive projection formation operator
and novel high-frequency robust term of the controller were used to guarantee that the
formation system was uniformly asymptotically stable under the influences of uncertainties
and disturbances, and could additionally enhance the transient performance and robustness.
Validation and comparative simulations were implemented to demonstrate the proposed
DMRRAC method. A platform based on ROS was used to verify the actual operability of
the controller.
The problem of controlling distributed formation vertical take-off and landing (VTOL)
UAVs under the action of unknown perturbations was studied in [27]. Two algorithms
for distributed formation control with a hierarchical structure were developed. Methods
were proposed to guarantee limited thrust, work without information about acceleration,
and compensation for the influence of external disturbances.
In [28], a method was proposed to ensure the overall stability of a quadrotor-type
UAV formation with significant variations in the parameters of individual UAVs. It was
presumed that the desired trajectory is known only to a single UAV in the formation (“the
leading UAV”). For the synthesis of local controllers, a signal-parametric adaptive control
based on the passification method [29,30] was employed. For illustration, the simulation
results for altitude control of a group of 25 quadrotors were shown. This result was extended
to the adaptive control of spatial motion for a group of quadrotors in [31]. Tomashevich and
Andrievsky [32] applied an approach of [33,34] to the synthesis of a high-order adaptive
algorithm for a multi-agent system control. Quadrotor attitude control was considered
as an example. To demonstrate the convergence of trajectories in a system of four agents,
a laboratory quadrotor setup [35] was used.
Recent results on this topic, including adaptive and variable-structure control ap-
proaches for the problem of spacecraft group motion control, can be found in the survey
by [36]. The application of an adaptive coding procedure for relative motion control of two
satellites over a packet erasure communication channel with a limited transmission rate
was discussed in [37]. A discrete-time adaptive controller for attitude control of a single
spacecraft using reaction wheels was studied in [38].
Contrary to the habitually adopted approach, when the aircraft model is derived with
the help of the Taylor linearization technique in the vicinity of the given flight conditions,
the feedback linearization method is applied to the nonlinear aircraft model. A rigorous
justification of this approach is obtained through the theory of flat systems; see [21,39–43].
The novelty of the present paper’s theory is that the uncertain mass is introduced into a
linearized system, in contrast to the usual approach in which the feedback-linearized plant
dynamics are described by double integrators. This makes it possible to apply adaptation
methods by taking into account the presence of the unknown parameter.
The remainder of this paper is organized as follows. The point-mass model of UAV
dynamics is provided in Section 2. Section 3 is devoted to the adaptive control of a single
UAV. Control of a multi-agent UAVs formation is considered in Section 4, and simulation
results are presented in Section 5. Section 6 presents a discussion of the obtained results
and compares them with other related works. Finally, our concluding remarks and future
work intentions are provided in Section 7.
Electronics 2022, 11, 4187 5 of 33

2. Point-Mass Model of UAV Dynamics

In this paper, we use the following point-mass model of unmanned aerial vehicle
(UAV) dynamics [44–46]:


 ẋ = Vg cos γ cos ψ,

ẏ = Vg cos γ sin ψ,





ḣ = Vg sin γ,




Fthrust − Fdrag


V̇g = − g sin γ, (1)
 m

 Flift sin ϕ
ψ̇ = ,




 V g m cos γ

 Flift cos ϕ g cos γ
γ̇ = − ,


Vg m Vg

where x, y, h are UAV translational coordinates; γ denotes the flight-path angle; ψ is the
1
heading angle; ϕ is the bank angle; Vg denotes the UAV ground speed; Flift = ρ a Vg 2 SCL ,
2
1 2
Fdrag = ρ a Vg SCD are the lift and drag forces, respectively; Fthrust denotes the thrust force;
2
m is the UAV mass; g stands for the gravity acceleration constant; S is the platform area
for a wing; and ρ a denotes the air density; finally, CL is the lift coefficient, CD is the drag
coefficient, and CD = CD0 + k c CL2 , where CD0 denotes zero lift drag coefficient, k c is the
induced drag factor. Following [40,47], let us adopt u = [ Fthrust , CL , ϕ]T as the vector of the
control signals.
In [39,41,42,48], a variant of (1) is used in which the load factor is defined as

Flift
nlf ,
mg

and used instead of the lift coefficient CL . Then, (1) takes the following form:


 ẋ = Vg cos γ cos ψ,

ẏ = Vg cos γ sin ψ,





ḣ = Vg sin γ,



Fthrust − Fdrag


V̇g = − g sin γ, (2)
m
n g sin ϕ


ψ̇ = lf


 ,



 Vg cos γ

g nlf cos ϕ − cos γ


γ̇ = .


Vg

Zhi et al. [26], Wang and Xin [43] used the following variant of (2):


 ẋ = Vg cos γ cos ψ,

ẏ = Vg cos γ sin ψ,





ḣ = Vg sin γ,




Fthrust − Fdrag


V̇g = − g sin γ, (3)
 m
 n g sin ϕ
ψ̇ = lf

,


Vg cos γ





 Flift cos ϕ g cos γ
γ̇ = − ,


Vg m Vg

where the control action is defined as in (2), i.e., u = [ Fthrust , nlf , ϕ]T .
Electronics 2022, 11, 4187 6 of 33

2.1. Feedback Linearization

Models (1)–(3) ues a common approach to the design of the control algorithm, namely,
the feedback linearization method, cf. [49]. This results in an equivalent linear system of
the form
ẍ = v1 ,
ÿ = v2 , (4)
ḧ = v3 ,
where v1 , v2 , v3 are virtual controls which can be uniquely converted into real ones. This
approach was used in [39,40,42,43], and, appears indirectly in [26]. A rigorous justification
for the applicability of this approach follows from the theory of flat systems [50,51]. It can
be shown that outputs x (t), y(t), h(t) of systems (1)–(3) are flat (in the sense of [50,51]) and
that variables Vg , ψ, γ can be expressed in terms of x, y, h and their derivatives, as well as in
terms of control signals. Let the real control signals in (1) be in the form u = [ Fthrust , Flift , ϕ]T ;
here, Flift is used instead of CL . Such a replacement is admissible because it follows from
the seventh equation in (1) that the required value of CL can be uniquely obtained from Flift .
Then, the inverse transformations of the control by dynamic feedback linearization can be
obtained in the following form:
Option 1.

Fthrust = Fdrag + m cos γ(v1 cos ψ + v2 sin ψ) + m sin γ( g + v3 ),

q
Flift = m (( g + v3 ) cos γ − sin γ(v1 cos ψ − v2 sin ψ))2 + (v2 cos ψ − v1 sin ψ)2 , (5)
v2 cos ψ − v1 sin ψ
 
ϕ = arctan .
( g + v3 ) cos γ − (v1 cos ψ + v2 sin ψ) sin γ

Option 2.

ẋv1 + ẏv2 + ḣ( g + v3 )

Fthrust = Fdrag + m
Vg
( g + v3 ) cos γ − (v1 cos ψ + v2 sin ψ) sin γ
Flift = m , (6)
cos ϕ
v2 cos ψ − v1 sin ψ
 
ϕ = arctan
( g + v3 ) cos γ − (v1 cos ψ + v2 sin ψ) sin γ

Option 3.

Fthrust = Fdrag + m cos γ(v1 cos ψ + v2 sin ψ) + m sin γ ( g + v3 ),

( g + v3 ) cos γ − (v1 cos ψ + v2 sin ψ)) sin γ
Flift = m ,
cos ϕ (7)
v2 cos ψ − v1 sin ψ
 
ϕ = arctan .
( g + v3 ) cos γ − (v1 cos ψ + v2 sin ψ) sin γ

Note that the respective expressions for Fthrust and Flift in all three options
are equivalent.
The admissible region in the phase space of the UAV model, in which the obtained
expressions (5)–(7) are defined for (1), is provided by the following inequalities:

Vg > 0, |γ| < π/2, | ϕ| < π/2. (8)

Similar expressions for systems (2) and (3) can be obtained to transform the virtual
control signals v1 , v2 , and v3 into the real ones.
Electronics 2022, 11, 4187 7 of 33

2.2. Feedback Linearization for Variable Mass m Case

The control signals in (5)–(7) are computed from the virtual controls v1 , v2 , and v3 for
system (4). Because Expression (4) does not contain the mass m in it, the mass change is not
taken into account when virtual controls are computed. In order to take into account the
influence of the mass, we can make transformations in (4)–(7) by employing the following
changes in variables:
mv1 = v̄1 ,
mv2 = v̄2 , (9)
m(v3 + g) = v̄3 .
Then,
1
v1 =
v̄ ,
m 1
1
v2 = v̄2 , (10)
m
1
v3 = v̄3 − g.
m
After substituting (10) into (4), we have

1
ẍ = v̄ ,
m 1
1
ÿ = v̄2 , (11)
m
1
ḧ = v̄3 − g,
m
where v̄1 , v̄2 , v̄3 are the new virtual controls.
Now, the transformations to the real control signals (5)–(7) are as follows:
Option 1.

Fthrust = Fdrag + cos γ(v̄1 cos ψ + v̄2 sin ψ) + v̄3 sin γ,

q
Flift = (v̄3 cos γ − sin γ(v̄1 cos ψ − v̄2 sin ψ))2 + (v̄2 cos ψ − v̄1 sin ψ)2 , (12)
v̄2 cos ψ − v̄1 sin ψ
 
ϕ = arctan .
v̄3 cos γ − (v̄1 cos ψ + v̄2 sin ψ) sin γ

Option 2.

