applied

sciences
Article
Adaptive Sliding Mode Control for a Robotic Manipulator with
Unknown Friction and Unknown Control Direction
Seung-Hun Han 1 , Manh Son Tran 2 and Duc-Thien Tran 2, *

1 Mechanical System Engineering Department, Tongyeong Campus, Gyeongsang National University,

Tongyeong 53064, Korea; [email protected]
2 Automatic Control Department, Ho Chi Minh City University of Technology and Education,
Ho Chi Minh City 700000, Vietnam; [email protected]
* Correspondence: [email protected]; Tel.: +84-988-862-588

Abstract: This paper is aimed at addressing the tracking control issue for an n-DOF manipulator
regardless of unknown friction and unknown control direction. In order to handle the above issues,
an adaptive sliding mode control (ASMC) is developed with a Nussbaum function. The sliding
mode control (SMC) in the proposed control guarantees the tracking problem and fast responses for
the manipulator. Additionally, there are adaptive laws for the robust gain in the SMC to deal with
the unknown external disturbance and reduce the chattering effect in the system. In practice, the
mistakes in the connection between actuators and drivers, named unknown control direction, cause
serious damage to the manipulator. To overcome this issue, the Nussbaum function is multiplied
by the ASMC law. A Lyapunov approach is investigated to analyze the stability and robustness
of the whole system. Finally, several simulations are implemented on a 3-DOF manipulator and




their results are compared with those of the existing controllers to validate the advantages of the

 proposed method.
Citation: Han, S.-H.; Tran, M.S.; Tran,
D.-T. Adaptive Sliding Mode Control Keywords: sliding mode control; manipulator; unknown control direction; Nussbaum function;
for a Robotic Manipulator with Lyapunov approach
Unknown Friction and Unknown
Control Direction. Appl. Sci. 2021, 11,
3919. https://doi.org/10.3390/
app11093919 1. Introduction
For the past few decades, manipulators have been widely applied in many areas,
Academic Editor: Daniel Vera
such as drilling [1], deburring [2], and object grasping [3,4]. Due to large intrinsic non-
linearities, modeling errors, high couplings, and substantial parametric uncertainties in
Received: 6 February 2021
robot dynamics, control design for the manipulator has continued to be an exciting task for
Accepted: 19 April 2021
Published: 26 April 2021
researchers. In order to obtain reasonable control performance, several robust advanced
control methods have been developed, such as adaptive control, backstepping control [5,6],
sliding mode control [7–10], and neural control [6,7].
Publisher’s Note: MDPI stays neutral
with regard to jurisdictional claims in
Sliding mode control is well known as a robust control for nonlinear systems to deal
published maps and institutional affil-
with uncertainties [11]. The principle of SMC is based on discontinuous control to force
iations. the control system’s error state variable toward zero. Therefore, this method has been
developed in many application fields [12]. However, the challenge of this method is the
chattering effect which is stimulated in situations of large control gain. This phenomenon
can excite high-frequency dynamics and cause instability in the controlled system [13]. In
order to overcome this issue, some methods, such as boundary layer methods [13–15] and
Copyright: © 2021 by the authors.
disturbance approximators [16–18], have been integrated into SMC. In [13–15], the sign
Licensee MDPI, Basel, Switzerland.
This article is an open access article
function in SMC was swapped with the sigmoid function, saturation function, and fuzzy
distributed under the terms and
logic system (FLS). Since these methods operate as low-pass filters, the chattering reduction
conditions of the Creative Commons and robustness should be traded off in these situations. Disturbance approximators [16–18]
Attribution (CC BY) license (https:// were used to estimate the disturbance and uncertainties and indirectly decrease the mag-
creativecommons.org/licenses/by/ nitude of the switching gain control and chattering phenomenon. In other approaches,
4.0/). switching gain adaptive laws [11,19] have been applied to SMC to adjust the switching

Appl. Sci. 2021, 11, 3919. https://doi.org/10.3390/app11093919 https://www.mdpi.com/journal/applsci

Appl. Sci. 2021, 11, 3919 2 of 14

gain regarding the uncertainties and disturbances without knowledge of them. In recent
years, the unknown control direction [20–22], which is the wrong connection in the actuator,
has been interesting to many researchers. The Nussbaum gain technique is well known
as a control gain estimator for handling the unknown control directions and it was first
presented in 1983 by Nussbaum in Systems and Control Letters. In [23], a fault-tolerant
control conducted based on feedback linearization is combined with a Nussbaum gain
function and an online parameter estimator for a hypersonic vehicle. In [24], a robust
adaptive and fault-tolerant control scheme was developed from a novel PID-based fault-
tolerant controller with a Nussbaum-type function for a wind turbine. In [25], a command
filter-based adaptive tracking controller was designed based on FLS, a command filter, and
Nussbaum functions for a class of nonlinear systems with unknown control directions.
In [26], Wang et al. developed an adaptive consensus tracking control based on some Nuss-
baum functions to manage the unknown control directions and the time-varying actuator
faults simultaneously. In [27], Liu Y., etc. provided an adaptive fuzzy prescribed perfor-
mance tracking control for a nonlinear switched stochastic system with unknown control
directions. In this study, the adaptive controller also used a Nussbaum-type function to
handle the control direction problem. From this literature, there has not been any study
integrating the unknown control direction issue into the SMC design for a manipulator.
Therefore, this is one of the motivations for our study.
This paper presents a novel adaptive SMC for an n-degree of freedom (DOF) robotic
manipulator regardless of the existence of the unknown friction and unknown control
direction. The proposed control is designed based on the SMC, adaptive switching gain
laws, and Nussbaum function. The adaptive switching gain laws are investigated to
regulate the robust gain of the SMC online. This approach helps in not only reducing the
chattering effect but also deal with the time-varying external disturbance in the system.
Moreover, a Nussbaum function is integrated into the adaptive SMC to manage the wrong
control connection. Additionally, the stability and robustness of the whole system with
the proposed control are analyzed by a Lyapunov theorem. Finally, some simulations
are implemented on a 3-DOF manipulator and their results are compared with those of
previous methods to demonstrate the superiority of the proposed control.
The rest of the paper is as follows: In Section 2, robotic dynamics and the preliminary
are exhibited. In Section 3, the design procedures of the ASMC and the proposed control
for the manipulator are presented and analyzed. Several simulations are presented in
Section 4. Finally, several conclusions and future works are discussed in Section 5.

