Proceedings of the 37th Chinese Control Conference

July 25-27, 2018, Wuhan, China

An Adaptive Sliding Mode Controller with The Exponential and

Power Reaching Law for Discrete Systems
Xiaofei Zhang1 , Hongbin Ma∗ 1,2 Zhuang Li3 , Wenchao Zao4
1. School of Automation, Beijing Institute of Technology, Beijing 100081, P. R. China
E-mail: [email protected]
2. State Key Laboratory of Intelligent Control and Decision of Complex Systems, Beijing Institute of Technology, Beijing 100081,
P. R. China
E-mail: [email protected]
3. School of Automation, Beijing Institute of Technology, Beijing 100081, P. R. China
E-mail: [email protected]
4. School of Automation, Beijing Institute of Technology, Beijing 100081, P. R. China
E-mail: [email protected]

Abstract: With the development of aeronautics and astronautics, the response speed of servo system need be faster. First, In
order to improve the dynamic quality of servo system, the exponential and power reaching law, which combines the advantages
of the exponential reaching law and the power reaching law, is introduced. Second, the chattering of the sliding mode controller
(SMC) with the exponential and power reaching law for discrete systems is investigated. Finally, the adaptive sliding mode
controller (ASMC) with the exponential and power reaching law is introduced. The stability of the ASMC with the exponential
and power reaching law for discrete systems is analyzed, and the simulation of this approach on one joint of a six degrees of
freedom robot is carried out. The experimental results indicate that the ASMC with the exponential and power reaching law is
effective in reducing the time of arriving the sliding mode surface. The experimental results also indicate that the ASMC with
the exponential and power reaching law may make output error reach zero in a shorter time.

Key Words: adaptive, sliding mode controller, reaching law

1 Introduction switching functions, Y. Eun[7] designed an SMC for dis-

crete system. G. Bartolini[8–11] reduced chattering of an
Sliding mode control algorithm has strong robust-
SMC for mechanical systems, which contain dynamics dis-
ness to the changes of system parameters and external
turbances and uncertain disturbances. M. Ertugrul[12] re-
disturbances[1]. It has been successfully applied to complex
alized the equivalent control based on neural network and
systems, e.g., motor and power control system, robot control
reduced effectively the chattering of sliding mode control.
system, flight control system. A discrete version of sliding
M. Ertugrul[13] designed a new SMC with neural network.
mode control is important when the implementation of the
It is worthwhile to note that the aforementioned litera-
control is realized by computers with a relatively slow sam-
tures mainly focus on the chattering problem. Reaching time
pling time. Because only the ideal continuous sliding mode
decreasing has been ignored by most literatures. However,
control has the equivalent control ueq and may produce vari-
with the development of aeronautics and astronautics, the re-
able structure control with switched logic, the discrete slid-
sponse speed of servo system need be faster. Reaching time
ing mode control may not be obtained from their continuous
has a close relationship with the dynamic quality of system.
counterpart by means of simple equivalence[2]. In practi-
In recent years, some scholars pay attention to terminal slid-
cal engineerings, the computer real-time control system is a
ing mode[14–17]. Besides, W. B. Gao[18] put forward an
discrete system, and the research and design of sliding mode
approach of sliding mode control, which may reduce buf-
control for discrete systems become an important component
feting using reaching law. Using the exponential reaching
of sliding mode control applications. In 1980s, the discrete
law s(k + 1) − s(k) = −εT sgn(s(k)) − qT s(k), the fre-
sliding mode control developed rapidly and was widely used
quent chattering of the control signal may be reduced, and
in engineering field[3]. In 1983, an SMC for a rigid manipu-
the dynamic quality of sliding mode control may be guar-
lator with two degrees of freedom was designed for the first
anteed by adjusting the parameters of reaching law. C. L.
time by J. J. In recent years, the robot control system has be-
Di[19] introduced an SMC with the exponential reaching
come one of the main applications for sliding mode control
law for discrete systems, analyzed the relationship between
theory[4]. However, some limitations are worth paying at-
reaching coefficient and buffeting, analyzed the chattering
tention. Chattering and reaching time are two main problems
reason and put forward the adaptive algorithm for adjusting
of sliding mode control. In order to reduce chattering, many
coefficient. More specifically, chattering reduction of slid-
scholars have done much research about switching functions
ing mode control may be assured by choosing a small con-
and boundary layers. W. C. Su[5] designed two filters (front
trol gain, as in [20–24]. While, the small control gain will
filter and back filter). H. Liu [6] introduced an SMC based
eventually make reaching time become longer and make sys-
on error prediction. By designing two order derivative of
tem have poor robustness [25]. The exponential term−qs(k)
This work is supported by the Foundation of Key Laboratory of Science of the exponential reaching law overcomes the disadvantage
and Technology for National Defense (No.9140C770205150C77319), the of the constant speed reaching law (The speed of constant
National Natural Science Foundation of China (61473038, 91648117) and
Beijing Natural Science Foundation (4172055).
speed reaching law is unchangeable), the system reaches the