Fthrust = Fdrag + ẋ v̄1 + ẏv̄2 + ḣv̄3 /Vg ,
v̄3 cos γ − (v̄1 cos ψ + v̄2 sin ψ) sin γ
Flift = ,
cos ϕ (13)
v̄2 cos ψ − v̄1 sin ψ
 
ϕ = arctan .
v̄3 cos γ − (v̄1 cos ψ + v̄2 sin ψ)) sin γ

Option 3.

Fthrust = Fdrag + cos γ(v̄1 cos ψ + v̄2 sin ψ) + v̄3 sin γ,

v̄3 cos γ − (v̄1 cos ψ + v̄2 sin ψ) sin γ
Flift = ,
cos ϕ (14)
v̄2 cos ψ − v̄1 sin ψ
 
ϕ = arctan .
v̄3 cos γ − (v̄1 cos ψ + v̄2 sin ψ) sin γ

Note that the right-hand sides of (12)–(14) do not include the mass m.
Electronics 2022, 11, 4187 8 of 33

3. Adaptive Control of Single UAV

3.1. Problem Statement and Control Goal
Now, we pose the following control goal: find controls for system (11) while ensuring
tracking of a given trajectory under the condition of an unknown (or slowly time-varying)
parameter m(t), assuming that the range of its possible values is known, i.e., ∀t : mmin 6
m(t) 6 mmax . First, we consider the following prescribed trajectory:

r (t) = [r x (t), ry (t), rh (t)]T , (15)

where the vector function r (t) ∈ R3 is formed from the reference coordinates r x (t), ry (t),
rh (t) along axes x, y, h, respectively. Assume that the time derivatives of r (t) up to r̈ (t)
inclusive are bounded for all t > 0 and the derivative r̈ (t) is a piecewise continuous
function of t.
To solve the posed problem, the following new variables are introduced:

ex1 = x1 − r x (t), ex2 = x2 − ṙ x (t),

ey1 = y1 − ry (t), ey2 = y2 − ṙy (t), (16)
eh1 = h1 − rh (t), eh2 = h2 − ṙh (t),

where x1 = x, x2 = ẋ, y1 = y, y2 = ẏ, h1 = h, h2 = ḣ, and

y x = α x ex1 + ex2 ,
yy = αy ey1 + ey2 , (17)
yh = αh eh1 + eh2 ,

with some α x > 0, αy > 0, αh > 0.

We now introduce the following goal function:

1 2 1 1 m m m
Qt (exyh , yxyh ) = ex1 + e2y1 + e2h1 + y2x + y2y + y2h . (18)
2 2 2 2 2 2
where exyh = (ex1 , ey1 , eh1 ), yxyh = (y x , yy , yh ). Here, the control goal is provided by the
asymptotic relation
lim Qt (exyh , yxyh ) = 0. (19)
t→∞

With respect to the new variables from (16), system (11) has the following form:

1
ëx1 = v̄ − r̈ x (t),
m 1
1
ëy1 = v̄2 − r̈y (t), (20)
m
1
ëh1 = v̄3 − g − r̈h (t).
m
Substituting the variables from (17) to (20), we obtain

ėx1 = −α x ex1 + y x ,
ėy1 = −αy ey1 + yy , (21)
ėh1 = −αh eh1 + yh ,

1
ẏ x = α x y x − α2x ex1 +
v̄ − r̈ x (t),
m 1
1
ẏy = αy yy − α2y ey1 + v̄2 − r̈y (t), (22)
m
1
ẏh = αh yh − α2h eh1 + v̄3 − r̈h (t) − g.
m
Electronics 2022, 11, 4187 9 of 33

Equation (21) describes the internal dynamics, while (22) is used for the external
dynamics. Obviously, the null dynamics of system (21) are exponentially stable [49].

3.2. Design of Main Control Loop

We take the control for the system (21)–(22) in the following form:

v̄1 = −m̂(k x y x − r̈ x (t)) + k̂ x y x ,

v̄2 = −m̂ k y yy − r̈y (t) + k̂ y yy , (23)
v̄3 = −m̂(k h yh − r̈h (t) − g) + k̂ h yh ,

where m̂, k̂ x , k̂ y , k̂ h are adjustable parameters and k x > 0, k y > 0, k h > 0 are fixed (design)
parameters.
Substituting (23) into (22), we have
!  
2 m̂ k̂ x m̂
ẏ x = −α x ex1 + α x − k x + yx + − 1 r̈ x (t),
m m m
!
k̂ y
 
2 m̂ m̂
ẏy = −αy ey1 + αy − k y + yy + − 1 r̈y (t), (24)
m m m
!  
2 m̂ k̂ h m̂
ẏh = −αh eh1 + αh − k h + yh + − 1 (r̈h (t) + g).
m m m

Now, the following Proposition can be stated.

Proposition 1. For system (21) and (22) with control (23) and with a known value of m, it is
always possible to find values of control parameters m̂, k x , k y , k h , k̂ x , k̂ y , k̂ h such that control
goal (19) is fulfilled and the closed-loop system is exponentially stable.

Proof of Proposition 1. Take the following parameter values:

m̂∗ = m,
k x = α x , k y = αy , k h = αh , (25)
k̂ x∗ = −λ x m, k̂ y∗ = −λy m, k̂ h∗ = −λh m,

where λ x > 0, λy > 0, λh > 0 are arbitrarily chosen positive constants (the design parameters).
Then, substituting (25) into (24) provides us with

ẏ x = −α2x ex1 − λ x y x ,
ẏy = −α2y ey1 − λy yy , (26)
ẏh = −α2h eh1 − λh yh .

Take the following Lyapunov function:

V (t) = Qt (exyh , yxyh ) (27)

and calculate the time derivative of V (t) along the trajectories of system (21) and (26):

V̇ = ex1 ėx1 + my x ẏ x + ey1 ėy1 + myy ẏy + eh1 ėh1 + myh ẏh . (28)

Grouping the terms for the x, y, and h axes in (28) results in

V̇ = V̇x + V̇y + V̇h , (29)

Electronics 2022, 11, 4187 10 of 33

where
V̇x = ex1 ėx1 + my x ẏ x ,
V̇y = ey1 ėy1 + myy ẏy , (30)
V̇h = eh1 ėh1 + myh ẏh .
Below, only the expression for V̇x is considered; however, transformations for V̇y , V̇h
can be made similarly. Making substitutions of expressions for derivatives, we obtains
 
V̇x = ex1 (−α x ex1 + y x )+ my x (−α2x ex1 − λ x y x ) = −α x e2x1 + 1 − mα2x ex1 y x − mλ x y2x . (31)

It can be easily seen that V̇x can be represented as a following quadratic form

V̇x = ηx T Hx ηx , (32)

where vector ηx = [ex1 y x ]T and matrix

−α x
0.5 1 − mα2x


Hx = (33)
0.5 1 − mα2x −mλ x

are introduced.
Matrix Hx = HxT is negative definite if its first angular minor is negative and its second
is positive. In our case, the first minor ∆1 = −α x < 0, and the second should satisfy the
inequality
 2
∆2 = mα x λ x − 0.25 1 − mα2x > 0. (34)

We can represent inequality (34) in the following form:

1  2 2
λx > mα x − 1 . (35)
4mα x

Obviously, with the known values of m and α x , there is always a value of λ x for
which inequality (35) is satisfied. In a similar way, by introducing vectors ηy = [ey1 yy ]T ,
ηh = [eh1 yh ]T and matrices
   
−αy 0.5 1 − mα2y
Hy =     (36)
0.5 1 − mα2y −mλy

−αh
0.5 1 − mα2h



Hh = , (37)
0.5 1 − mα2h −mλh
it can be shown that

V̇y = ηy T Hy ηy < 0, V̇h = ηh T Hh ηh < 0, (38)

is negative definite if the conditions

1  2 2 1  2 2
λy > mαy − 1 , λh > mαh − 1 . (39)
4mαy 4mαh

are satisfied.
As a result, we obtain the inequality

V̇ = ηx T Hx ηx + ηy T Hy ηy + ηh T Hh ηh < 0, (40)

It can be easily shown, cf. [49], that the above inequality implies the following one:

V̇ < −ρV, (41)

Electronics 2022, 11, 4187 11 of 33

where ρ > 0 is a certain constant. From this, it follows that the system (21) and (22) with
control (23) is exponentially stable and that control goal (19) is satisfied. This completes
the proof.

3.3. Synthesis of UAV Adaptation Law

To obtain the adaptive tuning law for control law (23), the adjustable parameters
m̂, k̂ x , k̂ y , k̂ h are employed with the Speed-gradient (SG) method, cf. [52–55].
Let the control aim be as given by (19). The time derivative of Qt along system (21)
and (22) trajectories with control (23) is as follows:

Q̇t = ex1 ėx1 + ey1 ėy1 + eh1 ėh1 + my x ẏ x + myy ẏy + myh ẏh . (42)

Taking into account (24), we have

Q̇t = ex1 ėx1 + ey1 ėy1 + eh1 ėh1 +

! 
 !
m̂ k̂ x m̂
+ my x −α2x ex1+ αx − k x + yx + − 1 r̈ x (t) +
m m m
! !
k̂ y
 
2 m̂ m̂
+ myy −αy ey1 + αy − k y + yy + − 1 r̈y (t) +
m m m
!   !
2 m̂ k̂ h m̂
+ myh −αh eh1 + αh − k h + yh + − 1 (r̈h (t) + g) (43)
m m m
= ex1 ėx1 + ey1 ėy1 + eh1 ėh1
 
− mα2x ex1 y x + mα x − m̂k x + k̂ x y2x + (m̂ − m)r̈ x (t)y x
 
− mα2y ey1 yy + mαy − m̂k y + k̂ y y2y + (m̂ − m)r̈y (t)yy
 
− mα2h eh1 yh + mαh − m̂k h + k̂ h y2h + (m̂ − m)(r̈h (t) + g)yh .

