2013 2nd International Symposium on Instrumentation and Measurement, Sensor Network and Automation (IMSNA)

Research of the discrete sliding mode variable structure algorithm based on the
exponential trending rule

LI Hong-ke YANG Hai-liang

Electronic Engineering College Electronic Engineering College
Naval University of Engineering Naval University of Engineering
Wuhan 430033,China Wuhan 430033,China
e-mail: [email protected]

HUANG Linshu XU Kai

Electronic Engineering College Electronic Engineering College
Naval University of Engineering Naval University of Engineering
Wuhan 430033,China Wuhan 430033,China
e-mail: [email protected]

Abstract—Instability characteristics of a shipborne satellite Satellite communications antenna servo systems are set
communication station tend to reduce the tracking accuracy in the host computer and the servo controller. If sliding mode
and communication quality. To improve the system control algorithm is used in them to some extent , it is
performance, an antenna servo controller based on DSP is helpful to overcome the deficiencies of the servo system and
developed in this paper. For the servo system, a new algorithm has important engineering significance. This algorithm is
based on the exponential trending rule discrete sliding mode mainly used in simulation against the interference generated
variable structure algorithm is proposed. The theoretical by friction torques. It achieves simulation experiment using a
analysis is derived and experimental simulating are developed. sliding mode control algorithm based on the rate of
Experiment results show that a discrete controller and its
exponential. Experiments results show that compared to the
control algorithm satisfies the stability. Compared with the
traditional PID law, the novel algorithm has much better
traditional the PID control algorithm, the sliding mode
performance on real-time adjustment. The sensitivity of the control algorithm have more good adaptability for
servo system to the continuous or unexpected interferences are interference.
greatly reduced and robustness are obviously strengthened.
II. DISCRETE VARIABLE STRUCTURE CONTROL
Keywords- servo control, satellite antenna, variable structure ALGORITHM DESIGNING
control,sliding mode The experimental platform of antenna servo controller is
based on the DSP , named the antenna servo controller
I. INTRODUCTION velocity servo system (speed control system), which consist
of the controller, power driver board, DSP board, USB bus,
When vessels are sailing, the platforms baseline of PC host computer and other devices.
satellite antenna tracking system will offset in three The main idea of the sliding mode variable structure
directions respectively of the azimuth , elevation and roll, in control is: to make the state of the system to run in
particularly, people should consider the disruption, accordance with the sliding surface preset, before the system
stabilization system to control the antenna pointing and state trajectory move to the sliding surface or the switch
automatically aiming the satellite. The antenna tracking of function, forcing the system state variables move to the
satellite communication earth station, initially uses single- sliding plane with the switch function, once encountered the
pulse mode with high precision, the beam is wide. With the sliding surface, the system status will slide along the sliding
satellite signal bandwidth increases, higher the operating surface has to end [1]. Sliding state of the system is
frequency becomes and the beam also is more narrow, and determined by the parameters of the trajectory of the sliding
some even reaches to several to near zero degree. Such a surface.
narrow beam width , when it is disturbed, especially on
satellite tracking accuracy and communication quality. Also consider the PWM generation circuit gain K A , you
Requires to satellite antenna servo system is improving can get the motor speed-PWM duty cycle instantaneous
continuously, which requires the servo system running with transfer function [2], as follows:
high accuracy , good interference resistance, automatic
aiming the satellite . Therefore, a set of advanced automatic
tracking system must be designed in the satellite earth station,
so that the receiver antenna can aiming the satellite
accurately in real time.