2711
Authorized licensed use limited to: Walchand College of Eng. Downloaded on January 09,2023 at 12:16:31 UTC from IEEE Xplore. Restrictions apply.
sliding mode surface at a changeable speed when the system buffeting and put forward the adaptive algorithm for adjust-
is not on the sliding mode surface. The power reaching law ing coefficient ε. When the sampling time is very small, the
may assure that control gain is small when the system is near existence of discrete sliding mode control and the reaching
to the sliding mode surface. In this paper, First, three funda- conditions are [26]
mental reaching laws (the constant speed reaching law, the
exponential reaching law and the power reaching law) are [s(k + 1) − s(k)]sgn(s(k)) < 0, (8)
introduced. In order to shorten the time of arriving the slid-
ing mode surface, the exponential and power reaching law, [s(k + 1) + s(k)]sgn(s(k)) > 0. (9)
which combines the advantages of the exponential reaching The discrete reaching law of the ASMC with the exponen-
law and the power reaching law, is introduced. Second, the tial reaching law is
chattering of SMC with the exponential and power reaching
law for discrete systems is investigated. Third, the ASMC |s(k)|
s(k + 1) − s(k) = −qT s(k) − T sgn(s(k)). (10)
with the exponential and power reaching law is introduced, 2
and this approach is applied to a joint simulation experiment
where |s(k)|/2 is control gain and, T = 0.001, T is sam-
of a six degrees of freedom robot. Finally, we compare the
pling time.
results of this approach with the ASMC with the exponential
The switching function is
reaching law.
2 Mathematical Model of One Joint of a Six De- s(k) = Ce E = Ce (R(k) − x(k)), (11)
grees of Freedom Robot
where q > 0, Ce = [c 1].
One joint of a six degrees of freedom robot is studied. The The control law of the ASMC with the exponential reach-
dynamic equation of one joint may be expressed as ing law is
J θ̈ + B θ̇ = u, (1) u(k) = (Ce B)−1 (Ce R(k + 1) − Ce Ax(k) − s(k) − ds(k)),
where J is the moment of inertia, B is the viscous damping (12)
coefficient, θ is the angular position, and u is the control law. where
The motor parameters are J = 1/133, B = 25/133. |s(k)|
The transfer function of one joint of a six degrees of free- ds(k) = −qT s(k) − T sgn(s(k)). (13)
2
dom robot may be expressed as
4 Three Fundamental Reaching Laws
133
G(S) = . (2) W. B. Gao put forward an SMC for weaking chattering
(S + 25)S
using reaching laws.
The sampling time is 0.001 s. The discrete model of this There are there fundamental reaching laws.
system is given by
4.1 The Constant Speed Reaching Law
x(k + 1) = Ax(k) + Bu(k), (3)
s(k + 1) − s(k) = −T εsgn(s(k)), ε > 0, (14)
where x(k) = [x1 (k)x2 (k)]T , x1 (k) = θ is the current an-
gular position of the motor. u(k) is the current control law. where ε is the speed of approching sliding mode surface.
r(k) is the position command. dr(k) is the change law of The smaller the speed of approching sliding mode surface
position command. R(k) = [r(k) dr(k)]T , R( k + 1) = is, the longer the time of arriving the sliding mode surface
[r(k +1) dr(k +1)]T . Obviously, from the above-mentioned becomes; If we make the speed become larger, the time will
model we know that the A matrix and B matrix are become short. When the speed of approching sliding surface
  is large, the chattering will become frequent.
1 0.001
A= , (4) 4.2 The Exponential Reaching Law
0 0.9753
 