The vector of function (43) has a gradient with respect to the adjustable parameters m̂,
k̂ x , k̂ y , k̂ h as follows:

−k x y2x + y x r̈ x (t) − k y y2y + yy r̈y (t) − k h y2h + yh (r̈h (t) + g)

 
 y2x 
∇ Q̇t =  . (44)
 y2y 
2
yh

The SG design procedure leads to the following tuning algorithms for variables m̂, k̂ x ,
k̂ y , k̂ h in (23):
 
m̂˙ = −γm −k x y2x + y x r̈ x (t) − k y y2y + yy r̈y (t) − k h y2h + yh (r̈h (t) + g) ,

k̂˙ x = −γx y2x ,

(45)
k̂˙ y = −γy y2y ,
k̂˙ = −γ y2 ,
h h h

where γm > 0, γx > 0, γy > 0, γh > 0 are the adaptation gains (i.e., the design parameters).
The following result can be derived.

Proposition 2. In system (21) and (22) with control (23) and tuning algorithms (45), the control
goal (19) is achieved and the tuning parameters m̂, k̂ x , k̂ y , k̂ h are bounded.
Electronics 2022, 11, 4187 12 of 33

Proof of Proposition 2. To prove this, we can check the conditions under which the SG
scheme is operable [52–55]:
1. The scalar objective function Qt (exyh , yxyh ) is non-negative and satisfies the growth
condition inf Qt (exyh , yxyh ) → ∞ as k[exyh , yxyh ]k → ∞. Note that this condition is
t >0
met for function (19).
2. Function Q̇t should be convex on adjustable parameters m̂, k̂ x , k̂ y , k̂ h . This condition
is met because Q̇t is linear with respect to the adjustable parameters.
3. The following reachability condition is satisfied: for any m from the range of its
possible values mmin 6 m 6 mmax , there exists a set of parameters m̂∗ , k̂ x∗ , k̂ y∗ , k̂ h∗
and a function f ( Qt ) ( f ( Qt ) > 0 as Qt > 0) such that for all x, y, h, t, the inequality


Q̇t ((exyh , yxyh , m̂∗ , k̂ x∗ , k̂ y∗ , k̂ h∗ ) 6 − f Qt (exyh , yxyh ) , (46)

is fulfilled, as follows from Proposition 1.

This completes the proof.

Remark 1. Relations (17) can be treated as an implicit reference model [29,30,53,56], where the
parameters α x , αy , αh set the goal function (18) convergence rate.

Remark 2. The efficiency of the adaptation algorithm (45) is preserved if the mass m in (11) is
time-varying with a sufficiently small rate of change. The rigorous consideration of the case of the
time-varying parameters can be performed based on the recent results in [57,58].

Remark 3. The adaptation algorithm (45) is a combined one. It includes the identification procedure
provided by the first expression in (45) and the direct adaptive control described by the last three
expressions in (45).

Remark 4. In practice, it is advisable to use the following regularized form of the adaptation
algorithm (45), cf. [53]:
 
m̂˙ = −γm −k x y2x + y x r̈ x (t) − k y y2y + yy r̈y (t) − k h y2h + yh (r̈h (t) + g) − σm (m̂ − m0 ),

k̂˙ x =
 
−γx y2x − σx k̂ x − k̂ x0 ,
(47)
k̂˙ y =
 
−γy y2y − σy k̂ y − k̂ y0 ,

k̂˙ h =
 
−γh y2h − σh k̂ h − k̂ h0 ,

where σm > 0, σx > 0, σy > 0, σh > 0 are parametric feedback gains (design parameters) and m0 ,
k̂ x0 , k̂ y0 , k̂ h0 are “guessed” values of the control law (23) tunable coefficients chosen based on the
prior information of the UAV parameters.

4. Control of Multi-Agent UAV Formation

4.1. Problem Description
Consider a group of N UAVs (agents) indexed as i = 1, 2, . . . , N. The dynamics of each
UAV can be described by (11), as follows:

1
ẍi = v̄ ,
mi 1i
1
ÿi = v̄ , (48)
mi 2i
1
ḧi = ¯ − g,
uv
mi 3i
Electronics 2022, 11, 4187 13 of 33

where mi is the mass of the i-th UAV and v̄1i , v̄2i , ū3i are the virtual controls for i-th UAV.
Here, masses mi are assumed to be unknown within the known range mmin i 6 mi 6 mmax i ,
i = 1, 2, . . . , N.

Remark 5. UAV communication channels and data exchange protocols are not considered in the
present paper, and it is assumed that the information exchange necessary to implement the proposed
algorithms is feasible in one form or another. Results regarding data transfer between UAVs in
formation can be found in [23,37,59–72]. Particularly, in [59,63,66,67], adaptive coding procedures
for navigation data exchanged between the UAVs in formation was studied in depth. For satellite
communications, Lin et al. [69] solved the problem of safe and efficient beamforming in multi-agent
satellite systems to simultaneously achieve data transmission security against interception and low
power consumption. Lin et al. [70] explored multicast satellite and aerial-integrated network with
rate-splitting multiple access, where the satellite and UAV components are controlled by the network
control center and operate in the same frequency band. An optimization problem was formulated to
maximize the total speed in terms of the signal–interference ratio and power limitations per antenna
in the UAV and satellite. Lin et al. [71] addressed the aim of providing efficient satellite-to-satellite
communication while minimizing the overall transmitted signal power and meeting the bit-rate
requirements for multiple blocked users. In [72], a broad overview of the state-of-the-art of UAV
communications from an industrial point of view was provided. It was shown that sub-6GHz
mainstream MIMO can successfully deal with cellular selection and interference problems; in
addition, the prospects for next-generation UAV communications were outlined and an assessment
of the prospective benefits for UAVs was provided. In Geraci et al. [72], the authors discussed the
main technological obstacles that stand in the way of further UAV communications development.

Unlike [5,23,27,42,73–75], in the present paper we assume that the reference trajec-
tory (15) along which the UAV group should move is given ahead of time. The group
formation is set by the vector relative position (the offset vector) for each UAV with respect
to trajectory (15), as follows:

δi = [δxi , δyi , δhi ]T , i = 1, 2, . . . , N. (49)

Then, the reference trajectory for the i-th UAV i is provided by the following vector
function:

ri (t) = [r x (t) + δxi , ry (t) + δyi , rh (t) + δhi ]T = [r xi (t), ryi (t), rhi (t)]T . (50)

By choosing appropriate offsets δi , we can set the required formation of the UAV
group. By analogy with (16) and (17), we can define the following variables for system (48):

ex1i = x1i − r xi (t), ex2i = x2i − ṙ xi (t),

ey1i = y1i − ryi (t), ey2i = y2i − ṙyi (t), (51)
eh1i = h1i − rhi (t), eh2i = h2i − ṙhi (t),

where x1i = xi , x2i = ẋi , y1i = yi , y2i = ẏi , h1i = hi , h2i = ḣi , and

y xi = α xi ex1i + ex2i ,
yyi = αyi ey1i + ey2i , (52)
yhi = αhi eh1i + eh2i ,

where α xi > 0, αyi > 0, αhi > 0, i = 1, 2, . . . , N.

Electronics 2022, 11, 4187 14 of 33

With these variables, system (48) can be written as

ėx1i = −α xi ex1i + y xi ,
ėy1i = −αyi ey1i + yyi , (53)
ėh1i = −αhi eh1i + yhi ,

1
ẏ xi = α xi y xi − α2xi ex1i +v̄ − r̈ xi (t),
mi 1i
1
ẏyi = αyi yyi − α2yi ey1i + v̄ − r̈yi (t), (54)
mi 2i
1
ẏhi = αhi yhi − α2hi eh1i + v̄ − r̈hi (t) − g.
mi 3i
Let the control goal for the i-th UAV be lim Qti (exyhi , yxyhi ) → 0, where exyhi = (ex1i , ey1i , eh1i ),
t→∞
y xyhi = (y xi , yyi , yhi ) are introduced. Then,

1 2 1 1 m m m
Qti (exyhi , yxyhi ) = e + e2 + e2 + i y2xi + i y2yi + i y2hi . (55)
2 x1i 2 y1i 2 h1i 2 2 2
Because no interaction between UAVs in a group is assumed, goal (55) can be achieved
using the earlier described control loop (23) as follows:

v̄1i = −m̂i (k xi y xi − r̈ xi (t)) + k̂ xi y xi ,

v̄2i = −m̂i k yi yyi − r̈yi (t) + k̂ yi yyi , (56)
v̄3i = −m̂i (k hi yhi − r̈hi (t) − g) + k̂ hi yhi ,

where m̂i , k̂ xi , k̂ yi , k̂ hi are tunable controller parameters and k xi > 0, k yi > 0, k hi > 0 stand
for the fixed (design) parameters.
The adaptation algorithm is as follows:
 
m̂˙ i = −γmi −k xi y2xi + y xi r̈ xi (t) − k yi y2yi + yyi r̈yi (t) − k hi y2hi + yhi (r̈hi (t) + g) ,

k̂˙ xi = −γxi y2xi ,

(57)
k̂˙ yi = −γyi y2yi ,

k̂˙ hi = −γhi y2hi ,

where γmi > 0, γxi > 0, γyi > 0, γhi > 0 are the adaptation gains (design parameters).
If the control goal for the entire UAV group is taken as the sum of the goals for the i-th
UAV, namely,
N
QtF (exyh , yxyh ) = ∑ Qti (exyh i , yxyh i ), (58)
i =1

then QtF (exyh , yxyh ) → 0, as t → ∞, because each term tends to zero.