978-1-4799-2716-6/13/$31.00 ©2013 IEEE 253

Authorized licensed use limited to: Walchand College of Eng. Downloaded on January 09,2023 at 12:13:59 UTC from IEEE Xplore. Restrictions apply.
 2013 2nd International Symposium on Instrumentation and Measurement, Sensor Network and Automation (IMSNA)

From formula(4) and (5),we can get:

ω (s) K AKt
= Ce R(k + 1) − Ce Ax(k ) − Ce Bu (k )
D( s) ( Ls + R )( Js + B ) + K t K e
− Ce R(k ) + Ce x(k ) ˄˅
K AKt 
= −qTs e (k ) − εT sgn( se (k ))
= JL
BL + JR K K + RB Assumed C e B reversible, so, discrete variable structure
s2 + s+ t e control law based on exponential reaching law can be
JL JL obtained by the formula (6), are expressed as formula (7)
Considering the system as following:
shown :
ω ( s) 6788489813
= 2
   u (k ) = (C e B ) −1 (C e R (k + 1) − C e Ax(k )
D( s) s + 7917.67 s + 718448.5
Real-time control systems for the computer, so the need − C e R (k ) + C e x(k ) 
converting the continuous system represented by formula (1) + qTs e (k ) + εT sgn( s e (k )))
to a discrete system. In this experimental system, the control
system set the sampling time, i.e. the sampling frequency of ˄˅
the system. Dispersing the continuous system of formula(1) Or it is presented as formula(8):
using a Z-transform method, specific operation is to calculate u (k ) =
use MATLAB function c2dm () and obtain the second-order
discrete system state equation of controlled object, as shown (Ce B ) −1 (Ce R (k + 1) ˄˅
in formula (2) below:
− Ce Ax(k ) − se (k ) − dse (k ))
ª x1 º ªx º
« x » = A« 1 » + Bu ˄˅ Where : dse ( k ) = −qTse ( k ) − εT sgn( se ( k ))) ˗
¬ 2¼ ¬ x2 ¼ ε > 0, q > 0, 1 − qT > 0 DŽ
In this experimental system, PWM signal generated by This is the form of discrete reaching law sliding mode
the DSP chip, the duty cycle is adjustable from 0 to 1, the variable structure control law based on the index-based
PWM generating circuit gain K A = 24 can be obtained realization.
under the supply voltage U N = 24V  III. SIMULATION AND ANALYSIS OF THE EXPERIMENTAL
Antenna servo system is a multi-loop feedback control RESULTS
system, servo system tracking performance depends largely
on the control algorithm design. The core of the whole control system is with TI's TMS
In the discrete sliding mode variable structure control, 320F2812 DSP chip, taking the Mitsubishi IPM as intelligent
switching function takes the form of formula (3) shown: servo drive module. Experimental conditions: when the
speed is 5000r/min, using an Agilent oscilloscope.
se (k ) = Ce xe (k ) = Ce ( R(k ) − x(k )) ˄˅ According to the parameters of the controlled object model

Where e
C = c[ 1. 0 ]
, which can be obtained by pole
and sliding mode variable structure control, the system
simulation experiment is carried on in the CCS integrated
assignment method. environment.
based on design ideas reaching law, switching function For servo control object, respectively, simulating using
and 2-order discrete system presented respectively by variable structure control and the classic PID control,
formula (3) and (2) , s e (k + 1) − s e ( k ) is calculated as respectively. By changing parameters of the controlled object,
and were added to successive interference sudden load
below:
torque , the simulation results of the response curves of the
se (k + 1) − se (k ) two algorithms and parameters before and after comparison.
= Ce xe (k + 1) − Ce xe (k ) When disturbed continuously, the robust performance of
the two strategies is observed. Speed Ref the reference value
= Ce ( R(k + 1) − x(k + 1)) − Ce ( R(k ) − x(k )) 
is set to 5000 rpm motor speed under the circumstances, the
use of hand-held motor shaft to increase its friction torque of
= Ce R(k + 1) − Ce Ax(k ) the motor is approximately equivalent to the disturbance or
− Ce Bu (k ) − Ce R(k ) + Ce x(k ) sudden load torque conditions, to observe PID algorithm and
discrete sliding mode control algorithm robust performance
˄˅ curve based on the index reaching law. Sampling time
Index reaching law proposed by Gao Wen-bin[2] is interval of systems is T s = 0 .00005 s , antenna servo
expressed as the following formula:
controller speed loop uses PID control strategy and approach
se (k + 1) − se (k ) = −qTs e (k ) − εT sgn( se (k )), control strategy robust performance base on the index based
 