2. Problem Formulation and Preliminary

2.1. Manipulator Dynamics Description
The n-DOF manipulator dynamics are presented by:
.. . .
M(θ)θ + C(θ, θ)θ + G(θ) + d(t) = Bτ (1)
. ..
where θ, θ, θ ∈ Rn×1 , respectively, presents angular position, angular velocity, and angular
acceleration vectors of each joint, M(θ) ∈ Rn×n is the symmetric and positive definite
.
matrix, C(θ, θ) ∈ Rn×n is the Coriolis and centrifugal matrix, G(θ) ∈ Rn×1 is the gravity
vector, τ derives the torque vector, B ∈ Rn× n presents the unknown control direction
. .
which is a diagonal matrix, M(θ) − 2C θ, θ is a skew-symmetric matrix, defined as
h .  . i
XT M(θ) − 2C θ, θ X = 0, and d(t) expresses the disturbance.
.
When we define x1 = θ ∈ Rn and x2 ∈ θ ∈ Rn , (1) can be rewritten as follows:
.
x1 = x2
. (2)
x2 = M−1 (x1 )(Bu − C(x1 , x2 )x2 − G(x1 ) − d(t))

where u = τ expresses the control input vector.

Appl. Sci. 2021, 11, 3919 3 of 14

.
Assumption 1. The joint trajectoriesx1d ∈ Rn×n , x2d ∈ Rn×n , and x2d ∈ Rn×n are bounded by
.
predefined thresholds as kx1d k∞ < δ1 , kx2d k∞ < δ2 , and kx2d k∞ < δ3 where δi (i = 1, 2, 3) are
positive constants.
Assumption 2. We suppose that the external disturbance,d(t), is bounded and the inequality
kd(t)k ≤ δ holds.
Control objective: Provide an advanced control for a manipulator under the presence
of unknown friction and unknown control direction to not only guarantee the tracking
requirements but also reduce the chattering effect in the control signal.

2.2. Nussbaum Function

A function N (ζ ) can be called a Nussbaum-like function if it satisfies the following
properties [28]:
Zs
lim sup N (ζ )dζ = +∞ (3)
s→∞
s0

Zs
lim inf N (ζ )dζ = −∞ (4)
s→∞
s0
2 π


In this study, the Nussbaum function N (ζ ) = eζ cos 2ζ is considered.
h 
Lemma 1 ([29,30]). Let V (t) and ζ i (t) be smooth functions defined on 0, t f with V (t) > 0, for
h 
any t ∈ 0, t f , if the following inequality is satisfied:

Zt n .
V ( t ) < c0 ± ∑ ( gi N (ζ i ) ± 1)ζ i dτ (5)
0 i =1

where gi is a non-zero constant and c0 expresses some suitable constant, then V (t) ζ i (t) and
Rt n . h 
0 ∑ ( gi N ( ζ i ) ± 1) ζ i dτ must be bounded on 0, t f .
i =1

3. Control Design
3.1. Adaptive Sliding Mode Control

Remark 1. This section presents an ASMC for the robotic manipulator under the existence of
disturbance. The unknown control direction does not happen in this case. Therefore, B = In×n .

e = x1 − x1d is the trajectory tracking error and xid presents the desired trajectory. A
sliding mode surface is selected as follows:
.
s = e + λe (6)

where λ = diag(λ1 , λ2 , . . . , λn ) ∈ Rn×n , (i = 1, 2) are positive definite matrices.

The time derivative of the sliding surface is computed as
. .. .
s = e + λe (7)

Replacing (2) into (7), it is rewritten as follows:

. . ..
s = λe + M−1 (x1 )(Bu − C(x1 , x2 )x2 − G(x1 ) − d(t)) − x1d (8)
Appl. Sci. 2021, 11, 3919 4 of 14

Based on assumption 2 and the sliding mode design procedure, an SMC can be
designed as
u = ueq + ur (9)
where
.. .
.
ueq = B−1 M(x1 ) x1d −λe + C(x1 , x2 ) x1d − λe + G(x1 )−ks



(10)
is an equivalent control with k = diag([k1 , . . . , k n ]) ∈ Rn×n , and

ur = −B−1 ηsign(s) (11)

is used to ensure the stability of the system with the existence of the uncertainties, η =
diag([η1 , . . . , ηn ]) ∈ Rn×n expresses a diagonal positive matrix which is chosen for the
condition ηi ≥ |δi |. The function is sign(x) = [sign( x1 ), . . . , sign( xn )] T ∈ Rn×1 . In practice,
the external disturbance usually changes with respect to time, and the robust gains can
cause the chattering effect when the robust gains are significantly larger than the external
disturbance. In order to overcome this disadvantage, fast adaptive laws are developed to
adjust the robust gain online with respect to the external disturbance. The robust control
(11) is replaced by the equation below:

^
ur = −B−1 ηsign(s) (12)

The adaptive laws are selected as follows:

  γ(t)
−1
 κ κ
 1i 2i i
 | s | γ(t) , if 0 < η̂i , min |si | ≥ ε 2
.  i
η̂ i = κ1i κ2i |si |−1 if η̂ = 0, min |si | ≥ ε 2 (13)


 i
0 otherwise


where γ(t) = sign(|si | − ε 1 ), κ ji ( j = 1, 2; i = 1, . . . , n) are positive constants, ε 1 are prede-

fined error boundaries, ε 1 = 10ε 2 ≥ 0.
These adaptive robust gain laws do not need knowledge of the upper boundary of the
external disturbance. The adaptive robust gains will rapidly grow and drive the sliding
surface si (i = 1, . . . , n) to the desired areas that are smaller than ε 1i . Otherwise, when the
sliding variable stays in these regions, the gains will gradually reduce. From the property
of the adaptive robust gain laws, the chattering effect is significantly reduced under the
existence of the time-varying external disturbance.

