Introduction:

The antenna azimuth position control system is crucial in control engineering because it determines and
maintains the horizontal orientation of an antenna. This system ensures that the antenna accurately points
toward a specific target or direction, such as a satellite or a communication source. In various fields like
telecommunications, radar systems, astronomy, and satellite communication, precise control over the
azimuth (horizontal) positioning of the antenna is vital for efficient signal reception, data transmission, and
information gathering. Control engineering ensures that the antenna maintains its intended azimuth position
despite external factors like wind, vibrations, or disturbances. Implementing robust control systems
enhances accuracy, reliability, and responsiveness, contributing significantly to the overall performance and
effectiveness of the communication or data acquisition process.

Project Objectives:
The purpose of this project is to get familiar with the control systems design, tools used to design the control
systems, technique of designing a control system and its application in the real-world.

➢ To apply the knowledge of the control system gained throughout the semester.
➢ To learn about the real-life application of the control systems.
➢ To learn about different techniques used to design a control system.
➢ To become familiar with the MATLAB tools in the design of control systems.
➢ To design lead compensator and PID controller to stabilize the transient response specifications as per
the requirements.
➢ To analyze and compare the results of the controllers.
Methodology:
The methodology includes the following steps:

First, we derived the open loop transfer function of the given system in the form of block diagram, checked
the step response of the open loop transfer function. We derived the closed loop transfer function of the
block diagram, applied step input and checked the response of the transfer function (closed loop system).
We designed the lead compensator using bode plot technique which uses the gain and phase margin of the
system to give the transfer function of the system and then we used sisotool command of the MATLAB to
design the lead compensator using root locus technique. Using the sisotool, we changed the value of gain
in the compensator editor.

We designed the PID Controller using the simulink model of the MATLAB and designed the PID
Controllers using integrators, differentiators and simple gain. Then we tuned the PID Controller to get the
desired transient response specifications. We compared and analyzed the results using compensator and
without the compensator.

System Description and Mathematical Modeling:

An antenna azimuth position control system consists of two potentiometers, a preamplifier, a power
amplifier, a motor, and a load. The position (direction) of the antenna can be changed via one potentiometer
(input transducer). The second potentiometer (output transducer) is used to obtain the actual azimuth
position of the antenna. The input potentiometer converts the desired angular position to a voltage, while
the differential amplifier provides the error signal between the desired and real antenna position signals. A
power amplifier is used to amplify the control signal applied to the motor. Antenna position control is
usually achieved with the help of a set of gears.

The detailed block diagram model of the antenna azimuth position feedback system is shown in figure:

The values of parameters used in the block diagram is:

Value of “a” will be: 50+23 = 73, because smallest registration no of this group is “22-EE-23”
 Part A
Open loop Transfer function

First, we derived the open loop transfer function which was very much unstable. The calculation is as
follows:

The MATLAB code and the pole zero analysis, step response, root locus plot, bode plot and the nyquist plot
is as follows:

MATLAB Code:

% Open Loop Transfer Function

k=1;

Kpot = 0.318;

K1 = 100

Kg = 0.2
Km= 2

a= 73

am=2.5

s = tf('s')

ts = Kpot*k*K1*Kg*Km/(s*(s+a)*(s+am))

step(ts)

Step Response:

The step response of the open loop system is unstable and results in a ramp output that will quickly saturate
the components of the system. This is due to a free integrator that exist in transfer function.
 Part-B
Close loop Transfer function
We derived the closed loop transfer function of the system and seen the step response of the system. The
calculation is as follows:

The MATLAB Code and step response of the system

MATLAB Code:

k=1;

Kpot = 0.318;

K1 = 100

Kg = 0.2
Km= 2

a= 73

am=2.5

num = K*Kpot*K1*Km*Kg;

den = [1 (am+a) (a*am) (K*Kpot*K1*Km*Kg)];

tcs = tf(num, den)

step(tcs)