ε > 0, q > 0, 1 − qT > 0 law continuous interference, from CCS graphics window

254

Authorized licensed use limited to: Walchand College of Eng. Downloaded on January 09,2023 at 12:13:59 UTC from IEEE Xplore. Restrictions apply.
 2013 2nd International Symposium on Instrumentation and Measurement, Sensor Network and Automation (IMSNA)

displays real-time operating data were crawled Figure 2 - algorithm based on the index rate is used to improve the
Figure 4. adaptability of the servo system, compared to the traditional
Compared by Figures 2 and 3, when the derivative time PID control strategy, and simulation experiments are carried
constant T d is 0.0001s and 0.001s, the greater is T d  on the experimental platform what cores is the TI's TMS
320F2812 DSP chip. Experimental results show that
the movement speed of the system will speed up and reduce feasibility of the discrete variable structure control algorithm
adjusting time. It is helpful to reduce overshoot and is applied to the shipboard satellite antenna servo system and
overcome oscillation. However, compared with Figures 2 they verify the continuous interference and the sudden
and 4, it is not helpful to robustness to the sensitivity interference caused by an instability sailing ship. Compared
interference. The integration time constant T i is used to with the PID control algorithm, interference susceptibility is
eliminate static error, to improve the system diversity. When reduced by using the sliding mode control algorithm and the
it is 0.3s, the greater is Ti, the weaker is the integral role. The system has better robustness and has a certain significance in
bigger integration time constant T i can reduce the engineering.
overshoot of the system, but will slow down the system the REFERENCES
response process at the same time. From Figure 4, the [1] Young K D, Utkin V I, Ozguner U. A control engineer's guide to
differential coefficient reflecting the change of the error sliding mode control[C]//Variable Structure Systems, 1996. VSS'96.
signal, i.e., prevent the occurrence of deviation of the Proceedings., 1996 IEEE International Workshop on. IEEE, 1996: 1-
differential depends on the time constant. 14.
In terms of robustness, the curve in Figure 3 fluctuates [2] Gao W B, Wang Y F, Homaifa A. Discrete-time variable structure
control systems[J]. IEEE Transactions on Industrial Electronics,
more greatly, comparison of the experimental results from 1995,42(2):117-122㧚
Figure 2 and Figure 4 shows that the less effective the [3] Palm R. Sliding mode fuzzy control[C]//Fuzzy Systems, 1992., IEEE
conventional PID control algorithm, the disturbance torque. International Conference on. IEEE, 1992: 519-526.
It reflects the motor speed fluctuates greatly and is not [4] Zhang F, Tian K. Sliding-mode variable structure control of the high
suitable for the actual motor’s running. To the existed same precision antenna servo system[J]. Manufact Autom, 2012, 34(3): 43-
disturbance torque, the robustness performance curve of the 45.
control strategy based on the index reaching law fluctuates [5] Leghmizi S, Liu S, Fraga R, et al. Dynamics modeling for satellite
smaller robust performance, showed better control, when antenna dish stabilized platform[J]. Advanced Materials Research,
subjected to continuous disturbance of the same size. 2012, 566: 187-196.
(2)In the case of sudden load , the robust performance of [6] Bedford C C, Davis K B, Fahle J H, et al. Antenna positioner system:
U.S. Patent 8,305,279[P]. 2012-11-6.
the observed performance of the two strategies . Sampling
[7] Rafael M C, Gonçalves J B, do Prado P P L. DEVELOPMENT OF
interval system is still s T = 0 .00005 s . In the case of AN AUTOMATED SYSTEM FOR MANEUVERING PARABOLIC
DISH ANTENNAS USED IN SATELLITE COMUNICATION[J].
continuous interferences, the robust performance of the 2012.
antenna servo controller speed loop based on exponential [8] Zhao P, Yao M, Hu Y, et al. Stabilization Control for SOTM Antenna
reaching law is compared with the control strategies PID Based on Adaptive PID Sliding Mode[J]. Electronics Optics &
control strategies by sudden interference ,their running Control, 2012, 3: 015.
results in real-time are shown in Figure 5 and Figure 6 [9] Moreno J A, Osorio M. A Lyapunov approach to second-order sliding
respectively. mode controllers and observers[C]//Decision and Control, 2008. CDC
As shown above, simulation curves waveforms in Figure 2008. 47th IEEE Conference on. IEEE, 2008: 2856-2861.
6 are much gentler than that in Figure 5 subjected to the [10] Wu L, Shi P, Gao H. State estimation and sliding-mode control of
Markovian jump singular systems[J]. Automatic Control, IEEE
same sudden disturbance torque. It is also reflected in Figure Transactions on, 2010, 55(5): 1213-1219.
5 that the conventional PID control system algorithm to
[11] Gwo-Ruey Yu, Rey-Chue Hwang. Optimal PID Speed Control of
control the sudden disturbance in the adjustment time is too Brushless DC Motors Using LQR Approach[J]. IEEE International
large. Robust sliding mode control speed fluctuation curves Conference On Systems, Man and Cybernatics. 2004㧦473-478.
show the algorithm is much less than under conventional [12] K. David Young, Senior Member, IEEE, Vadim I. Utkin, Senior
PID control anti-windup, and here applied PID control Member, IEEE, and Umit Ozguner, Member, IEEE. A Control
parameters has undergone a number of debugging, Engineer̉Guide to Sliding Mode Control[J]. [4] IEEE Transactions
optimization. Experimental results show that the index On Control Systems Technology. Vol.7, NO.3, MAY 1999:328-
342㧚
reaching law of discrete variable structure control algorithm,
[13] Wu L, Ho D W C. Sliding mode control of singular stochastic hybrid
compared to the conventional PID control strategy, the systems[J]. Automatica, 2010, 46(4): 779-783.
sensitivity for weak interference, robustness enhanced with [14] Yu X. Sliding-mode control with soft computing: A survey[J].
better real-time adaptive Industrial Electronics, IEEE Transactions on, 2009, 56(9): 3275-3285.