0.0001
B= . (5) s(k + 1) − s(k) = −T εsgn(s(k)) − qT s(k), ε > 0, q > 0,
0.1314
(15)
Linear extrapolation is used to predict r(k+1) and dr(k+
where−qT s(k) is the exponential term. Its solution is
1)
s(k) = s(0)e−qkT . The exponential term −qs(k) over-
r(k + 1) = 2r(k) − r(k − 1), (6)
comes the disadvantage of the constant speed reaching law
dr(k + 1) = 2dr(k) − dr(k − 1). (7) (The speed of constant speed reaching law is unchangeable)
and provides the system with access to the sliding mode sur-
3 Design of the Adaptive Sliding Mode Controller
face at a changeable speed when the system is not on the slid-
with the Exponential Reaching Law
ing mode surface. The system reaches sliding mode surface
L. C. Di[19] introduced the ASMC with the exponential in a relatively short time. Meanwhile, when system is near
reaching law for discrete systems and analyzed the buffet- to the sliding mode surface, the system approaches the slid-
ing reason caused by reaching coefficient. Besides, he also ing mode surface at a constant speed, and the system reaches
analyzed the relationship between reaching coefficient and sliding mode mode surface in a finite time.

2712
Authorized licensed use limited to: Walchand College of Eng. Downloaded on January 09,2023 at 12:16:31 UTC from IEEE Xplore. Restrictions apply.
4.3 The Power Reaching Law The sampling time T is very small, T << 1.0, α = 2. By
Eq 18, we know
|s(k + 1)|
s(k + 1) − s(k) = −KT |s(k)|α sgn(s(k)), K > 0, α > 0. |ρ| = , (20)
|s(k)|
(16)
where KT |s(k)|α is control gain. When the system is near to ρ = 1 − qT − KT |s(k)|. (21)
sliding mode surface, the smaller control gain is guaranteed,
and the chattering is reduced. Obviously, ρ < 1. For Eq 21, the following three kinds of
4.4 The Exponential and Power Reaching law conditions are analyzed.
(1)When |s(k)| < (2 − qT )/(KT ),
The exponential reaching law provides the system with
access to a sliding mode surface at a changeable speed when 2 − qT
the system is not on the sliding mode surface. The power ρ > 1 − qT − KT = −1. (22)
KT
reaching law assuee that control gain is small when system
is near to the sliding mode surface. This paper introduces the so, |ρ| < 1, |s(k + 1)| < |s(k)|, |s(k)| diminishes.
exponential and power reaching law combining the advan- (2)When |s(k)| > (2 − qT )/(KT ),
tages of the power reaching law and the exponential reaching
law. 2 − qT
ρ < 1 − qT − KT = −1. (23)
α KT
s(k + 1) − s(k) = − KT |s(k)| sgn(s(k)) − qT s(k),
K > 0, α > 0, q > 0. so, |ρ| > 1, |s(k + 1)| > |s(k)|, |s(k)| increasees.
(17) (3)When |s(k)| = (2 − qT )/(KT ),
The curves of these four reaching laws are compared and 2 − qT
analyzed. The experimental results are shown in Fig. 1. It ρ = 1 − qT − KT = −1. (24)
KT
may be seen from Fig. 1 that the exponential and power
reaching law combines the advantages of the exponential so, |ρ| = 1, |s(k + 1)| = |s(k)|, |s(k)| goes into instable
reaching law and the power reaching law. When s(k) > 0.7, state.
the ds(k) of the exponential and power reaching law is
greater than the ds(k) of the exponential reaching; When 6 Design of Adaptive Sliding Mode Controller
s(k) < 0.7, the ds(k) of the exponential and power reaching with The Exponential and Power Reaching Law
law smoothly approaches zero. for Discrete Systems
From the Eq22−Eq24 we may obtain that only when
|S(k)| < (2 − qT )/KT , the state s(k) will diminish, which
means the q should satisfy