Theorem 1. When we consider the external disturbance in the robot manipulator with the known
control direction described in (2), the sliding mode surface presented in (6), and the controller
defined in (9), then the sliding variables,si , are bounded. Additionally, the chattering effect is
reduced significantly.

Proof. The Lyapunov function is selected as follows:

1 T 1 ~T ~
V1 = s M(x1 )s + η κ1−1 κ2 η (14)
2 2
where ηei = ηi∗ − η̂i , (i = 1, . . . , n) are robust gain errors.
~T ~
Assumption 3. The robust gain error term (14) is bounded. The condition 21 η κ1−1 κ2 η ≤ ζ η holds.
Taking the derivative of the Lyapunov function, V1 , with respect to time, we have
. . .
. ~T ~
V 1 = sT M(x1 )s + 12 sT M(x1 )s + η κ1−1 κ2 η
. . (15)
. .. ~T ~
= sT M(x1 ) λe + M−1 (x1 )(Bu − C(x1 , x2 )x2 − G(x1 ) − d1 (t)) −xd + 21 sT M(x1 )s + η κ1−1 κ2 η



Appl. Sci. 2021, 11, 3919 5 of 14

By applying the control laws (9) and skew-matrix property in (15), the result is
presented as follows:
.
  .
^ ~T ~
V 1 = sT −ks − ηsign(s) − d(t) + η κ1−1 κ2 η
.
n n ~ ^
−1
≤ −sT ks − ∑ (η̂i − |di (t)|)|si | − ∑ ηi κ1i κ2i ηi (16)
i =1  i =1
n  . 
2 − 1
≤−∑ k i si − ηei |si | − κ1i κ2i η̂ i
i =1

Now, we take into account two adaptive cases: ksk∞ ≥ ε 1 and ksk∞ < ε 1 . In the first
case, ksk∞ ≥ ε 1 , the differential Lyapunov function is presented as follows:

. n n
V 1 = − ∑ k i s2i ≤ − ∑ k i ε21 (17)
i =1 i =1

This result shows that V1 is decreasing and bounded since 0 ≤ V1 (t) ≤ V1 (0) ≤
∞. It means that the sliding variable si reaches the small region of the sliding surface,
i.e., |si | ≤ ε 1 .
In the other cases, ksk∞ ≤ ε 1 , it means that all sliding variables si approach the regions
.
≤
| i | ε 1 , and they may move in or leave the regions because V 1 is not guaranteed to be
s
semi-negative in the vicinity of the sliding manifold. If any sliding variable si moves out of
.
the region |si | ≤ ε 1 , then V 1 will become negative again. This result will force the variable
to go back toward the sliding surfaces.
When all the sliding variables, si , stay in the region |si | ≤ ε 1 , the upper bound of the
sliding variable is computed as follows:

1 T 1 1~ ~
s M(x1 )s ≤ V1 ≤ sT M(x1 )s + ηκ1−1 κ2 η (18)
2 2 2
~ ~
Since 12 ηκ1−1 κ2 η in (18) is bounded, we have

1 1
V1 < λmax ( M ( x1 ))nε2 + ζ η (19)
2 2
Based on (18) and (19), we obtain
n
1
2 λmin ( M ( x1 )) ∑ s2i < 12 λmax ( M ( x1 ))nε21 + 12 ζ η
i =1 (20)
n
λmax ( M ( x1 )) 2 ζη
∑ s2i < λmin ( M( x1 ))
nε 1 + λmin ( M ( x1 ))
i =1

This means that

s s
n
λmax ( M ( x1 )) 2 ζη
∑ s2i < λmin ( M( x1 ))
nε 1 +
λmin ( M ( x1 ))
(21)
i =1

Equation (21) means that sliding variables, si , are ultimately uniform bounded. The
upper boundary in (21) can be adjusted by the parameters κ1 , κ2 , ε 1 . 

3.2. Proposed Control

The schematic of the proposed control presented in Figure 1 is developed from the
adaptive SMC. Therefore, the proposed control can deal with the external disturbance
and chattering effect. Moreover, a Nussbaum function is added in the adaptive SMC to
overcome the unknown control direction in the robot system.
 upper boundary in (21) can be adjusted by the parameters κ1 , κ 2 , ε1 .□

3.2. Proposed Control

The schematic of the proposed control presented in Figure 1 is developed from the
Appl. Sci. 2021, 11, 3919 adaptive SMC. Therefore, the proposed control can deal with the external disturbance and
6 of 14
chattering effect. Moreover, a Nussbaum function is added in the adaptive SMC to over-
come the unknown control direction in the robot system.

Proposed control
External
disturbance
Adaptive robust
gain laws
n-DOF Manipulator

Unknown
Trajectory Sliding Sliding mode Nussbaum Manipulator
control
planning surface control function dynamics
direction

Adaptive sliding mode control

Figure 1.
Figure 1. Schematic of
of the
the proposed
proposed control.
control.

Remark
Remark 2.2. The
The proposed
proposed control
control is
is developed
developed to
to overcome
overcome the
the unknown
unknown control
control direction
direction which
which
means
means that
that the
the signal
signal of
of the
the control
control input
input gains
gains can
can be
be nonpositive.
nonpositive.