IV. CONCLUSIONS
Aiming shipboard satellite antenna servo system may be
disturbed, a novel discrete variable structure control

255

Authorized licensed use limited to: Walchand College of Eng. Downloaded on January 09,2023 at 12:13:59 UTC from IEEE Xplore. Restrictions apply.
 2013 2nd International Symposium on Instrumentation and Measurement, Sensor Network and Automation (IMSNA)


Figure 1. Robust performance simulation curve PID control,where
K p = 1 .5 T = 0.3 ; Td = 0.0001 ; TC
; i
= 0 .2



Figure 2. Robust performance simulation curve PID control,
K p = 1.5 Ti = 0.3 Td = 0.001 TC = 0 . 2
where ; ; ;


Figure 3. Robust performance simulation curve index reaching law of discrete sliding mode control,
where c = 2000 ; q = 4000 ; ξ = 100 ; Sat = 1 .0


K p = 1 .5 Ti = 0.3 Td = 0.0001 TC = 0 .2
Figure 4. Robust performance simulation curve PID control,where ; ; ;



Figure 5. Robust performance simulation curve index reaching law of discrete sliding mode control,

where c = 2000  q = 4000  ξ = 100  Sat = 1 .0 

Authors
Li hong-ke˄1979—˅At present, automatic control and research of electronic information technology.
Huang lin-shu˄1975—˅,Hunan Province , At present, communication and information systems and radio wave
propagation aspects of research.
Xu kai˄1967—˅Associate Professor. The main research direction for the overall satellite technology.
E-mail: [email protected]

256

Authorized licensed use limited to: Walchand College of Eng. Downloaded on January 09,2023 at 12:13:59 UTC from IEEE Xplore. Restrictions apply.