2 − KT |s(k)|
q< . (25)
T

In this paper, we choose q = 1500 − K|s(k)|, the inequality

is established because the sampling time T = 0.001.
The discrete reaching law is

s(k + 1) − s(k) = − (1500 − K|s(k)|)T s(k)

(26)
− KT |s(k)|α sgn(s(k)).

The switching function is

s(k) = Ce E = Ce (R(k) − x(k)), (27)

Fig. 1: The curves of these four reaching laws
where q > 0, T = 0.001, Ce = [c 1].
The control law of ASMC with the exponential and power
5 Chattering Analysis of Sliding Mode Controller
reaching law for the discrete system is
with The Exponential and Power Reaching Law
for Discrete Systems
u(k) = (Ce B)−1 (Ce R(k + 1) − Ce Ax(k) − s(k) − ds(k)),
Eq 17 can be rewritten as follows: (28)
where
s(k + 1) = ρs(k), (18)
ds(k) = − KT |s(k)|2 sgn(s(k))
where (29)
ρ = 1 − qT − KT |s(k)|α−1 . (19) − (1500 − K|s(k)|)T s(k).

2713
Authorized licensed use limited to: Walchand College of Eng. Downloaded on January 09,2023 at 12:16:31 UTC from IEEE Xplore. Restrictions apply.
6.1 Stability Analysis
Aimed at one joint of a six degrees of freedom robot Eq3
and combined the discrete reaching law Eq26 and the control
law Eq28, we get

(s(k + 1) − s(k))sgn(s(k)) < 0, (30)

and
(s(k + 1) + s(k))sgn(s(k)) > 0. (31)
Thus, the discrete reaching law Eq 28 satisfies the existence
and reaching conditions of discrete sliding mode controller,
so the designed control system is stable.
7 Analysis of Experimental Result
Fig. 3: The control signal change curve of ASMC with the
C. L. Di proposed an ASMC with the exponential reaching
exponential reaching law
law for discrete systems. This paper introduces an ASMC
with the exponential and power reaching law for discrete
systems. These two approaches are applied to a joint sim-
ulation experiment of a six degrees of freedom robot. The
parameters of ASMC with the exponential and power reach-
ing law for discrete systems are K = 30, α = 2, c = 10.
The parameters of AMSC with the exponential reaching law
for discrete systems are c = 10, q = 30. Initial states of
controlled object is [-0.5 -0.5].