The
The proposed
proposed control
control is
is chosen
chosen as
as follows:
follows:

u(t) u=( t )N=(N

ζ )(u
ζ ) u∈0 ∈RRn×1 n ×1
0 (22)
(22)
u = ( M ( x1 )( 
xd ..- λe ) + .C
( x1 , x 2 )( x d − λe ). +G ( x1 )
− ks − ηˆ sign ( s ) ) ∈ R ^
n ×1
 
where u 0=
where M(x1 ) xd −λe + C(x1 , x2 ) xd − λe +G(x1 ) − ks − ηsign(s) ∈ R n ×1
0
π 
N (ζ ) = diag (
  N (ζ 1 ) ... N (ζ n )

N (ζ ) = diag N (ζ 1 ) . . . N (ζ n ) ∈R n × n
)
 ∈ R n×n , N (ζ i ) = eζ i cos
2

ζ
, N (ζ i ) = e cos
i
2  πζ i 
.
 2 2 ζ i .
. (23)
(23)
ζ i = si u0i , (i = 1, . . . , n)
ζ i = si u0i , ( i = 1,..., n )

Theorem 2. When we consider the external disturbance in the robot manipulator with unknown
Theorem 2. When we consider the external disturbance in the robot manipulator with unknown
control direction described in (2), the SM surface presented in (6), the controller defined in (22),
control direction described in (2), the SM surface presented in (6), the controller defined in (22),
and the adaptive robust gain laws provided in (13), then the sliding variables,si , are bounded and
and the adaptive robust gain laws provided in (13), then the sliding variables, si , are bounded and
the unknown control direction is overcome. Consequently, the closed-loop system is proven to be
the unknown uniformly
semi-globally control direction is overcome.
ultimately Consequently, the closed-loop system is proven to be
bounded (SGUUB).
semi-globally uniformly ultimately bounded (SGUUB).
Proof of Stability. The stability of the whole system is proven by using the Lyapunov
Proof of (14),
function stability
and the analysis is implemented as follows:
The stability
The time of the whole
derivative of the system is proven
Lyapunov by (14)
function usingis the Lyapunov
presented function (14), and
as follows:
the analysis is implemented as follows: .
. . . ~T ~
The
V1 = time
sT Mderivative
(x1 )s + 1 sTof
M(the
x1 )sLyapunov
+ η κ−1 κ2function
η (14) is presented as follows:
2 1 . (24)
. ..
. ~ T −1 ~
= sT M 1


( x ) λe − x + Bu − C ( x , x )
 ( x ) s + 1η κ2 κ dηx − λe − G ( x ) − d ( t )) + η κ κ η
V1 = s1 M ( x1 ) s +
T d s M T
1
T −
1
1
2
1 1 1 2
2 (24)
sT ( M
By replacing= the ( x1 )( λe − xcontrol
proposed d ) + Bu − C ( xinto
(22) 1 , x 2 )((24),
x d − λe
the) −result ( t ) ) + η T κas
G ( x1 ) −isd1yielded −1

1 κfollows:
2η

.
. . By
. replacing
~ T the proposed
~ control (22) into (24), the result is yielded as follows:
V1 = sT M(x1 )s + 21 sT M
(x1 )s + η κ1−1 κ2 η

. .. . .. .

= sT M(x1 ) λe − xd − C(x1 , x2 ) λe − xd − G(x1 ) − d1 (t) + BN(ζ ) M(x1 ) xd −λe
.
. ^ ~T ~
+C(x1 , x2 ) xd − λe + G(x1 ) − ks − ηsign(s) ) ) + η κ1−1 κ2 η


. ..
. ^ ^
sT M(x1 ) λe − xd − C(x1 , x2 ) xd − λe − G(x1 ) + ks + ηsign(s) − ks − ηsign(s) − d1 (t)


= (25)
.
.. .
.
^ ~ T −1 ~
+BN(ζ ) M(x1 ) xd −λe +C(x1 ,x2 ) xd − λe +G(x1 ) − ks − ηsign(s))) + η κ1 κ2 η
. .
~T
~  ~
= −sT ks − sT ηsign(s) + ε − (1 − BN(ζ ))ζ + η κ1−1 κ2 η
.
≤ −cV − (1 − BN(ζ ))ζ + δ
Appl. Sci. 2021, 11, 3919 7 of 14

h i
By multiplying two sides of (25) with ect and integrating over 0, t f , the result is
presented as follows:
  Z t .
δ δ −ct f
0 ≤ V ( t ) ≤ + V (0) − e − e−ct (1 − BN(ζ ))ζecτ dτ (26)
c c 0

R tf .
From Lemma 1 and (26), we conclude that V (t), ζ i , and 0 (1 − BN(ζ ))ζecτ dτ are
bounded and semi-globally uniformly ultimately bounded (SGUUB). 

4. Numerical Simulations
4.1. Simulation Descriptions
In this section, two simulations on a 3-DOF manipulator are implemented to show the
efficiencies of the proposed method. In the first simulation, the controllers are applied on
the manipulator under the presence of the friction model and known control direction. In
the second simulation, besides the friction model, the unknown control direction is taken
into account in the manipulator dynamics. The simulation environment is MATLAB 2020a
with a sampling time of 10−3 s. The 3-DOF manipulator dynamics are mentioned in [5].
Their parameters are presented in Table 1.

Table 1. Parameters of the 3-DOF robotic manipulator.

Variable Description Variable Description

l1 = 0.46 m Length of the 1st link m1 = 0.7 kg Mass of the 1st link
l2 = 0.4 m Length of the 2nd link m2 = 0.6 kg Mass of the 2nd link
l3 = 0.2 m Length of the 3rd link m3 = 0.3 kg Mass of the 3rd link
g = 9.81 ms−2 Gravity constant

The friction model in the manipulator dynamics consists of the viscous and Coulomb
frictions, presented as follows:

d(t) = bx2 (t) + csign(x2 (t)) (27)

where b = 0.05I3×3 (Nms/rad) and c = 0.05I3×3 .

The trajectory reference in the Cartesian coordination is presented as xd = 0.4 +
0.15 cos(0.2π )(m), yd = 0(m), zd = −0.15 cos(0.2π + π/2)(m), and the rotary angle
around the z-axis is zero.

4.2. Simulation Comparisons

In order to demonstrate the advantages of the proposed control, the results of the
proposed control are compared with those of the sliding mode control (SMC) and adaptive
SMC (ASMC). The SMC is presented in (6), (9), (10) and (11). The ASMC is computed by
(6), (9), (10), (12) and (13).

