SLIDING MODE CONTROLLER DESIGN FOR

CONTROLLING THE SPEED OF A DC MOTOR

1
Vyoma Singh, 2Dr. Veena Sharma
1
M.tech student (Signal Processing and Control)
Electrical Engineering Department, NIT Hamirpur, India
2
Associate Professor Electrical Department NIT Hamirpur, India  

 
ABSTRACT: The objective of this paper is to fail, they do not provide the desired
control the speed of a DC motor efficiently output in desired time. For this reason, some
by using SMC. Model of a separately excited better approach has to be taken. In this paper,
DC motor has been choose. First PID SMC (Sliding Mode Controller) is that
controller is used for the speed con- trol of approach, a controller which has ability to
the motor (both linear and non-linear model) work satisfactorily under linear as well as
and then SMC is used for the same non-linear factors of the system.
non-linear model. The modeling and
simulation has been done in MATLAB. This paper has taken two types of non-linearities
that are present in a DC motor. One of them is
The observations shows that SMC gives friction that occurs in the motor itself and the
better results as compared to PID. While other non-linear factor is backlash. Backlash is
PID is not capable to give the desired result generated at the gear box which is driven by the
satisfactorily in case of non-linear systems, motor.
SMC has robust nature in presence of such
disturbances. When the motor is required at low speed level
KEY WORDS: Sliding mode controller, such as in chemical processing application in
chattering, non-linearity in dc motor, industries, the non-linear factors do affect the
modeling of dc motor. output of the system that can not be eliminated
I. INTRODUCTION  by the conventional control- lers in real time.
DC motors have been widely used in industrial, SMC is designed for avoiding such un-
agricultural and domestic sectors with various certainties. It has many advantages over PID
applica- tions. Different fields require different controller. Some of them are robustness,
speed ranges of the motors with different
ranges of load changes. Usual- ly where a disturbance rejection and independent of he
higher speed of dc motor is required, the non system parameters. The results shown in the
– linear factors of the motor are neglected since
these factors do not affect much. Also, in such paper proves that SMC is much better than PID
cases, conven- tional controllers such as PI for the speed control of a DC motor
(Proportional-Integral) con-troller and PID especially when the non-linear model of the
(Proportional-Integral-Derivative) con- troller
are effective. motor is considered
When we come across non-linear systems
the above controllers are likely to II. SLIDING MODE CONTROL (SMC)
Sliding mode controller works by switching the
trajectory of the system from one structure to
other and in between sliding on a specific line,
plane or surface in state space. The motion of
the system trajectory along a chosen path in

ISSN(PRINT):2394-3408,(ONLINE):2394-3416,VOLUME-2,ISSUE-1,2015
94
 INTERNATIONAL JOURNAL OF ADVANCED COMPUTING AND ELECTRONICS TECHNOLOGY(IJACET)

state space is called the sliding mode and the d(s2)/dt < - ɳ|s| -6
controller designed with the aim to achieve the
slid- ing motion is called sliding mode When the state trajectory changes its
controller. The path such chosen is called the structure from one path to other ie; as it is on the
sliding surface or switching surface. verge to slide on the sliding surface, due to the
stronger condition been applied to it, the
While choosing the sliding surface, there are transition from one structure of the path to the
some requirements that has to be taken care of. other is not smooth. In fact it forms a zig
-zag path along the sliding surface. This
The system stability has to be confined to the happens due to the practical limitations of the
sliding (switching) surface. Also, the system devices used for the
measurements. Such an undesirable
trajectory should converge to the sliding surface phenomenon is known as chattering.
within finite time
REQUIREMENT I. Chattering occurs due to the presence of
System stability: Consider the following system the sig- num function used in the controller.
Hence, to avoid or reduce chattering, signum
dynamics   function has to be replaced by some smoother
ẋ1 = A11 x1 + A12 x2 -1 functions may be tangent function or saturation
ẋ2 = A21 x1 + A22 x2 + Bu -2 -2function.
- 1 ẋ2 =Many
A21 xtypes of functions
1 + A22 x2 + Bu -2have been
used in different papers for reducing the
Let the sliding function chattering ear- lier and many are still working
chosen be : on it.
s = Cx1 + x2 = 0 -3 -3III. DC MOTOR ( LINEAR MODEL)
Equation 1 can be DC motors have a   wide area of application in 
rewritten as: all sorts of sectors, be it industrial, agricultural 
ẋ1 = A11 x1 - A12 Cx1 = (A11 - A12C)x1 -4 or domes‐ tic. Moreover, they have been used 
-4in  scientific  re‐  search  and  technology  works 
Taking C > 0 such that the solution of equation 4 too.  The  electric  circuit  of  the  DC  motor  is 
lies to the left hand side of the phase plane, the shown  in  Fig.  1.  Objective  here  will  be 
dynam- ics of x1 can be made stable at the sliding controlling  the  speed  of  the  motor  through 
surface. From equation 3 the dynamics of x2 arma‐ ture voltage. 
already becomes stable at the sliding surface.  
Thus this sliding surface validates the first
requirement.