Fig. 4: The sliding surface change curve of ASMC with the

exponential reaching law

Fig. 2: The motor angle change curve of ASMC with the

exponential reaching law

The simulation results of ASMC with the exponential

reaching law are shown in Fig. 2, Fig. 3 and Fig. 4.
The simulation results of ASMC with the exponential and
power reaching law are shown in Fig. 5, Fig. 6 and Fig.
7. Fig. 2 plots the motor angle change of ASMC with the
exponential reaching law; Fig. 3 shows that the control sig-
nal change of ASMC with the exponential reaching law; In
Fig. 4. the curve denotes the sliding mode surface function Fig. 5: The motor angle change curve of ASMC with the
s(k) of ASMC with the exponential reaching law; Fig. 5 exponential and power reaching law
plots the motor angle change of ASMC with the exponen-
tial and power reaching law; Fig. 6 shows that the con-
trol signal change of ASMC with the exponential and power reach the target position in a limited time. The motor an-
reaching law; In Fig. 7 the curve denotes that the sliding gle error of ASMC with the exponential and power reaching
mode surface function s(k) of ASMC with the exponential law reaches 0.0057191, at time 0.566 s. The motor angle
and power reaching law; The motor angle of ASMC with error of ASMC with the exponential reaching law reaches
the exponential and power reaching law and the motor an- 0.0057387, at time 0.6 s. Because the sampling time is
gle of the ASMC with the exponential reaching law may 0.001s, ASMC with the exponential and power reaching law

2714
Authorized licensed use limited to: Walchand College of Eng. Downloaded on January 09,2023 at 12:16:31 UTC from IEEE Xplore. Restrictions apply.
 stant speed reaching law, the exponential reaching law, the
power reaching law and the exponential and power reaching
law. The results show that when s(k) > 0.7, the ds(k) of
the exponential and power reaching law is bigger than the
ds(k) of the exponential reaching law; When s(k) < 0.7,
the exponential and power reaching law smoothly approches
zero. Second, we analyze the chattering of SMC with ex-
ponential and power reaching law for discrete systems and
find the stability condition of SMC with the exponential and
power reaching law for discrete systems. Finally, we intro-
duce the ASMC with exponential and power reaching law.
The experimental results show that the sliding surface func-
tion of ASMC with exponential and power reaching law
reaches zero earlier than ASMC with exponential reaching
Fig. 6: The control signal change curve of ASMC with the law. The experimental results also show that ASMC with
exponential and power reaching law exponential and power reaching law has 34 sampling times
less than ASMC with exponential reaching law, when output
error reaches 0.0057. The motor position of ASMC with ex-
ponential and power reaching law reach target position ear-
lier than ASMC with exponential reaching law. A robot is

Table 1: The first 20 sampling points

k Time(k) s1 (k) s2 (k)
1 0.001 1015.5 1015.5
2 0.002 14.172 14.172
3 0.003 15.505 15.505
4 0.004 15.032 14.177
5 0.005 14.573 12.869
6 0.006 14.129 11.58
7 0.007 13.698 10.31
8 0.008 13.28 9.0583
9 0.009 12.875 7.8242
Fig. 7: The sliding surface change curve of ASMC with the 10 0.010 12.483 6.6074
exponential and power reaching law 11 0.011 12.102 5.4075
12 0.012 11.733 4.224
13 0.013 11.375 3.0567
has 34 sampling times less than ASMC with the exponential 14 0.014 11.028 1.9051
reaching law, when the error reaches 0.0057. Besides, there 15 0.015 10.692 0.76874
is an obvious difference on the sliding mode surface func- 16 0.016 10.365 -0.35268
tion s(k). At time 0.020 s, the sliding mode surface function 17 0.017 10.049 0.17634
s(k) of ASMC with the exponential and power reaching law 18 0.018 9.7428 -0.08817
is -0.022043, and the sliding mode surface function s(k) of 19 0.019 9.4457 0.044085
ASMC with the exponential reaching law is 9.1576. Mean- 20 0.020 9.1576 -0.022043
while, the experimental results of ASMC with the exponen-
tial reaching law also show that the sliding mode surface a coupled and nonlinear dynamic system. The parameters
function s(k) of ASMC with the exponential reaching law of robot often change following the change of time. For ex-
is 0.022493, at time 0.214 s. It implies that the ASMC with ample, the dynamic characteristics of motor vary with the
the exponential and power reaching law reaches zero earlier inertia torque and the load friction. Model errors and dis-
than ASMC with the exponential reaching law. turbances are also important factors[27]. The time-varying
In order to further analyze the change trend of the sliding uncertainty of parameters or structure, which often makes
mode surface functions s(k) of those two approaches, the the system swing and makes dynamic and static performance
first 20 sampling points are listed in table 1, where s1 (k) is of the system have a change, brings difficulty to the high-
the sliding mode surface function of ASMC with the expo- precision trajectory control of a robot. The motion of robot
nential reaching law, and s2 (k) is the sliding mode surface may be divided into point-to-point motion and continue mo-
function of ASMC with the exponential and power reaching tion. The point-to-point motion refers to the motion of point
law. From this table, we learn that when k is 20, time(k) is to point in space. In this paper, we introduce ASMC with ex-
0.020, s2 (k) is -0.022043, which is far smaller than s1 (k) ponential and power reaching law and prove than one joint
(s1 (k) is 9.1576). In this paper, first, we introduce the system of six degrees of freedom robot is stable using this
exponential and power reaching law combining the advan- approach. However, the one joint mathematical model of a
tages of the exponential reaching law and the power reach- six degrees of freedom robot does not contain disturbances.
ing law. Meanwhile, we compare the advantages of the con-