Remark 1. The parameters of the three controllers are presented in Table 2. As the SMC is designed
based on the manipulator dynamics, the ASMC is developed from the SMC and the proposed control
inherits from the ASMC, so the parameters of the ASMC inherit from the SMC and the parameters
of the proposed control inherit from the ASMC to maintain the fairness in comparisons.

Table 2. Coefficients of the SMC, ASMC, and the proposed control.

Controllers Parameters
 T 
SMC λ = 10I3×3 , k = 3I3×3 , η = 15diag 2 1 1
λ = 10I3×3 , k = 3I3×3 , ε 1 = 0.001, κ1 = 10−2  2.5 2.5 1 , κ2 = 5·10−4 1
   
ASMC 1 1
λ = 10I3×3 , k = 3I3×3 , ε 1 = 0.001, κ1 = 10−2 2.5 2.5 1 , κ2 = 5·10−4 1
  
Proposed control 1 1
 before they are plotted in the figures of Section 4.3.

Table 2. Coefficients of the SMC, ASMC, and the proposed control.

Controllers Parameters
Appl. Sci. 2021, 11, 3919
SMC (
λ = 10I 3×3 , k = 3I 3×3 , η = 15diag [ 2 1 1]
T
) 8 of 14

ASMC λ = 10I 3×3 , k = 3I 3×3 , ε1 = 0.001 , κ1 = 10 −2 [ 2.5 2.5 1] , κ 2 = 5.10 −4 [1 1 1]

Proposed control λ = 10I 3×3 , k = 3I 3×3 , ε1 = 0.001 , κ1 = 10 −2 [ 2.5 2.5 1] , κ 2 = 5.10 −4 [1 1 1]
Remark 2. The data are multiplied with a constant to convert from radian units to degree units
before they are plotted in the figures of Section 4.3.
Remark 3. In order to improve the ability of the Nussbaum functions in control direction detection,
the
Remark 3. In orderNussbaum function
to improve in theofsimulation
the ability is modified
the Nussbaum as follows:
functions in control direction detection,
the Nussbaum function in the simulation is modified as follows:
 N2(ζ ) =sign  ζ  π  2

i π  e cos  ζ i  
i
(28)
N (ζ i ) = sign eζ i cos ζ i  2  (28)
2

4.3. Simulation4.3.Results
Simulation Results
In the first simulation
In the first case, the results
simulation in Figure
case, the results2inshow the2 output
Figure show theperformances at
output performances at
three joints of the joints
three SMC (blue
of the line),
SMC the ASMC
(blue (black
line), the ASMCline), and line),
(black the proposed control (red
and the proposed control (red
line), trackedline),
to the reference
tracked to thesignals.
referenceFigure 3 presents
signals. Figure 3the error responses
presents of the SMC
the error responses of the SMC
with a blue line,
with thea blueASMC
line, with a black
the ASMC line,
with and the
a black line,proposed control with
and the proposed a red
control line.
with a red line.
Figure 4 presents
Figure the control signal
4 presents performance
the control of SMC in of
signal performance Figure
SMC 4a, ASMC4a,
in Figure inASMC
Figurein4b,Figure 4b,
and proposed andcontrol
proposed in Figure
control4c.
in These
Figureresults show
4c. These the effectiveness
results of the adaptive
show the effectiveness of the adaptive
laws for robustlaws gain
for in chattering
robust gain insuppression in the ASMC
chattering suppression andASMC
in the the proposed controller.controller.
and the proposed

ppl. Sci. 2021, 11, x FOR PEER REVIEW 9 of 15

(a) (b)

(c)
Figure
Figure 2. Output responses of the SMC, 2. Output
ASMC, responses
and proposed of the
control in SMC,
(a) linkASMC, and2,proposed
1, (b) link (c) link 3.control in (a) link 1, (b) link 2, (c) link 3.

(a) (b)
 (c)
Appl. Sci. 2021, 11, 3919 9 of 14
Figure 2. Output responses of the SMC, ASMC, and proposed control in (a) link 1, (b) link 2, (c) link 3.

(a) (b)

(c)
Figure
Figure3.3.Error
Errorefforts
effortswith
withthe SMC,
the ASMC,
SMC, andand
ASMC, proposed control
proposed in (a)
control inlink 1, (b)1,link
(a) link (b) 2, (c)2,
link link
(c)3.link 3.

In the second case, three wrong connection situations are implemented to prove the
superiority of the proposed control. These situations are the
 wrong connection in an
 T  T 
actuator, B = diag −1 1 1 , two actuators B = diag −1 −1 1 , and all
 T 
three actuators B = diag −1 −1 −1 . Figure 5 shows the output responses of the
3-DOF manipulator with the proposed control with the wrong connection in actuator 1.
These results show good tracking performances. Figure 6 presents the control signals of
the proposed control at each joint of the manipulator. Comparing to Figure 4c, the control
signal of joint 1 in Figure 6 has a different direction, which exhibits the wrong connection
in actuator 1. Figure 7 plots the Nussbaum function at each joint to help the controlled
system overcome the unknown control direction. The response of the Nussbaum function
in Figure 7a changed from +1 to −1 to compensate for the connection issue in actuator 1.
 Appl. Sci. 2021, 11, x FOR PEER REVIEW 10 of 15
Appl. Sci. 2021, 11, 3919 10 of 14

(a) (b)

(c)
Appl. Sci. 2021, 11, x FOR PEER REVIEW
FigureFigure
4. Control performance
4. Control of (a) SMC,
performance of (b)
(a) ASMC, andASMC,
SMC, (b) (c) proposed controller
and (c) proposedin the 3-DOF manipulator.
controller in the 3-DOF
manipulator.
In the second case, three wrong connection situations are implemented to prove the
superiority of the proposed control. These situations are the wrong connection in an actu-
(
ator, B = diag [ −1 1 1]
T
) , two actuators (
B = diag [ −1 −1 1]
T
) , and all three actuators
(
B = diag [ −1 −1 −1]
T
) . Figure 5 shows the output responses of the 3-DOF manipulator
with the proposed control with the wrong connection in actuator 1. These results show
good tracking performances. Figure 6 presents the control signals of the proposed control
at each joint of the manipulator. Comparing to Figure 4c, the control signal of joint 1 in
Figure 6 has a different direction, which exhibits the wrong connection in actuator 1. Fig-
ure 7 plots the Nussbaum function at each joint to help the controlled system overcome
the unknown control direction. The response of the Nussbaum function in Figure 7a
changed from +1 to −1 to compensate for the connection issue in actuator 1.