REQUIREMENT II.
Convergence to s = 0 : For this requirement to
be ful- filled, the reachability condition has to
satisfied which states that in order to make sure
that sliding mode starts at some time t > 0,
irrespective of the initial state x(0), we should be
sure that the state trajectory is always moving
towards s = 0, for s ≠ 0.
 
d(s2)/dt < 0 -5 -5
Figure 1. Electric circuit of DC motor
As the state trajectory comes closer to the
sliding surface, the rate of convergence The parameters shown in figure 1. are given
becomes slow. So along with reachability
condition, there is one more condition called ɳ below: Va : armature voltage (input to the
condition that ensures the conver- gence of the
trajectory towards s = 0 within finite time as the motor)
trajectory approaches towards the same. For the Ra : armature resistance
example taken above, its ɳ condition can be La : armature inductance
written as: Eb : back emf
ω : angular velocity (output of the motor)

ISSN(PRINT):2394-3408,(ONLINE):2394-3416,VOLUME-2,ISSUE-1,2015
95
 INTERNATIONAL JOURNAL OF ADVANCED COMPUTING AND ELECTRONICS TECHNOLOGY(IJACET)

if : field current
J : moment of inertia
b : constant of moment of inertia
T : torque produced

The motor rotates with a constant speed

which is gained by balancing the load torque
and the torque pro- duced by the motor. The
motor is excited by the field current with input
as armature voltage applied at the terminals. Figure 2. Simulink model of DC
This induces a voltage in the motor known as motor (linear)
back emf. The output speed of the motor rotates
the shaft attached to it which is actually the
load. IV . DCMOTOR(NON-LINEARMODEL)

The equations that govern the working of Two non-linearities considered in this paper
the motor are written as: are friction and backlash. Friction has three
com- ponents namely static friction,
Torque in terms of shaft parameters:  coulomb friction and viscous friction.
Whenever two surfaces are in contact they are
T = J(dω/dt) + Bω -7 -7having some static friction which one surface
has to overcome so as to move. Static friction
Torque produced by armature current: is also called stiction. As the surface of one
object starts to move it glides over the other
T = Kt ia -8 -8object with some velocity and in a direction.
Op- posite to this velocity and direction,
Now applying KVL in the circuit shown in coulomb force is always present which is
fig- ure 1., we get: independent of the mag- nitude of the
velocity. It depends on the nature (properties)
Va - Eb = Raia + La(dia/dt) -9 of the two surfaces in contact. The static
friction is greater than the coulomb friction.
(dia/dt) = Va /La - Eb /La - (Ra/La) ia -10 The friction proportional to the velocity is
viscous friction. It has the same direction as
Where , Eb = Kb ω -11
that of the moving object ie; the velocity.
And ,(dω/dt) = - (b/J )ω + (Kt/J) ia Below is a figure describing the relationship
-12
between the three com- ponents of the friction
With ω and ia as state variables, Va as the and the velocity.
input, the state model of the motor can be
represented by equations 10 and 12. Finally
by putting the values in equations 11 and 12,
the system model becomes:
(dω/dt) = - ω + 47.9041916 ia -13

(dia/dt) = - 66.66667ω - 50ia + 83.3333Va -14

Equations 13 and 14 are obtained by using
the following values:
Ra = 0.6Ω La = 0.012H Figure 3. Friction - velocity plot
Kt = 0.8Nm/A Kb = 0.8Vs/rad Backlash occurs in every mechanical system
J = 0.0167Kgm2/s2 b = 0.0167 that has some movement. It is like hysteresis
(that occurs in electric systems), but in
Simulink model of DC motor (linear) is
shown in fig- ure 2. below. mechanical sys- tems. When the motor
drives the shaft, it rotates with certain speed
and the gear box attached pro- duces
backlash. The plot of input angle (driving
motor angle output) vs output angle (driven