2715
Authorized licensed use limited to: Walchand College of Eng. Downloaded on January 09,2023 at 12:16:31 UTC from IEEE Xplore. Restrictions apply.
So, in the furture we will try to prove that this approach is [13] S. Purwar, I.N. Kar, and A.N. Jha. Neuro sliding mode
stable when it is applied to one joint system of a six degrees control of robotic manipulators. volume vol.1, pages 595
of freedom robot, which contains disturbances. we will ap- – 600, Piscataway, NJ, USA, 2005//. neuro sliding mode
ply this approach to a six degrees of freedom robot in prac- control;robotic manipulators;neuro-adaptive control;nonlinear
tical engineering. systems;neural network based controllers;Chebyshev neural
network;unknown dynamics;adaptive switching gain;neural
References approximation error compensation;disturbance compensa-
[1] J. K. Liu. Sliding Mode Variable Structure Control MATLAB tion;single layer CNN;Lyapunov stability theory;ultimate uni-
Simulation. PTsinghua University Press, 2010. form boundedness;tracking error;.
[2] A. Homaifa W. B. Gao, Y. F. Wang. Discrete-time variable [14] Z. Wang, Y. X Su and L. Y. Zhang. A New Nonsingular
structure control systems. IEEE Transaction on Industrial Terminal Sliding Mode Control for Rigid Spacecraft Attitude
Electronics, 42(2):117–122, 1995. Tracking. Journal of Dynamic Systems Measurement and c
[3] J. K. Liu. Sliding Mode Variable Structure Control MATLAB Control, 140(5):051006.1-051006.9, 2017.
Simulation. PTsinghua University Press, 2012. [15] S. Mobayen Fast terminal sliding mode controller design for
[4] X. H. Yu and J. X. Xu. Advances in Variable Structure Systems. nonlinear second-order systems with time-varying uncertain-
World Scientific Publishing,, 2000. ties. COMPLEXITY, 21(2):239-244, 2015
[5] W. C. Su, S. V.Drakunov, and K. D. Young. Sliding mode with [16] Q. S. Xu. Piezoelectric Nanopositioning Control Using
chattering reduction in sampled data systems. In Proceedings Second-Order Discrete-Time Terminal Sliding-Mode Strat-
of the 32nd IEEE Conference On Decision and Control, pages egy IEEE TRANSACTIONS ON INDUSTRIAL ELECTRON-
2452–2457, 1993. ICS ,62(12):7738-7748, 2015
[6] H. Liu. Smooth sliding mode control of uncertain system based [17] S. Mobayen, D. Baleanu and F. Tchier Second-order fast
on a prediction error. International Journal of Robust and Non- terminal sliding mode control design based on LMI for a
linear Control, 7(4):414–422, 1997. class of non-linear uncertain systems and its application to
[7] H. Liu. Discrete-time variable structure controller with a de- chaotic systems. JOURNAL OF VIBRATION AND CONTROL,
coupled disturbance compensator and its application to a cnc 23(18):2912-2925, 2017
servomechanism. IEEE TRANSACTIONS ON CONTROL SYS- [18] W. B. Gao. Theoretical Research and Design Method of Vari-
TEMS TECHNOLOGY, 11(1):156–157, 1999. able Structure Control. Science Press, 2012.
[8] G. Bartolini, A. Ferrara, and E. Usani. Chattering avoidance [19] C. L. Di and Z. M. Wu. Variable structure control method for