Figure 5.5.The
Figure Theoutput responses
output of the proposed
hresponses of control under
the proposed the presence
control of the
under the wrong connection
presence of the wrong co
( )
i  T
in actuator 1, B =−diag
in actuator 1, B = diag
tion [ ]
1 1 −11 1 . 1 T .
Appl. Sci. 2021, 11, 3919 Figure 5. The output responses of the proposed control under the presence of the wrong connec-
11 of 14

(
tion in actuator 1, B = diag [ −1 1 1] ).T

Figure 5. The output responses of the proposed control under the presence of the wrong connec-
(
tion in actuator 1, B = diag [ −1 1 1]
T
).

Figure 6.
Figure Figure
Control
6.Control 6. Controlofresponses
responses
responses ofthe of the proposed
theproposed
proposed controllercontroller
with
controller with
the the
with wrongtheconnection
wrongwrong connection
ininactuator
actuator
in actuator
connection 1, 1,
1,
B=
B (
diag [ −1−11 11] 1 .
= diag
h
()
B = diagT [ −1 1 1] .
iT  T
. )

(a) (b)

(a) (b)

(c)
Figure 7. Nussbaum function
Figure 7. performance
Nussbaum functionofperformance
the proposedof
control in (a) joint
the proposed 1, (b) joint
control in (a)2,joint
and (c) joint
1, (b) 3. 2, and (c)
joint
joint 3.
Next, the advantages(c)of the proposed control are taken into account with the wrong
Figure 7. Nussbaum function
Next, the
connections in actuators 1 and 2,control
performance
advantages
of the
of
proposed
the proposed
control in (a) joint
are[ −taken
B = diag
1, (b)
1 −1 into
joint
T
(
12,]T  . account
and (c)
)
Figure 8with
is thethe
joint 3.
wrong
control responses
connections in actuators 1 and 2, B = diag −1 −1 1 . Figure 8 is the control
responses of the proposed controller. The results show that the control signals in actuator
Next, the advantages of the proposed control are taken into account with the wrong
1 and 2 are different directions when compared to the control signals in actuator 1 and 2
connections in actuators
with the normal 1 and
case in Figure TheBeffectiveness
4c. 2, (
= diag [ −1 −is1 shown
T
)
1] . Figure 8 is
in Figure the control
9, which responses
presents
Nussbaum function responses at each joint with the wrong connection in actuator 1 and 2.
The results in Figure 9a,b turn from +1 into −1 to overcome the wrong connections.
 In the final subcase,
 the wrong connection at all three actuators, B = diag
T
−1 −1 −1 , is considered. Figure 10 presents the control signals of the pro-
posed controllers. The sign of the control responses of the proposed control in three
actuators is different compared to the control responses in the normal case, as shown in
Figure 4c. Additionally, Figure 11 shows the Nussbaum function responses which turn into
−1 to help the proposed control to deal with the unknown control direction.
 of the proposed controller. The results show that the control signals in actuator 1 and 2
of the proposed controller. The results show that the control signals in actuator 1 and 2
are different directions when compared to the control signals in actuator 1 and 2 with the
are different directions when compared to the control signals in actuator 1 and 2 with the
normal case in Figure 4c. The effectiveness is shown in Figure 9, which presents Nuss-
normal case in Figure 4c. The effectiveness is shown in Figure 9, which presents Nuss-
baum function responses at each joint with the wrong connection in actuator 1 and 2. The
Appl. Sci. 2021, 11, 3919 baum function responses at each joint with the wrong connection in actuator 1 and
12 of 14 2. The
results in Figure 9Error! Reference source not found.a,b turn from +1 into −1 to overcome
results in Figure 9Error! Reference source not found.a,b turn from +1 into −1 to overcome
the wrong connections.
the wrong connections.

Figure 8. Control
of responses of the proposedwith
controller with connection
the wrong connection
Control
Figure 8. Control responses the proposed controller the wrong
wrong connection in actuatorin11actuator
in actuator and 1
and actuator 2, B = diagT [ −
( 1 −1 1] .
T
() )
h iT 
and actuator
actuator 2, B 2, B = diag−[1−1 −−11 11] . .
= diag

(a) (b)
(a) (b)

(c)
Figure 9. Nussbaum
Figure 9. Nussbaum function
h performance
(c) proposed
function performance
of the of the proposed
control incontrol in (a)
(a) joint 1, joint 1, (b)2,joint
(b) joint and2, (c)
andjoint
(c) 3,
( )
i  T
[ −1 joint ] B .= diag of−1the−1proposed
Appl. Sci. 2021, 11, x FOR T
Figure 9. Nussbaum B PEER
= diagREVIEW
function−1 3,
1performance 1
13 of 15
. control in (a) joint 1, (b) joint 2, and (c) joint 3,

(
B = diag [ −1 −1 1]
T
). In the final subcase, the wrong connection at all three actuators,

In ( T
)
the final [ subcase,] the wrong connection at all three actuators,
B = diag −1 −1 −1 , is considered. Figure 10 presents the control signals of the pro-

( T
)
posed controllers. The sign of the control responses of the proposed control in three actu-
B = diag [ −1 −1 −1] , is considered. Figure 10 presents the control signals of the pro-
ators is different compared to the control responses in the normal case, as shown in Figure
posed controllers. The sign ofFigure
4c. Additionally, the control responses
11 shows of the proposed
the Nussbaum control inwhich
function responses three turn
actu-into −1 to
ators is different
helpcompared to the
the proposed control
control responses
to deal inunknown
with the the normal case, direction.
control as shown in Figure
4c. Additionally, Figure 11 shows the Nussbaum function responses which turn into −1 to
help the proposed control to deal with the unknown control direction.