4.2. Consensus Algorithm for Multi-Agent Formation Control

Control laws (56) and (57) are applied independently and separately to each UAV in
the formation, and rely on information about the UAVs’ relative positions. For the group
control problem, additional signals employing the consensus algorithm can significantly
increase accuracy in the presence of noise (cf. [20,76–78]). Therefore, to improve formation
positioning accuracy, a consensus algorithm (protocol) can be applied for multi-agent
formation control. Unlike [28,31,32], in which the reference position r (t) is known only to
a single UAV in the formation (“the leading UAV”), here we assume, as above, that each
UAV has information about r (t). This makes it possible to employ the consensus-based
approach described below instead of relying on the leader’s position tracking.
Electronics 2022, 11, 4187 15 of 33

4.2.1. Basic Information on Consensus Algorithm

Following [10,20,75–79], we define graph G as a pair (V , E ), where V = 1, . . . , n is a
set of nodes (agents) and E ∈ V × V is a set of edges in which each edge is represented
by an ordered pair of different nodes. Edge (i, j) shows that node i is a neighbor to node
j, and node j can receive information from node i. A graph is called directed if, for each
(i, j) ∈ E , ( j, i ) ∈ E . The path from node i1 to node il is a sequence of ordered edges with
the form (ik , ik+1 ), k = 1, ..., l − 1. An undirected graph is connected if there are paths to all
other nodes for any i ∈ V .
Let graph G contains n nodes. The adjacency matrix A = aij  ∈Rn×n is defined as
 

aii = 0, aij = 1 if ( j, i ) ∈ E , and 0 otherwise. The Laplace matrix L = lij ∈ Rn×n is defined
as lii = ∑ Nj=1 aij and lij = − aij , i 6 = j. A directed tree is a directed graph in which every node
has exactly one parent except for one node, called the root, which has no parent and has a
directed path to every other node. A spanning tree of a directed graph is a directed tree
formed by graph edges that connect all the nodes of the graph. Let us assume that the
communication graph G is undirected and connected; examples of undirected connected
graphs are depicted in Figure 1.

3 2 4

Figure 1. Communication graph.

Assume that ξ i ∈ R contains information about the state of the i-th agent. For infor-
mation states with first-order dynamics, the following fundamental first-order consensus
algorithm is known:
ξ̇ i = ui (59)
where ui ∈ R has the following form:
n
∑ aij


u i = − ρ0 ξi − ξ j , (60)
j =1

and ρ0 > 0 is a certain constant.

It is assumed for algorithms (59) and (60) that consensus is achieved asymptotically among
several agents when for any ξ i (0) and for all i 6= j it is valid that lim ξ i (t) − ξ j (t) = 0.
t→∞

4.2.2. UAV Formation Control based on the Consensus Algorithm

Consider the following extended control algorithm for system (53) and (54):

v̄1i = −m̂i (k xi y xi − r̈ xi (t)) + k̂ xi y xi + v̄1ci ,

v̄2i = −m̂i k yi yyi − r̈yi (t) + k̂ yi yyi + v̄2ci , (61)
v̄3i = −m̂i (k hi yhi − r̈hi (t)) + k̂ hi yhi + v̄3ci ,
Electronics 2022, 11, 4187 16 of 33

where v̄1ci , v̄2ci , v̄3ci are additional control signals based on the consensus algorithm by Ren
and Beard [76]:
N
v̄1ci = −k xc y xi − ρ x ∑ aij y xi − y xj ,


j =1
N
∑ aij


v̄2ci = −k yc yyi − ρy yyi − yyj , (62)
j =1
N  
v̄3ci = −k hc yhi − ρh ∑ ij hi hj ,
a y − y
j =1

where k xc > 0, k yc > 0, k hc > 0, ρ x > 0, ρy > 0, ρh > 0 denote the consensus algorithm
parameters and aij represents the adjacency matrix elements, thereby defining a directed
connections graph (digraph) G between the UAVs for the consensus algorithms (59) and
(60). In matrix form, (62) can be written as

v̄1c = −(k xc IN + ρ x L)y x ,

v̄2c = −(k yc IN + ρy L)yy , (63)
v̄3c = −(k hc IN + ρh L)yh ,

where IN denotes the N × N identity matrix, L is the Laplace matrix corresponding to the
adjacency matrix and the directed connection graph G , and the following vector notations
are used: y x = [y x1 , y x2 , . . . , y xN ]T , yy = [yy1 , yy2 , . . . , yyN ]T , yh = [yh1 , yh2 , . . . , yhN ]T , v̄1c =
[v̄1c1 , v̄1c2 , . . . , v̄1cN ]T , v̄2c = [v̄2c1 , v̄2c2 , . . . , v̄2cN ]T , v̄3c = [v̄3c1 , v̄3c2 , . . . , v̄3cN ]T .
The following Proposition can now be stated.

Proposition 3. System (53) and (54) with control (61) and (62), adaptation algorithms (57),
control goal (58), and the settings m̂i , k̂ xi , k̂ yi , k̂ hi , and i = 1, 2, . . . , N is stable if the associated
communication digraph G contains a spanning tree.

Sketch of Proof for Proposition 3. We can prove this proposition, using the SG scheme by
checking the following conditions (cf. the proof of Proposition 2):
1. The scalar objective function QtF (exyh , y xyh ) is non-negative and satisfies the growth
condition inf QtF (exyh , y xyh ) → ∞ as k(exyh , y xyh )k → ∞. This is true for function (58).
t>0
2. Function Q̇tF should be convex on the adjustable parameters m̂i , k̂ xi , k̂ yi , k̂ hi , i =
1, 2, . . . , N. This condition is met because Q̇tF is linear with respect to the parameters.
3. Reachability condition: for any set of values mi , i = 1, 2, ..., N from the ranges of possible
values mmin i 6 mi 6 mmax i , there exists a set of parameters m̂∗ = [m̂∗1 , m̂∗2 , . . . , m̂∗ N ]t rn,
k̂ x∗ = [k̂ x∗1 , k̂ x∗2 , . . . , k̂Tx∗ N ], k̂ y∗ = [k̂ y∗1 , k̂ y∗2 , . . . , k̂ y∗ N ]T , k̂ h∗ = [k̂ h∗1 , k̂ h∗2 , . . . , k̂ h∗ N ]T
and function f ( QtF ) ( f ( QtF ) > 0 as QtF > 0) for which for all x, y, h, t hold:

Q̇tF (exyh , y xyh , m̂∗ , k̂ x∗ , k̂ y∗ , k̂ h∗ ) 6 − f ( QtF (exyh , y xyh )). (64)

To prove this, we can consider the Lyapunov function:

VF (t) = QtF (exyh , y xyh ). (65)

Because QtF (exyh , y xyh ) is the sum of the control objective functions for each UAV
and the control signal based on the consensus algorithm enters additively into the
overall control signal (61), by taking into account expressions (27) and (41) the proof
of Proposition 1 implies

N
V̇F (t) = − ∑ ρi Qti (exyhi , y xyhi ) + f c (exyh , y xyh ), (66)
i =1
Electronics 2022, 11, 4187 17 of 33

where ρi > 0, i = 1, 2, . . . , N and f c (exyh , y xyh ) is an addition to the derivative of the

consensus algorithm (63). This function is as follows:

f c (exyh , y xyh ) = −yTx (k xc IN + ρ x L)y x − yTy (k yc IN + ρy L)yy − yTh (k hc IN + ρh L)yh . (67)

By assumption, the associated communication digraph G contains a spanning tree;
therefore (see [76]), the Laplace matrix L is positive semi-definite, and hence all
the matrices inside the parentheses of expression (67 ) are positive definite. Then,
f c (exyh , y xyh ) < 0, and for (66) there always exists ρ∗ > 0 such that

V̇F (t) < −ρ∗ QtF (exyh , y xyh ) (68)

is valid. Therefore, condition (64) is satisfied.

This completes the proof.

Remark 6. In practice, it is preferable to use the robust form of the tuning algorithm (57), cf. [53]:
 
m̂˙ i = −γmi −k xi y2xi + y xi r̈ xi (t) − k yi y2yi + yyi r̈yi (t) − k hi y2hi + yhi (r̈hi (t) + g)
− σm (m̂i − m0i ),
k̂˙ xi = −γxi y2xi − σxi k̂ xi − k̂ x0i ,
 
(69)
k̂˙ yi = −γyi y2yi − σyi k̂ yi − k̂ y0i ,
 

k̂˙ hi = −γhi y2hi − σhi k̂ hi − k̂ h0i ,

 

where σmi > 0, σxi > 0, σyi > 0, σhi > 0; m0i , k̂ x0i , k̂ y0i , k̂ h0i are some “guessed” values of the
adjustable control law parameters.

Remark 7. The obtained adaptation and control algorithms for a single UAV in (23) and (47) for
group control of UAVs in (61) and (62), as well as the tuning algorithms (69), are applicable for
UAV models with the form (1) and expressions (12)–(14) when conditions (8) are met.

5. Simulation Results
We carried out computer simulations for various scenarios with the proposed adaptive
control algorithms in the Matlab/Simulink software environment.
For converting virtual controls into real ones, the control plant model was taken as (1)
with expressions (12). We used the control laws for a single UAV in (23) and (47), for group
control in (61) and (62), for and tuning algorithms (69).