Step Response:

Comments:
By hand calculations of open as well as close loop transfer function matches with the MATLAB provided
transfer function. Attached graphs of step responses for open and close loop system depicts that the system
is unstable at this stage
 System Analysis and Report
Part-A

Open loop pole-zero analysis, Root locus plot, Bode plot, Nyquist plots

Pole-Zero Analysis:

Root Locus:
Bode Plot:

Nyquist Plot:
 Part-B

Close loop Step Response

MATLAB Code:

k=1;

Kpot = 0.318;

K1 = 100

Kg = 0.2

Km= 2
a= 73

am=2.5

num = K*Kpot*K1*Km*Kg;

den = [1 (am+a) (a*am) (K*Kpot*K1*Km*Kg)];

tcs = tf(num, den)

step(tcs)

Step Response:
 Part-C

Based on (i) and (ii) steps above, comment on the system stability, type of response and
uncompensated time domain performance such as rise time, settling time, overshoot and steady state
error.

By looking at the response of the uncompensated system, we can say that:

Open loop transfer function shows that system is unstable due to ramp output. PZmap shows that system is
marginally stable because of one pole at origin. Root Locus also predict that system is less stable because
it branches towards right. But Bode plot is showing system is stable , why ? Because there is not any single
pole at right half plane so MATLAB considers this system as a stable system instead of unstable system or
marginally stable.

Close loop step response shows that our system can be regarded as an overdamped & is not showing any
unstability or any overshoot but still settling time and rise time is very larger.

Controller Designs and Simulation

The objective is to design two appropriate controllers (phase compensator and PID controller) to achieve
the following performance specifications:

1. Overshoot of less than 10%

2. Rise time is half of the uncompensated system or equal to (Tr<=0.5Tr (OL))

3. Settling time should be less than or equal to 2 sec. (Ts<= 2sec.)

4. Phase margin between 45 and 65 degree, and Gain margin of at least 6dB

Part-A

Phase Lead Compensator Design (Analytically)

MATLAB Code:

k = 1;

H = 1;

sys = tf(12.72*k,[1 755.5 182.5 0])

s = tf('s')

Gc =(12.18*s+1)/(34.17*s+1)

com = feedback(Gc*sys, H)

step(com)
figure

bode(com);

Step Response of Compensated System:

Through compensated step response we can observe that still settling time, rise time & overshoot
requirements are not fulfilled.
Bode Plot for Compensated System:

The compensated bode plot has clearly fulfilled the design requirement of GM & PM i-e:

• Gain Margin of 87.3 dB at frequency 12.1 rad/s.

• Phase Margin of 102 (deg) at frequency 0.051 rad/s

From seeing the bode plot, it is clear that the phase margin and the gain margin are not satisfied as per
requirements.

So we have to apply someother way to get the require specificatons. But the question is what would be the
alternate way of getting the desired specifications? The other way is to design the lead compensator using
Sisotool technique of design. This technique is used by deploying the root locus of the transfer functio in
sisotool.

Part-B

Phase Lead Compensator Design (Sisotool)

By using MATLAB command sisotool(openloop transfer function), we will get a window opened which
will enable us to design the lead compensator to get the desired specifications.

We will first go to the root locus tab and then we will add the Percentage Overshoot value and settling
time as per required specifications i.e.

• Overshoot of less than 10%

• Settling time should be less than or equal to 2 sec. (Ts<= 2sec.)
The new root locus after a new design:
To further adjust the design we need to add real zeros and poles on root locus, for this purpose we use
compensator editor of Sisotool as follows:

And we obtained the transfer function from this as:

MATLAB Code:
h=1;

k=1

sys = tf(12.72*k,[1 75.5 182.5 0])

K = 143.09;

sisotool(sys)

csis_sys = tf([K K*4.9], [1 25.4])

coms = feedback(csis_sys*sys, h)

step(coms)

stepinfo(coms)
figure

bode(coms)

margin(coms)

rlocus(coms)