Figure
Figure 10. Controlresponses
10. Control responses
ofof
thethe proposed
proposed controller
controller with
with the the wrong
wrong connection
connection in allinthree
all three
( )
iT 
diag [ −1−1−1 −
−11] −
T
h
actuators, = diag
actuators, BB= .1 .
 Figure 10. Control responses of the proposed controller with the wrong connection in all three
Appl. Sci. 2021, 11, 3919 13 of 14
(
actuators, B = diag [ −1 −1 −1]
T
).

(a) (b)

(c)
Figure 11.
Figure 11. Nussbaum Nussbaum
function functionofperformance
performance of
hthe 3-DOF
the 3-DOF manipulator manipulator
with the proposedwith the proposed
control in (a) jointcontrol in 2,
1, (b) joint
( )
iT 
[ −1joint 1] .(c) joint 3, B = diag
T
= diag
and (c) joint 3,(a)Bjoint 1, (b) −1 2,−and −1 −1 −1 .

5. Conclusions5. Conclusions
This paper was
Thisaimed
paper at addressing
was the tracking
aimed at addressing thecontrol issue
tracking for an
control n-DOF
issue for anmanipu-
n-DOF manip-
lator with the existence
ulator with oftheunknown
existence friction and unknown
of unknown friction andcontrol direction.
unknown To direction.
control overcomeTo over-
these issues,come
an adaptive sliding
these issues, mode control
an adaptive sliding(ASMC) is developed
mode control (ASMC)with a Nussbaum
is developed with a Nuss-
function. Thebaum
SMCfunction. The SMC
in the proposed in theensures
control proposed thecontrol
trackingensures
problem theand
tracking problem and fast
fast responses
responses
for the manipulator. for the manipulator.
Additionally, there are Additionally,
adaptive lawsthere arerobust
for the adaptive
gainlaws forSMC
in the the robust
to gain
deal with theinunknown
the SMCfriction
to deal and
withsuppress
the unknown friction and
the chattering suppress
effect the chattering
in the system. effect in the
In practice,
the mistakessystem. In practice, the
in the connection mistakes
between in the connection
actuators and drivers,between
named actuators
unknown andcontrol
drivers, named
unknown control direction, cause serious damage to the manipulator.
direction, cause serious damage to the manipulator. To manage this issue, the Nussbaum To manage this is-
sue, the Nussbaum function is multiplied by the ASMC law.
function is multiplied by the ASMC law. A Lyapunov approach is investigated to analyze A Lyapunov approach is
the stability and robustness of the whole system in theory. Finally, several simulations are Finally,
investigated to analyze the stability and robustness of the whole system in theory.
implemented several simulations
on a 3-DOF are implemented
manipulator and theironresults
a 3-DOF aremanipulator
compared with and their
thoseresults
of theare com-
pared with those of the existing controllers
existing controllers to prove the advantages of the proposed method. to prove the advantages of the proposed
method.
For future works, some experiments will be carried out to demonstrate the advantages
of the proposed methods. The results in the paper can be used to conduct advanced
fault-tolerant controllers for robotic manipulators.

Author Contributions: Conceptualization, S.-H.H. and D.-T.T.; Data curation, M.S.T.; Formal analy-
sis, S.-H.H. and M.S.T.; Investigation, S.-H.H.; Methodology, S.-H.H. and D.-T.T.; Resources, S.-H.H.
and D.-T.T.; Software, S.-H.H. and M.S.T.; Supervision, D.-T.T.; Validation, S.-H.H., M.S.T. and D.-T.T.;
Visualization, S.-H.H. and M.S.T.; Writing—original draft, S.-H.H. and D.-T.T.; Writing—review &
editing, D.-T.T. All authors have read and agreed to the published version of the manuscript.
Funding: This research received no external funding.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Zanchettin, A.M.; Rocco, P.; Robertsson, A.; Johansson, R. Exploiting task redundancy in industrial manipulators during drilling
operations. Proc. IEEE Int. Conf. Robot. Autom. 2011, 128–133. [CrossRef]
2. Kazerooni, H.; Bausch, J.J.; Kramer, B.M. An Approach to Automated Deburring by Robot Manipulators. J. Dyn. Syst. Meas.
Control. 1986, 108, 354–359. [CrossRef]
3. Li, G.; Huang, H.; Guo, H.; Li, B. Dynamic Modeling and Control for a Deployable Grasping Manipulator. IEEE Access 2019, 7,
23000–23011. [CrossRef]
4. Li, Z.; Zhao, T.; Chen, F.; Hu, Y.; Su, C.Y.; Fukuda, T. Reinforcement Learning of Manipulation and Grasping Using Dynamical
Movement Primitives for a Humanoidlike Mobile Manipulator. IEEE/ASME Trans. Mechatron. 2018, 23, 121–131. [CrossRef]
Appl. Sci. 2021, 11, 3919 14 of 14