5.1. General Data for Simulations

A circle with a constant radius of 200 m at a constant height of 100 m was used as the
motion trajectory:

r x (t) = 200 cos(0.25t), ry (t) = 200 sin(0.25t), rh (t) = 100. (70)

The nominal parameters of the UAV were as follows: m = 500 kg nominal mass;
S = 7.98 m2 platform area for a wing; ρ a = 1.225 kg/m3 air density, CD0 = 0.02 zero-lift
drag coefficient, k c = 0.1 induced drag factor.
The simulation results of UAV motion along a given trajectory are presented in Figure 2.
The simulation results of group UAV motion along a given trajectory are presented in
Figure 3. In (17), the following output parameters y x , yy , yh were used: α x = αy = αh =
0.0397. The control law (23) constant parameters were set to k x = k y = k h = 0.0796. The
adaptation algorithm (47) parameters were: γm = 2, σm = 0.008, γx = γy = γh = 0.75,
σx = σy = σh = 0.008. The following initial values of the adjustable parameters were used:
Electronics 2022, 11, 4187 18 of 33

m0 = 325 kg, k̂ x0 = k̂ y0 = k̂ h0 = −30. To prevent negative values of the thrust force, it was
limited from below by 50 N.
For the group flight simulation, the control law parameters used for each UAV were
the same as for the single UAV simulation.

Figure 2. UAV trajectory for adaptive control case.

Figure 3. UAV trajectory for group control case.

5.2. Control of Single UAV

The following values were used as the initial parameters of the simulation. Initial
coordinates: x (0) = 50 m, y(0) = 0 m, h(0) = 75 m; initial speed Vg (0) = 50 m/s; initial
attitude: γ(0) = 0 rad, ψ(0) = π/2 rad.

5.2.1. Control of Single UAV with Linearly Varying Mass

For the simulations, the UAV mass was considered in the form of a linear function in
time with an initial value m = 400 kg and slope −0.4 kg/s.
The simulation results are shown in Figure 4.
Electronics 2022, 11, 4187 19 of 33

150

100

50

0 50 100 150 200 250 300 350 400

(a)

500

400

300

200
0 50 100 150 200 250 300 350 400

(b)

-40

-60

-80

-100

-120
0 50 100 150 200 250 300 350 400

(c)

800

600

400

200

0
0 50 100 150 200 250 300 350 400

(d)
Figure 4. Coordinate errors (a), estimated mass value (b), controller gains (c), and thrust force (d) for
a scenario in which the UAV mass changes linearly over time.

5.2.2. Control of Single UAV with Jump-changing Mass

Unknown UAV mass changes linearly until t = 200 s; its initial value is 400 kg, and
the slope is −0.2 kg/s. At time instant t = 200 s, jumplike mass decreases 50 kg.
The simulation results are shown in Figure 5.
Electronics 2022, 11, 4187 20 of 33

150

100

50

0 50 100 150 200 250 300 350 400

(a)

550

500

450

400

350

300

250
0 50 100 150 200 250 300 350 400

(b)

-40

-60

-80

-100

-120
0 50 100 150 200 250 300 350 400

(c)

800

600

400

200

0
0 50 100 150 200 250 300 350 400

(d)
Figure 5. Coordinate errors (a), estimated mass value (b), controller gains (c), and thrust force (d) for
scenario of jumplike UAV mass changing.

5.3. UAV Group Control

Next, we consider the control problem for a group of four autonomous UAVs. The
formation is configured by offsets (49) and (50) relative to the reference trajectory (70):

δx = [0, −25, −25, −50]T , δy = [0, 25, −25, 0]T , δh = [0, 0, 0, 0]T . (71)

The consensus algorithm in form (62) was used to implement the group control. Data
exchange between agents was carried out according to the connections graph shown in
Figure 1.
The consensus algorithm parameters are as follows: k xc = k yc = k hc = 350, ρ x = ρy =
ρh = 250.
The following common initial conditions for all UAVs of the group were used: initial
speed Vg (0) = 50 m/s; initial attitudes γ(0) = 0 rad, ψ(0) = π/2 rad.
Electronics 2022, 11, 4187 21 of 33

For the first UAV, the initial uncertain mass was m = 500 kg, and the coordinates were
x (0) = 50 m, y(0) = 0 m, h(0) = 75 m. For the second UAV, the initial parameters were
m = 450 kg and the initial coordinates x (0) = 100 m, y(0) = 30 m, h(0) = 75 m. For the
third UAV, the initial uncertain mass was taken as m = 400 kg and the initial coordinates
x (0) = 0 m, y(0) = 0 m, h(0) = 75 m. For the fourth UAV, the initial uncertain mass was
taken as m = 350 kg and the initial coordinates x (0) = 75 m, y(0) = 0 m, h(0) = 75 m.
The masses of the agents change linearly with the same slope of −0.22 kg/s; in addi-
tion, each UAV loses 50 kg of mass once at time instants 100 s, 150 s, 200 s, and 250 s.
The simulation results are shown in Figures 6–9. Figure 6 shows the errors in coordi-
nates for each UAV in the group, Figure 7 demonstrates the mass estimates, Figure 8 shows
the controller gains time histories, and Figure 9 shows the thrust forces time histories.

100

50

0 50 100 150 200 250 300 350 400 450

(a)

80

60

40

20

-20
0 50 100 150 200 250 300 350 400 450

(b)

50

-50
0 50 100 150 200 250 300 350 400 450

(c)

40

20

-20

-40
0 50 100 150 200 250 300 350 400 450

(d)
Figure 6. Coordinate errors for control the case of a group of UAVs of variable mass with initial mass
values: 500 kg (a), 450 kg (b), 400 kg (c), 350 kg (d).
Electronics 2022, 11, 4187 22 of 33

550

500

450

400

350

300
0 50 100 150 200 250 300 350 400 450

(a)

500

450

400

350

300
0 50 100 150 200 250 300 350 400 450

(b)

500

450

400

350

300

250
0 50 100 150 200 250 300 350 400 450

(c)

400

350

300

250

200
0 50 100 150 200 250 300 350 400 450

(d)
Figure 7. Estimated mass values when controlling a group of UAVs of variable mass with initial mass
values of 500 kg (a), 450 kg (b), 400 kg (c), and 350 kg (d).
Electronics 2022, 11, 4187 23 of 33

-30

-35

-40

-45

-50

-55
0 50 100 150 200 250 300 350 400 450

(a)

-30

-32

-34

-36

-38
0 50 100 150 200 250 300 350 400 450

(b)

-30

-35

-40

-45

-50

-55
0 50 100 150 200 250 300 350 400 450

(c)

-30

-31

-32

-33

-34

0 50 100 150 200 250 300 350 400 450

(d)
Figure 8. Coefficients of regulators when controlling a group of UAVs of variable mass with initial
mass values of 500 kg (a), 450 kg (b), 400 kg (c), and 350 kg (d).
Electronics 2022, 11, 4187 24 of 33

600

400

200

0
0 50 100 150 200 250 300 350 400 450

(a)

600

400

200

0
0 50 100 150 200 250 300 350 400 450

(b)

400

300

200

100

0
0 50 100 150 200 250 300 350 400 450

(c)

300

200

100

0
0 50 100 150 200 250 300 350 400 450

(d)
Figure 9. Thrust forces for the case of a group of variable mass UAVs control with initial mass values
of 500 kg (a), 450 kg (b), 400 kg (c), and 350 kg (d).

5.4. UAV Group Control under External Disturbances

To test the robustness properties of the obtained algorithms and compare them with
other algorithms that solve similar problems, the results by Zhi et al. [26] for a four-
agent formation were simulated. Control laws for group control (61) and (62) and tuning
algorithms (69) were used.
The desired trajectory was

r x (t) = 60t m, ry (t), ≡ 0, rh (t) = 80 m. (72)

Electronics 2022, 11, 4187 25 of 33

The following nominal parameters of the UAV were used: mass m = 20 kg; wing area
S = 1.37 m2 ; air density ρ a = 1.225 kg/m3 ; zero-lift drag coefficient CD0 = 0.02; induced
drag factor k c = 0.1. In the simulation, at the 25-th second of the flight the coefficients CD0
and k c = 0.15 increased abruptly by 50% (up to CD0 = 0.03, k c = 0.15).
The formation was configured by offsets (49) and (50) relative to the reference trajectory
(72)
δx = [0, −100, 0, 100]T , δy = [100, 0, −100, 0]T , δh = [0, 0, 0, 0]T . (73)
The control laws (52), (61), (62) and tuning algorithm (69) parameters were identical for
all the UAVs in the formation. In (52), the following output parameters y xi , yyi , yhi were
used: α xi = αyi = αhi = 0.2182. The constant parameters of control law (61) were set to
k xi = k yi = k hi = 0.4583. The following adaptation algorithm (47) parameters were used:
γmi = 15.5, σmi = 0.008, γxi = γyi = γhi = 15.5, σxi = σyi = σhi = 0.008, and the initial
values of adjustable parameters were set to m0i = 20 kg, k̂ x0i = k̂ y0i = k̂ h0i = −100.
The consensus algorithm in form (62) was used to implement the group control. Data
exchange between agents was carried out according to the connections graph shown in
Figure 1. The consensus algorithm parameters were as follows: k xc = k yc = k hc = 850,
ρ x = ρy = ρh = 700. The initial parameters were the same as in Zhi et al. [26].
The following common initial conditions were used for all UAVs in the group: initial
attitudes γ(0) = 0 rad, ψ(0) = 0 rad.
For the first UAV, the initial uncertain mass was m = 24 kg with initial coordinates
x (0) = 0 m, y(0) = 200 m, h(0) = 95 m, speed Vg (0) = 70 m/s. For the second UAV,
the initial uncertain mass was m = 17 kg with initial coordinates x (0) = 0 m, y(0) = 60 m,
h(0) = 90 m, speed Vg (0) = 60 m/s. For the third UAV, the initial uncertain mass was
m = 22 kg with initial coordinates x (0) = 0 m, y(0) = −200 m, h(0) = 70 m, speed
Vg (0) = 40 m/s. For the fourth UAV, the initial uncertain mass was m = 22 kg with initial
coordinates x (0) = 0 m, y(0) = −60 m, h(0) = 80 m, speed Vg (0) = 50 m/s.
In addition, in the simulations the masses of agents changed linearly with the identical
slope of −0.083 kg/s.
External perturbations d x , dy , dh were added to the first three equations of (1). The ex-
ternal disturbances were the same as in Zhi et al. [26], namely, d x1 = 0.5 cos(t),
dy1 = 0.8 sin(t), dh1 = 0.5 cos(t), d x2 = 4 sin(t), dy2 = 3 sin(t), dh2 = 4 sin(t),
d x3 = cos(2t), dy3 = 2 sin(t), dh3 = cos(2t), d x4 = 2 cos(t), dy4 = sin(t), dh4 = 2 cos(t).
The simulation results are shown in Figures 10–12. Figure 10 shows the 3D tracking
trajectories, Figure 11 shows the tracking error comparison curves of three positions and
three velocities, and Figure 12 shows the mass estimates.
Comparison of these results with those obtained by Zhi et al. [26] allows us to
conclude that the algorithms presented in our paper show results no worse than in [26],
while at the same time having a much simpler structure and lower number of ad-
justable parameters.
Electronics 2022, 11, 4187 26 of 33