Step Response:

Bode Plot:
Root Locus:

Through compensated step response we observe that requirements of %OS, Ts, Tr are fulfilled i-e:

• Percentage Overshoot (%OS) = 3.29 < 10%

• Rise time (Tr) = 0.908 < 0.5 (Tr (OL))
• Settling Time (Ts) = 2.3s

The reason for settling time to be slightly greater than than 2 is, because by nature this system has the a
slightly greater settling time for the chosen value ‘a = 50+registration number’

From Bode plot we can observe that design requirement of GM & PM remain fulfilled i-e:

• Gain Margin of 38.2 dB at frequency 38.9 rad/s.

• Phase Margin of 64.2 (deg) at frequency 1.95 rad/s

Part-D

Simulate the system with the designed compensator for unit step {u(t)} and 10 unit step input
{10u(t)}and observe the time domain performances (overshoot and settling time).

MATLAB Code:

h=1;

k=1

sys = tf(12.72*k,[1 75.5 182.5 0])

K = 143.09;

csis_sys = tf([K K*4.9], [1 25.4])

coms = feedback(csis_sys*sys, h)

step(coms)

Step Response for u(t):

MATLAB code for 10u(t)

h=1;
k=1

sys = tf(12.72*k,[1 75.5 182.5 0])

K = 143.09;

csis_sys = tf([K K*4.9], [1 25.4])

coms = feedback(csis_sys*sys, h)

t=[-0.5:0.01:3]
unitstep=zeros(size(t))

unitstep(t>=0)=10

lsim(coms,unitstep,t)
title("10u(t) system Response")

Step Response for 10*u(t):

As you can see there is a slight when we applied the unit step input with magnitude 10. But it is giving the
result as per requirements. arise due to the nonlinear effects or dynamic changes introduced by the
compensator in the system.

Part-B

PID Controller Design

We used ziggler nicholas method to design the PID Controller which involves the following steps:

• First have any value of the proportional gain initially.

• Then ultimately increase the value of the Proportional gain and see the value for which the output
sustains the oscillation.
• Then that value will be called as critical value Ku (Critical Value of gain).
• Then we have to find the value of the time period of the sustained oscillatory wave/signal named as
Pu.
• Then we will find the values for KP, KI and KD using formulae.
• Placing these values in the gain will give us the PID controller.

The above method is the one way to design the PID Controller. There are many other ways to design the
controller such as tuning method of PID Controller.

The simulink circuits are as follows:

The waveforms of the circuits are as follows respectively:
For values of KI, KP and KD, we have a table of formulae which we will use in the code for getting the
value.

So, writing the MATLAB code as per formulae for PID Controller, we get:

MATLAB Code:

Ku=1075;

Pu=0.465; % It is the time period of the sustained oscilating wave

% Look up for appropriate value for KP,Ti,Td

Ki=Ku*0.6;
Ti=Pu/2;

Td=Pu/8;

% Compute the value of KP, Ki and Kd

KP = Ki

KI=KP/Ti

KD=KP*Td

The result is shown below:

KP =645, KI =2.7742e+03, KD =37.4906

Now, putting these gain values, we get:

The wavefotm of the above simulink circuit is as follows:

Now implementing it using MATLAB, we will get:

The response of both compensated and uncompensated response (both) is:

As it is seen that when we applied the PID Controller, the step response does not meet the requirements.
So, in order to to meet the specifications, we have to tune the PID controller. The tuning will be done by
selecting the block of PID controller in the simulink. So, doing this we will get new values of KP, KI and
KD.

The step respone we got is:

Now, we will write matlab code and see the step response of the system with PID controller i.e. tuned
controller.

k = 1;

plant = tf(12.72*k,[1 75.5 182.5 0]);

sys1=feedback(plant,1)

kp = 251.8953

ki = 240.6821

kd = 64.2582
compensator=pid(kp, ki, kd);

sys2=feedback(plant*compensator,1)

step(sys2);

legend(Compensated')

The ouput will be:

So, our requirement is fulfill now with the satisfying results:

%OS <10 and Ts < 2.