5. Tran, D.T.; Dao, H.V.; Dinh, T.Q.; Ahn, K.K. Output feedback control via linear extended state observer for an uncertain
manipulator with output constraints and input dead-zone. Electronics 2020, 9, 1355. [CrossRef]
6. Tran, D.-T.; Truong, H.-V.-A.; Ahn, K.K. Adaptive backstepping sliding mode control based RBFNN for a hydraulic manipulator
including actuator dynamics. Appl. Sci. 2019, 9, 1265. [CrossRef]
7. Yen, V.T.; Nan, W.Y.; van Cuong, P. Robust Adaptive Sliding Mode Neural Networks Control for Industrial Robot Manipulators.
Int. J. Control. Autom. Syst. 2019, 17, 783–792. [CrossRef]
8. Van, M.; Mavrovouniotis, M.; Ge, S.S. An adaptive backstepping nonsingular fast terminal sliding mode control for robust
fault-tolerant control of robot manipulators. IEEE Trans. Syst. Man, Cybern. Syst. 2019, 49, 1448–1458. [CrossRef]
9. Le, Q.D.; Kang, H.-J. Finite-Time Fault-Tolerant Control for a Robot Manipulator Based on Synchronous Terminal Sliding Mode
Control. Appl. Sci. 2020, 10, 2998. Available online: https://www.mdpi.com/2076-3417/10/9/2998 (accessed on 25 April 2020).
[CrossRef]
10. Lee, J.; Chang, P.H.; Jin, M. Adaptive Integral Sliding Mode Control With Time-Delay Estimation for Robot Manipulators. IEEE
Trans. Ind. Electron. 2017, 64, 6796–6804. [CrossRef]
11. Tran, D.T.; Ba, D.X.; Ahn, K.K. Adaptive Backstepping Sliding Mode Control for Equilibrium Position Tracking of an Electrohy-
draulic Elastic Manipulator. IEEE Trans. Ind. Electron. 2020, 67, 3860–3869. [CrossRef]
12. Utkin, V.; Guldner, J.; Shi, J. Sliding Mode Control in Electro-Mechanical Systems; CRC Press: Boca Ration, FL, USA, 2009.
13. Perruquetti, W.; Barbot, J.-P. Sliding Mode Control in Engineering; CRC Press: Boca Ration, FL, USA, 2002.
14. Roopaei, M.; Jahromi, M.Z. Chattering-free fuzzy sliding mode control in MIMO uncertain systems. Nonlinear Anal. Theory
Methods Appl. 2009, 71, 4430–4437. [CrossRef]
15. Amer, A.F.; Sallam, E.A.; Elawady, W.M. Adaptive fuzzy sliding mode control using supervisory fuzzy control for 3 DOF planar
robot manipulators. Appl. Soft Comput. J. 2011, 11, 4943–4953. [CrossRef]
16. Jung, S. Improvement of Tracking Control of a Sliding Mode Controller for Robot Manipulators by a Neural Network. Int. J.
Control. Autom. Syst. 2018, 16, 937–943. [CrossRef]
17. Zhao, L.; Li, Q.; Liu, B.; Cheng, H. Trajectory Tracking Control of a One Degree of Freedom Manipulator Based on a Switched
Sliding Mode Controller with a Novel Extended State Observer Framework. IEEE Trans. Syst. Man Cybern. Syst. 2019, 49,
1110–1118. [CrossRef]
18. Wang, Y.; Jiang, S.; Yan, F.; Chen, B. Sliding Mode Control of A Newly Designed Cable-Driven Manipulator with Time Delay
Estimation. In Proceedings of the 17th International Conference on Control, Automation and Systems (ICCAS), Jeju, Korea, 18–21
October 2017; No. 51705243. pp. 1605–1609.
19. Roy, S.; Kar, I.N. Adaptive sliding mode control of a class of nonlinear systems with artificial delay. J. Franklin Inst. 2017, 354,
8156–8179. [CrossRef]
20. Yu, J.; Shi, P.; Lin, C.; Yu, H. Adaptive Neural Command Filtering Control for Nonlinear MIMO Systems with Saturation Input
and Unknown Control Direction. IEEE Trans. Cybern. 2020, 50, 2536–2545. [CrossRef] [PubMed]
21. Zhai, D.; An, L.; Li, X.; Zhang, Q. Adaptive fault-tolerant control for nonlinear systems with multiple sensor faults and unknown
control directions. IEEE Trans. Neural Netw. Learn. Syst. 2018, 29, 4436–4446. [CrossRef]
22. Wang, H.; Karimi, H.R.; Liu, P.X.; Yang, H. Adaptive neural control of nonlinear systems with unknown control directions and
input dead-zone. IEEE Trans. Syst. Man Cybern. Syst. 2018, 48, 1897–1907. [CrossRef]
23. Xu, B.; Qi, R.; Jiang, B. Adaptive Fault-Tolerant Control for HSV with Unknown Control Direction. IEEE Trans. Aerosp. Electron.
Syst. 2019. [CrossRef]
24. Habibi, H.; Nohooji, H.R.; Howard, I. Adaptive PID control of wind turbines for power regulation with unknown control
direction and actuator faults. IEEE Access 2018, 6, 37464–37479. [CrossRef]
25. Xia, J.; Zhang, J.; Feng, J.; Wang, Z.; Zhuang, G. Command Filter-Based Adaptive Fuzzy Control for Nonlinear Systems With
Unknown Control Directions. IEEE Trans. Syst. Man Cybern. Syst. 2019, 1–9. [CrossRef]
26. Wang, C.; Wen, C.; Guo, L. Adaptive Consensus Control for Nonlinear Multi-Agent Systems With Unknown Control Directions
and Time-Varying Actuator Faults. IEEE Trans. Autom. Control. 2020, 1. [CrossRef]
27. Liu, Y.; Ma, H. Adaptive Fuzzy Tracking Control of Nonlinear Switched Stochastic Systems With Prescribed Performance and
Unknown Control Directions. IEEE Trans. Syst. Man Cybern. Syst. 2020, 50, 590–599. [CrossRef]
28. Ge, S.S.; Wang, J. Robust adaptive neural control for a class of perturbed strict feedback nonlinear systems. IEEE Trans. Neural
Netw. 2002, 13, 1409–1419. [CrossRef]
29. Boulkroune, A.; Tadjine, M.; M’Saad, M.; Farza, M. Fuzzy adaptive controller for MIMO nonlinear systems with known and
unknown control direction. Fuzzy Sets Syst. 2010, 161, 797–820. [CrossRef]
30. Ramezani, Z.; Arefi, M.M.; Zargarzadeh, H.; Jahed-Motlagh, M.R. Neuro observer-based control of pure feedback MIMO systems
with unknown control direction. IET Control Theory Appl. 2017, 11, 213–224. [CrossRef]