Figure 10. 3D tracking trajectories.

100 20
50 10

0 0

-50 -10

-100 -20
0 10 20 30 40 50 60 0 10 20 30 40 50 60

100 20
50 10

0 0

-50 -10

-100 -20
0 10 20 30 40 50 60 0 10 20 30 40 50 60

3
20
2
10 1
0
0
-1
-10 -2
0 10 20 30 40 50 60 -3
0 10 20 30 40 50 60

Figure 11. Tracking error comparison curves of three positions and three velocities.
Electronics 2022, 11, 4187 27 of 33

Figure 12. Estimated masses of UAVs in group.

6. Discussion
At present, a great deal of work is being devoted to the UAV formation control
problem. A detailed explanation of the differences between the present work and others
is provided in Table 1. Each row (item) of the table contains the distinctive features of
this paper as compared to those noted in the corresponding item; for brevity and clarity,
similar properties are not indicated. The following characteristics are notable: generation
of the desired trajectory (whether it is predefined or generated depending on the current
situation); type of UAV; control/adaptation method; set of uncertain UAV parameters;
control vector components; whether or not consensus is ensured; model of UAV motion;
and communication graph topology.
Electronics 2022, 11, 4187 28 of 33

Table 1. Comparison with other works.

# Other Works This Work

[5] – tracking the moving ground target in the real-time;
– tracking the given reference trajectory;
1 method: fuzzy-logic model-reference adaptive control
method: implicit reference model (IRM) adaptive control
(FL MRAC)
[20] – UAV parameters are known; – uncertain parameters: m, CD0 , k c , S;
2
method: non-adaptive control method: IRM adaptive control
[27] – UAV type: VTOL; – UAV type: fixed-wing aircraft;
control vector: command forces and torques; control vector: u = [ Fthrust , CL , ϕ]T ;
3
UAV parameters are known; uncertain parameters: m, CD0 , k c , S;
method: adaptive estimation of disturbance method: adaptive control with the IRM
[26] – control vector: u = [ Fthrust , nlf , ϕ]T ; – control vector: u = [ Fthrust , CL , ϕ]T ;
4 uncertain parameters: CD0 , k c , S; uncertain parameters: m(t), CD0 , k c , S;
method: robust adaptive controller method: adaptive control with the IRM
[28,31,32] – the reference trajectory is known only to the
– each UAV has an information about the desired trajectory;
leading UAV;
5 consensus is ensured;
no consensus is ensured;
UAV type: fixed-wing aircraft
UAV type: rotating-wing (quadrotor)
[41] – single UAV; – group of UAVs;
regulation of Vg , γ, tracking ψ∗ (t); tracking the given reference trajectory;
point-mass dynamics model without kinematic relations; kinematic relations are included;
6
control vector: u = [ Fthrust , nlf , ϕ]T ; control vector: u = [ Fthrust , CL , ϕ]T ;
uncertain parameters: CD0 , k c ; uncertain parameters: m(t), CD0 , k c , S;
method: adaptive control for non-affine plants method: IRM adaptive control
[42] – mission: passing a waypoint by several UAVs;
– mission: tracking the given reference trajectory;
control vector: u = [ Fthrust , nlf , ϕ]T ;
control vector: u = [ Fthrust , CL , ϕ]T ;
parameters are known;
7 uncertain parameters: m(t), CD0 , k c , S;
method: optimal design of the behavioral approach of
method: IRM adaptive control;
decentralized control;
communication graph: directed
communication graph: undirected (ring topology)
[43] – tracking the desired trajectory on the plane
( xd (t), yd (t)); – tracking the desired trajectory in 3-D space;
model: kinematic on the plane at constant height; model: 3-D nonlinear point model;
8 control vector: u = [Vg , ϕ]T ; control vector: u = [ Fthrust , CL , ϕ]T ;
parameters: estimated by the Gaussian Process Regression uncertain parameters: m(t), CD0 , k c , S;
training procedure; communication graph: directed
– communication graph: undirected

7. Conclusions
In this paper, we have proposed a new feedback-linearized UAV model with point
mass. A linear model with an unknown value of the UAV mass was obtained via transfor-
mation. The velocity gradient method was applied to obtain a combined adaptive control
law that includes equations for estimating the unknown mass value and for tuning the pa-
rameters of a linear controller with an implicit reference model. The obtained new adaptive
algorithms were used to solve the problem of decentralized control of a UAV formation
with an unknown mass based on a consensus algorithm. Rigorous stability conditions for
control and identification are presented. Computer simulations of the obtained algorithms
for both a single UAV and a UAV formation show good performance of the proposed ap-
proach under conditions of variable UAV mass. Further research might involve considering
the influence of wind on the movement of the UAVs, as well as considering situations in
which not every UAV has access to information about the movement trajectory.

Author Contributions: Conceptualization, A.M.P.; data curation, D.G.K. and A.A.S.; formal analysis,
A.M.P.; funding acquisition, A.M.P.; investigation, D.G.K. and A.A.S.; methodology, A.M.P.; project
administration, A.M.P.; software, A.M.P., D.G.K. and A.A.S.; supervision, A.M.P.; writing—original
draft, A.M.P. and B.A.; writing—review and editing, A.M.P. and B.A. All authors have read and
agreed to the published version of the manuscript.
Electronics 2022, 11, 4187 29 of 33

Funding: This work was carried out in BSTU “VOENMEH” with financial support from the Ministry
of Science and Higher Education of the Russian Federation (government contract agreement No.
075-03-2020-045/2 of 9 June 2020).
Data Availability Statement: Not applicable.
Conflicts of Interest: The authors declare no conflict of interest.

Abbreviations
The following abbreviations are used in this manuscript:

DMRRAC Distributed Model Reference Robust Adaptive Controller

IRM Implicit Reference Model
MRAC Model Reference Adaptive Control
SG Speed gradient
UAV Unmanned Aerial Vehicle
Nomenclature
x, y, h UAV translational coordinates
γ flight-path angle
ψ heading angle
ϕ bank angle
Vg UAV ground speed
Flift lift force
Fdrag drag force
Fthrust thrust force
m UAV mass
mmin , mmax UAV mass m ∈ [mmin , mmax ]
g gravity acceleration constant
S wing platform area
ρa air density
CL lift coefficient
CD drag coefficient
CD0 zero-lift drag coefficient
kc induced drag factor
nlf load factor
u vector of control signals
v1 , v2 , v3 , v̄1 , v̄2 , v̄3 virtual controls
r prescribed trajectory
r x , ry , rh reference coordinates along the corresponding axes
e x1 , ey1 , eh1 coordinate errors
exyh vector of coordinate errors
e x2 , ey2 , eh2 speed errors
y x , yy , yh implicit reference model outputs
yxyh vector of (y x , yy , yh )
α x , αy , αh implicit reference model parameters
Qt goal function
QtF goal function for the group of UAVs
k x , ky , kh fixed control parameters
m̂ estimated mass
k̂ x , k̂ y , k̂ h adjustable controller gains
m̂∗ , k̂ x∗ , k̂ y∗ , k̂ h∗ stabilizing controller gains
m0 , k̂ x0 , k̂ y0 , k̂ h0 “guessed” controller gains
γm , γ x , γy , γ h adaptation gains
σm , σx , σy , σh parametric feedback gains
N number of UAVs (agents)
i index of UAVs i = 1, 2, . . . , N
δi relative position vector for i-th UAV
Electronics 2022, 11, 4187 30 of 33

δxi , δyi , δhi relative offset of i-th UAV in a group for corresponding axes
G communication graph
A adjacency matrix
L Laplace matrix
v̄1ci , v̄2ci , v̄3ci control signals based on the consensus algorithm
ρ0 , k xc , k yc , k hc , ρ x , ρy , ρh consensus algorithm parameters
IN N × N identity matrix
d x , dy , dh external perturbations
t time
Subsripts
UAV translational coordinates in the appropriate normal
x, y, h
Earth’s reference frame
m UAV mass
i index of the UAV, i = 1, 2, . . . , N