ISSN(PRINT):2394-3408,(ONLINE):2394-3416,VOLUME-2,ISSUE-1,2015
96
 INTERNATIONAL JOURNAL OF ADVANCED COMPUTING AND ELECTRONICS TECHNOLOGY(IJACET)

shaft angular position) is shown below in

figure 4. VI. SI M UL ATI O N RESULTS

First comparison between the output of PID

controller for DC motor linear and
non - linear model has been done. Then the
results of PID and SMC for non - linear DC
motor model are compared

Figure 4. Backlash in gear box

Here ’a’ is the perpendicular distance be- tween
the teeth of the motor driving the shaft and the
shaft been driven (as an example).

The Simulink model of DC motor is taking into

consideration the friction and backlash non-
linearities, is shown in figure 5. Below Figure 6. Modeling of PID controller for
DC motor speed

Figure 5. Simulink model of DC motor (non-linear)

V. CONTROLDESIGN Figure 7. Speed control of linear

DC motor model using PID
controller
Take x1 = ω(t) ; x 2 = x1 ; u = Va
 
The system then becomes:

ẋ1 = x2 -15
ẋ2 = - 51.39 x1 - 51 x2 + 3992.015 u - 16 y = x1
-17

Selecting the sliding surface as:

s = Ce + x2 = 0 -18

where C is negative and

e = reference speed - actual speed e =
 
r - x1
Figure 8. Speed control of non-linear
DC motor model using PID controller
The controller ‘u’ is defined as:  
 
u= K , s<0 -19
=-K , s>0

where K should be a positive value. In the simu

lation given below, K = | C | .

ISSN(PRINT):2394-3408,(ONLINE):2394-3416,VOLUME-2,ISSUE-1,2015
97
 INTERNATIONAL JOURNAL OF ADVANCED COMPUTING AND ELECTRONICS TECHNOLOGY(IJACET)

be efficient than PID controller, be it in terms

of peak overshoot or the settling time
REFERENCE S 

. Vadim I. Utkin, “Sliding Mode Control

Design Prin- ciples and Applications to
Electric Drives”, IEEE Transactions on
Industrial Electronics, February
1993.
  2. K. David Young, Vadim I. Utkin and U mit
Figure 9.  Speed control of DC motor using PID  O zgun- er, “A Control Engineer,s Guide
to Sliding Mode Control”, IEEE
controller  Transactions on Control Systems
  Technology, Vol.7, No.3, May 1999.
     3. Mohammed Godam Sarwer, Md. Abdur
Rafiq and B.C.Gosh, “Sliding Mode
  Speed Controller of a D.C. Motor Drive”,
    Journal of Electrical Engineer-
ing, The Institution of Engineers,
  Bangladesh, Dec
  2004.
  4. Uma Maheshwararao, Y.S.Kishore
Babu, K.Amaresh, “Sliding Mode Speed
  Control of a DC
Figure 10. Modeling of SMC for DC motor Motor”, 2011 International Conference
on Com-
speed munication Systems and Network
Technologies.
5. Ned Mohan, "Electric Drives: An
Integtalive Ap- proach”, MNPERE
Minneapolis, 2001.
6. J. K. Liu and F. C. Sun, "Research and
development on
theory and algorithms of sliding mode
control", Journal
of Control Theory and Applications, vol.
24, no. 3,
2007.
 
 
Figure 11. Speed control of DC model
(non-linear) using SMC
VII. CO NCLUSI O NS
In this paper conventional controller PID
and a different approach SMC have been
applied for speed control of DC motor. As far
as linear model of the motor is concerned,
PID is capable to control the speed within
desirable time and er- rors under limitations.
But when it comes to con- trol the speed
under the presence of non- linearities, SMC
is far better than PID. However, results of
PID controller can be improved by changing
the gains of P, I, D terms but SMC proves to

ISSN(PRINT):2394-3408,(ONLINE):2394-3416,VOLUME-2,ISSUE-1,2015
98