by second-order sliding mode control. IEEE Transactions on discrete time systems. Journal of Shanghai Jiao Tong Univer-
Automatic Control, 43(2):241–246, 2002. sity, 34(5):719–722, 2000.
[9] G. Bartolini, A. Ferrara, E. Usai, and V.I. Utkin. On [20] O. Barambones and P.Alkorta. Postion control of the induc-
multi-input chattering-free second-order sliding mode con- tion motor using an adaptive sliding-mode controller and ob-
trol. IEEE Transactions on Automatic Control, 45(9):1711 – servers. IEEE Trans. Ind. Electron, 61(12):6556–6565, 2014.
17, Sept. 2000. sliding mode control;variable structure sys- [21] O. Drazenovic. The invariance conditions in variable structure
tems;uncertain systems;nonlinear systems;state space;MIMO systems. Automatica, 5(3):287–295, 1969.
systems;. [22] J. Zhang, G. Feng, and Y. Xia. Design of estimator-
[10] G. Bartolini and E. Punta. Chattering elmination with second- based sliding-mode output-feedback controllers for discrete-
order sliding modesrobust to coulomb friction. journal of Dy- time systems. IEEETrans. Ind. Electron, 61(5):2432–2440,
namic SystemMeasurement and cotrol, 122(4):679–686, 2000. 2014.
[11] G. Bartolini, A. Pisano, E. Punta, and E. Usai. A survey of ap- [23] K. Furuta. Sliding mode control of a discrete system. Syst.
plications of second-order sliding mode control to mechanical ControlLett, 14:144–152, 1990.
systems. International Journal of Control, 76(9-10):875 – 92, [24] B. Veselic, B. Perunicic-Drazenovic, and C. Milosavlje-
0010/06/15. second-order sliding mode control;mechanical vic. Improved discrete-time sliding-mode position control
systems application;switching control;time derivatives;sliding using euler velocity estimation. IEEE Trans. Ind. Electron,
variable;finite-time convergence;unmodelled actuator dy- 57(11):3840–3847, 2010.
namics;uncertain system;second-order differential inequal- [25] W. Gao and J. C.Hung. Variable structure control of nonlinear
ities;differentiation algorithms;phase trajectory;robotic ma- systems a new approach. IEEE Trans. Ind. Electron, 40(1):45–
nipulators;overhead cranes;Lyapunov-based control;position 55, 1993.
control;active pantographs control;Jacobian matrix;. [26] S. Z. Sarpturk, Y. Istefanopulos, and O. Kaynak. On the
[12] Hiroshi Moroika, Kenzo Wada, Asif Sabanovic, and Karel stability of discrete-time sliding mode control system. IEEE
Jezernik. Neural network based chattering free sliding mode Transactions on Automatic Control, 32(10):930–932, 1987.
control. pages 1303 – 1308, Hokkaido, Jpn, 1995. Chat- [27] M. Zhuang and X. P. Liu. Robot trajectory planning method
tering;Continuous function;Equivalent control;On-line estima- based on combined sine curve function. Mechanical Engineer-
tion;Seesaw system;Sliding mode control;Systems dynamics;. ing and Automation, 1:64–68, 2008.

2716
Authorized licensed use limited to: Walchand College of Eng. Downloaded on January 09,2023 at 12:16:31 UTC from IEEE Xplore. Restrictions apply.