References
1. Du, N.; Zhou, Y.; Deng, W.; Luo, Q. Improved chimp optimization algorithm for three-dimensional path planning problem.
Multimed. Tools Appl. 2022, 81, 27397–27422. [CrossRef]
2. Sefati, S.; Halunga, S.; Farkhady, R. Cluster selection for load balancing in flying ad hoc networks using an optimal low-energy
adaptive clustering hierarchy based on optimization approach. Aircr. Eng. Aerosp. Technol. 2022, 94, 1344–1356. [CrossRef]
3. Trujillo, M.; Becerra, H.; G?mez-Guti?rrez, D.; Ruiz-Le?n, J.; Ram?rez-Trevi?o, A. Hierarchical task-based formation control and
collision avoidance of UAVs in finite time. Eur. J. Control 2021, 60, 48–64. [CrossRef]
4. Luo, L.; Wang, X.; Ma, J.; Ong, Y. GrpAvoid: Multigroup Collision-Avoidance Control and Optimization for UAV Swarm.
IEEE Trans. Cybern. 2021 . [CrossRef] [PubMed]
5. Muslimov, T.; Munasypov, R. Multi-UAV cooperative target tracking via consensus-based guidance vector fields and fuzzy
MRAC. Aircr. Eng. Aerosp. Technol. 2021, 93, 1204–1212. [CrossRef]
6. Asami, K.; Bai, Y.; Svinin, M.; Hatayama, M. Survivor searching in a dynamically changing flood zone by multiple unmanned
aerial vehicles. Artif. Life Robot. 2022, 27, 292–299. [CrossRef]
7. Islam, S.; DIas, J.; Sunda-Meya, A. Distributed Tracking Synchronization Protocol for a Networked of Leader-follower Unmanned
Aerial Vehicles with Uncertainty. In Proceedings of the Industrial Electronics Conference (IECON 2021), Toronto, Canada, 13–16
October 2021. [CrossRef]
8. Cucker, F.; Smale, S. Emergent Behavior in Flocks. IEEE Trans. Autom. Control 2007, 52, 852–862. [CrossRef]
9. Song, Y.; Gu, M.; Choi, J.; Oh, H.; Lim, S.; Shin, H.S.; Tsourdos, A. Using Lazy Agents to Improve the Flocking Efficiency of
Multiple UAVs. J. Intell. Robot. Syst. Theory Appl. 2021, 103. [CrossRef]
10. Zhou, P.; Chen, B. Semi-global leader-following consensus-based formation flight of unmanned aerial vehicles. Chin. J. Aeronaut.
2022, 35, 31–43. [CrossRef]
11. Ahmed, A.; Naeem, M.; Al-Dweik, A. Joint Optimization of Sensors Association and UAVs Placement in IoT Applications With
Practical Network Constraints. IEEE Access 2021, 9, 7674–7689. [CrossRef]
12. Chen, L.; Zhao, R.; He, K.; Zhao, Z.; Fan, L. Intelligent ubiquitous computing for future UAV-enabled MEC network systems.
Clust. Comput. 2022, 25, 2417–2427. [CrossRef]
13. Mahmood, A.; Vu, T.; Khan, W.U.; Chatzinotas, S.; Ottersten, B. Optimizing Computational and Communication Resources for
MEC Network Empowered UAV-RIS Communication. Preprint 2022.
14. Lin, Z.; Lin, M.; Wang, J.B.; de Cola, T.; Wang, J. Joint Beamforming and Power Allocation for Satellite-Terrestrial Integrated
Networks With Non-Orthogonal Multiple Access. IEEE J. Sel. Top. Signal Process. 2019, 13, 657–670. [CrossRef]
15. Güzey, H.; Güzey, N. Adaptive hybrid formation-search and track controller of UAVs. Int. J. Syst. Sci. 2022, 53, 2301–2317.
[CrossRef]
16. Mahfouz, M.; Hafez, A.; Ashry, M.; Elnashar, G. Target Assignment for Cooperative Quadrotors Unmanned Aerial Vehicles.
J. Phys. Conf. Ser. 2021, 2128. [CrossRef]
17. Han, Q.; Zhou, Y.; Liu, X.; Tuo, X. Event-Triggered Finite-Time Attitude Cooperative Control for Multiple Unmanned Aerial
Vehicles. Appl. Bionics Biomech. 2022, 2022, 5875004. [CrossRef]
18. Wu, X.; Wu, Y.; Xu, Z.; Chen, Q. Fixed-time flocking formation of nonlinear multi-agent system with uncertain state perturbation.
Int. J. Control. 2022. [CrossRef]
19. Gong, J.; Jiang, B.; Ma, Y.; Mao, Z. Distributed Adaptive Fault-Tolerant Formation Control for Heterogeneous Multiagent Systems
With Communication Link Faults. IEEE Trans. Aerosp. Electron. Syst. 2022, 1–11. [CrossRef]
20. Popov, A.M.; Kostin, I.; Fadeeva, J.; Andrievsky, B. Development and Simulation of Motion Control System for Small Satellites
Formation. Electronics 2021, 10, 3111. [CrossRef]
Electronics 2022, 11, 4187 31 of 33

21. Popov, A.M.; Kostrygin, D.G.; Krashanin, P.V.; Shevchik, A.A. Development of Algorithm for Guiding the Swarm of Unmanned
Aerial Vehicles. In Proceedings of the 29th Saint Petersburg International Conference on Integrated Navigation Systems (ICINS
2022), Saint Petersburg, Russia, 30 May–1 June 2022; Volume 1, pp. 1–4. [CrossRef]
22. Gamagedara, K.; Lee, T. Geometric Adaptive Controls of a Quadrotor Unmanned Aerial Vehicle with Decoupled Attitude
Dynamics. J. Dyn. Syst. Meas. Control. ASME 2022, 144, 031002. [CrossRef]
23. Chatterjee, B.; Dutta, R. Studying the Effect of Network Latency on an Adaptive Coordinated Path Planning Algorithm for UAV
Platoons. In Proceedings of the 2022 8th Workshop on Micro Aerial Vehicle Networks, Systems, and Applications (DroNet 2022),
Part of MobiSys 2022, Portland, OR, USA, 27 June–1 July 2022; pp. 7–12. [CrossRef]
24. Ullah, N.; Mehmood, Y.; Aslam, J.; Wang, S.; Phoungthong, K. Fractional order adaptive robust formation control of multiple
quad-rotor UAVs with parametric uncertainties and wind disturbances. Chin. J. Aeronaut. 2022, 35, 204–220. [CrossRef]
25. Liang, W.; Chen, Z.; Yao, B. Geometric Adaptive Robust Hierarchical Control for Quadrotors With Aerodynamic Damping and
Complete Inertia Compensation. IEEE Trans. Ind. Electron. 2022, 69, 13213–13224. [CrossRef]
26. Zhi, Y.; Liu, L.; Guan, B.; Wang, B.; Cheng, Z.; Fan, H. Distributed robust adaptive formation control of fixed-wing UAVs with
unknown uncertainties and disturbances. Aerosp. Sci. Technol. 2022, 126, 107600. [CrossRef]
27. Huang, Y.; Meng, Z. Bearing-Based Distributed Formation Control of Multiple Vertical Take-Off and Landing UAVs. IEEE Trans.
Control Netw. Syst. 2021, 8, 1281–1292. [CrossRef]
28. Andrievsky, B.; Tomashevich, S. Passification based signal-parametric adaptive controller for agents in formation. IFAC-
PapersOnLine 2015, 48, 222–226. [CrossRef]
29. Andrievskii, B.; Fradkov, A. Method of passification in adaptive control, estimation, and synchronization. Autom. Remote Control
2006, 67, 1699–1731. [CrossRef]
30. Andrievskii, B.R.; Selivanov, A.A. New Results on the Application of the Passification Method. A Survey. Automat. Remote
Control 2018, 79, 957–995. [CrossRef]
31. Tomashevich, S.; Andrievsky, B. Adaptive control of quadrotors spatial motion in formation with implicit reference model.
AIP Conf. Proc. 2018, 2046, 020103. [CrossRef]
32. Tomashevich, S.; Andrievsky, B. High-order adaptive control in multi-agent quadrotor formation. Math. Eng. Sci. Aerosp. 2019,
10, 681–693.
33. Furtat, I.B. Simple Adaptive Algorithm for Plants with Input Delay and Disturbances. IFAC-PapersOnLine 2017, 50, 4270–4275.
[CrossRef]
34. Tomashevich, S.I. Control for a system of linear agents based on a high order adaptation algorithm. Autom. Remote Control 2017,
78, 276–288. [CrossRef]
35. Tomashevich, S.; Belyavskyi, A. 2DoF indoor testbed for quadrotor identification and control. In Proceedings of the 23rd Saint
Petersburg International Conference Integrated Navigation Systems (ICINS 2016), Saint Petersburg, Russia, 30 May–1 June 2016;
pp. 373–376.
36. Andrievsky, B.; Popov, A.M.; Kostin, I.; Fadeeva, J. Modeling and Control of Satellite Formations: A Survey. Automation 2022,
3, 511–544. [CrossRef]
37. Andrievsky, B.; Fradkov, A.; Kudryashova, E. Control of two satellites relative motion over the packet erasure communication
channel with limited transmission rate based on adaptive coder. Electronics 2020, 9, 2032. [CrossRef]
38. Kuznetsov, N.V.; Andrievsky, B.; Kudryashova, E.V.; Kuznetsova, O.A. Stability and hidden oscillations analysis of the spacecraft
attitude control system using reaction wheels. Aerosp. Sci. Technol. 2022, 131, 107973. [CrossRef]
39. Menon, P.K.A. Short-range nonlinear feedback strategies for aircraft pursuit-evasion. J. Guid. Control. Dyn. 1989, 12, 27–32.
[CrossRef]
40. Menon, P.K.; Sweriduk, G.D.; Sridhar, B. Optimal Strategies for Free-Flight Air Traffic Conflict Resolution. J. Guid. Control. Dyn.
1999, 22, 202–211.
41. Boskovic, J.D.; Chen, L.; Mehra, R.K. Adaptive Control Design for Nonaffine Models Arising in Flight Control. J. Guid. Control.
Dyn. 2004, 27, 209–217.
42. Kim, S.; Kim, Y. Optimum design of three-dimensional behavioural decentralized controller for UAV formation flight. Eng.
Optim. 2009, 41, 199–224. [CrossRef]
43. Wang, J.; Xin, M. Integrated Optimal Formation Control of Multiple Unmanned Aerial Vehicles. IEEE Trans. Control Syst. Technol.
2013, 21, 1731–1744. [CrossRef]
44. Beard, R.W.; McLain, T.W. Small Unmanned Aircraft: Theory and Practice; Princeton University Press: Princeton, NJ, USA, 2012.
45. Dobrokhodov, V. Kinematics and Dynamics of Fixed-Wing UAVs. In Handbook of Unmanned Aerial Vehicles; Springer Netherlands:
Dordrecht, The Netherlands, 2015; pp. 243–277. [CrossRef]
46. Miele, A. Flight Mechanics: Theory of Flight Paths; Dover Publications Inc.: New York, NY, USA, 2016.
47. Zhao, Y.; Tsiotras, P. Time-Optimal Path Following for Fixed-Wing Aircraft. J. Guid. Control. Dyn. 2013, 36, 83–95.
[CrossRef]
48. Anderson, M.; Robbins, A. Formation flight as a cooperative game. In Proceedings of the Guidance, Navigation, and Control Conference
and Exhibit; AIAA: Boston, MA, USA, 1998; p. 244. [CrossRef]
49. Khalil, H.K. Nonlinear Systems, 3rd ed.; Prentice-Hall: Upper Saddle River, NJ, USA, 2002.
Electronics 2022, 11, 4187 32 of 33

50. Fliess, M.; Lévine, J.; Martin, P.; Rouchon, P. Flatness and defect of non-linear systems: Introductory theory and examples.
Int. J. Control 1995, 61, 1327–1361.
51. Lévine, J. Analysis and Control of Nonlinear Systems A Flatness-Based Approach; Springer: Berlin/Heidelberg, Germany, 2009.
[CrossRef]
52. Fradkov, A.L. Speed-gradient scheme and its application in adaptive control problems. Autom. Remote Control 1980, 40, 1333–1342.
53. Fradkov, A.L.; Miroshnik, I.V.; Nikiforov, V.O. Nonlinear and Adaptive Control of Complex Systems; Kluwer: Dordrecht, The
Netherlands, 1999.
54. Andrievsky, B.R.; Fradkov, A.L. Speed Gradient Method and Its Applications. Autom. Remote Control 2021, 82, 1463–1518.
[CrossRef]
55. Fradkov, A.L.; Andrievsky, B. Speed-Gradient Method in Mechanical Engineering. Adv. Struct. Mater. 2022, 164, 171–194.
[CrossRef]
56. Fradkov, A.L. Synthesis of an adaptive system for linear plant stabilization. Autom. Remote Control 1974, 35, 1960–1966.
57. Fradkov, A. Lyapunov-Bregman functions for speed-gradient adaptive control of nonlinear time-varying systems.
IFAC-PapersOnLine 2022, 55, 544–548. [CrossRef]
58. Fradkov, A.; Andrievsky, B. Adaptive Synchronization of Time-Varying Non-linear Systems with Application to Signal Transmis-
sion. Stud. Syst. Decis. Control 2022, 414, 325–344. [CrossRef]
59. Fradkov, A.; Tomashevich, S.; Andrievsky, B.; Amelin, K.; Kaliteevskiy, I. Adaptive Coding For Data Exchange Between
Quadrotors In The Formation. IFAC-PapersOnLine 2016, 49, 275–280. [CrossRef]
60. Marantos, P.; Koveos, Y.; Kyriakopoulos, K. UAV State Estimation Using Adaptive Complementary Filters. IEEE Trans. Control
Syst. Technol. 2016, 24, 1214–1226. [CrossRef]
61. Yu, Y.; Peng, S.; Li, Q.; Dong, X.; Ren, Z. Cooperative Navigation Method Based on Adaptive CKF for UAVs in GPS Denied Areas.
In Proceedings of the IEEE CSAA Guidance, Navigation and Control Conference, CGNCC 2018, Xiamen, China, 10–12 August
2018. [CrossRef]
62. Silantyev, A.; Tereshchenko, D.; Kazakov, L.; Selyanskaya, E. Adaptive system of mutual positioning for controlling the groups
of unmanned aerial vehicles. In Proceedings of the 2018 Systems of Signal Synchronization, Generating and Processing in
Telecommunications (SYNCHROINFO 2018), Minsk, Belarus, 4–5 July 2018. [CrossRef]
63. Tomashevich, S.; Andrievsky, B.; Fradkov, A. Formation control of a group of unmanned aerial vehicles with data exchange
over a packet erasure channel. In Proceedings of the 2018 IEEE Industrial Cyber-Physical Systems (ICPS 2018), Saint Petersburg,
Russia, 15–18 May 2018; pp. 38–43. [CrossRef]
64. Zhong, W.; Xu, L.; Liu, X.; Zhu, Q.; Zhou, J. Adaptive beam design for UAV network with uniform plane array. Phys. Commun.
2019, 34, 58–65. [CrossRef]
65. Darabkh, K.; Alfawares, M.; Althunibat, S. MDRMA: Multi-data rate mobility-aware AODV-based protocol for flying ad-hoc
networks. Veh. Commun. 2019, 18. [CrossRef]
66. Tomashevich, S.; Andrievsky, B.; Dokuchaeva, A.; Emelyanov, V. Navigation data exchange between UAVs in the formation by
means of the adaptive coding procedure. Math. Eng. Sci. Aerosp. 2019, 10, 463–478.
67. Amelin, K.; Andrievsky, B.; Tomashevich, S.; Fradkov, A. Data Exchange with Adaptive Coding between Quadrotors in a
Formation. Autom. Remote Control 2019, 80, 150–163. [CrossRef]
68. Lim, S.; Song, Y.; Choi, J.; Myung, H.; Lim, H.; Oh, H. Decentralized Hybrid Flocking Guidance for a Swarm of Small UAVs.
In Proceedings of the 2019 International Workshop on Research, Education and Development on Unmanned Aerial Systems,
RED-UAS 2019, Cranfield, UK, 25–27 November 2019; pp. 287–296. [CrossRef]
69. Lin, Z.; An, K.; Niu, H.; Hu, Y.; Chatzinotas, S.; Zheng, G.; Wang, J. SLNR-based Secure Energy Efficient Beamforming in
Multibeam Satellite Systems. IEEE Trans. Aerosp. Electron. Syst. 2022, 1–4. [CrossRef]
70. Lin, Z.; Lin, M.; de Cola, T.; Wang, J.B.; Zhu, W.P.; Cheng, J. Supporting IoT With Rate-Splitting Multiple Access in Satellite and
Aerial-Integrated Networks. IEEE Internet Things J. 2021, 8, 11123–11134. [CrossRef]
71. Lin, Z.; Niu, H.; An, K.; Wang, Y.; Zheng, G.; Chatzinotas, S.; Hu, Y. Refracting RIS-Aided Hybrid Satellite-Terrestrial Relay
Networks: Joint Beamforming Design and Optimization. IEEE Trans. Aerosp. Electron. Syst. 2022, 58, 3717–3724. [CrossRef]
72. Geraci, G.; Garcia-Rodriguez, A.; Azari, M.M.; Lozano, A.; Mezzavilla, M.; Chatzinotas, S.; Chen, Y.; Rangan, S.; Renzo, M.D.
What Will the Future of UAV Cellular Communications Be? A Flight From 5G to 6G. IEEE Commun. Surv. Tutorials 2022,
24, 1304–1335. [CrossRef]
73. Dovgal, V. Making decisions about the placement of unmanned aerial vehicles based on the implementation of an artificial
immune system in relation to information processing. In Proceedings of the International Conference on Industrial Engineering,
Applications and Manufacturing (ICIEAM 2021), Sochi, Russia, 17–21 May 2021; pp. 828–833. [CrossRef]
74. Wu, P.; Xiao, F.; Huang, H.; Sha, C.; Yu, S. Adaptive and Extensible Energy Supply Mechanism for UAVs-Aided Wireless-Powered
Internet of Things. IEEE Internet Things J. 2020, 7, 9201–9213. [CrossRef]
75. Wei, X.; Yang, J.; Fan, X. Fully distributed guidance laws for unmanned aerial vehicles formation flight. Trans. Inst. Meas. Control
2020, 42, 965–980. [CrossRef]
76. Ren, W.; Beard, R.W. Distributed Consensus in Multi-Vehicle Cooperative Control; Springer: London, UK, 2008. [CrossRef]
77. Olfati-Saber, R.; Murray, R. Consensus problems in networks of agents with switching topology and time-delays. IEEE Trans.
Autom. Control 2004, 49, 1520–1533. [CrossRef]
Electronics 2022, 11, 4187 33 of 33

78. Olfati-Saber, R.; Fax, J.A.; Murray, R.M. Consensus and Cooperation in Networked Multi-Agent Systems. Proc. IEEE 2007,
95, 215–233. [CrossRef]
79. Kabiri, M.; Atrianfar, H.; Menhaj, M. Formation control of VTOL UAV vehicles under switching-directed interaction topologies
with disturbance rejection. Int. J. Control 2018, 91, 33–44. [CrossRef]