Now we will see the bode plot of the system compensated with the PID Controller:

Uncompensated System Analysis in Time and Frequency Domain:

PM & GM Rise Time Settling Time Overshoot ess

GM = 53.9
Uncompensated 20.21 36.40 0 0.0032
PM = 87.5

Analysis:
The uncompensated system is overall not stable but good, as we observe the root locus plot, pzmap (pole-
zero analysis), we see that open-loop two poles are at left half of s-plane and one is at origin which mean
system is marginally stable. It can either be unstable or marginally stable depending on how many poles
are there at origin. But they will never remain strictly stable. If it is single pole, i.e, s = 0, the
corresponding time domain graph is of constant straight with zero slope line which neither diverges nor
converges.

Time domain performance (observed from close loop) is not very good such as rise time, settling time and
overshoot is 20.21 sec, 36.40 sec & 0 respectively.

Bode plot is showing the value of phase margin (PM=87.5), gain margin (GM=53.9) for open loop
system. The values of PM and GM does not fulfil our requirements, in-short not fulfilling our
requirements in time and frequency domain.

Therefore, we need to design a phase compensator which fulfils our requirements.

Compensators Design and Evaluation

PM & GM Rise Time Settling Time Overshoot ess

GM = 62.4
Compensator 1 24.7 93.8 13.2 2.95e+04
(Analytical) PM = 61.5

GM = 29.1
Compensator 2 0.508 1.78 8.85 0.0055
(Sisotool) PM = 61.1

Evaluation of Phase lead Compensators:

In part 03 of this PBL we first designed the phase lead compensator analytically (by hand) then implanted
on MATLAB. This model satisfied the conditions of our gain and phase margin, but still other time
domain parameters were not fulfilled.

So we designed another compensator by using MATLAB Sisotool, in which we changed the position of
zeros and poles, this amended our first compensator. We observed in the second design all the asked
requirements are fulfilled. This makes it the most suitable option for designing antenna azimuth position
control system.

In second compensator the performance of our system is very good than the first one as well as original
one. We can see as settling time reduce from 36.40 sec to 1.78 sec. Similarly this compensated system is
faster as rising time is 0.508 sec instead of 20.21 sec.
Evaluation of PID controller

PM & GM Rise Time Settling Time Overshoot ess

GM = 41.2
PID controller 0.165 1.9 8.34 0.0017
PM = 74

According to next design constraint of our PBL we had to design a PID controller which can improve our
performance parameters, so we completed this task by using Simulink blocks and constructed a simple
PID controller as provided.

This controller performs robustly and satisfies time domain requirements of our system, we can observe
that settling time reduces from 36.40 sec to 1.9 sec, overshoot is maintained below 10%, and system is
now faster than ever as rising time is now 0.165 sec instead of 20.21 sec.

But PID controller design doesn’t meet our frequency domain requirements i-e GM & PM which makes
this controller not the best option for designing a position control system.
 Results and Comparative Analysis

PM & GM Rise Time Settling Time Overshoot ess

GM = 53.9
Uncompensated 20.21 36.40 0 0.0032
PM = 87.5

GM = 29.1
Compensator 1 0.508 1.78 8.85 0.0055
(Sisotool) PM = 61.1

GM = 41.2
Compensator 2 0.165 1.9 8.34 0.0017
(PID) PM = 74

Analysis:
Comparison between uncompensated and compensated system can be observed clearly, without
compensator phase margin didn’t lie in the required range but after adding the appropriate compensator
(Sisotool) in the system both phase margin and gain margin fulfils our requirement, for this purpose we
used Bode plot for analysis in frequency domain.

Now come towards time domain, we observe a clear difference between uncompensated and compensated
system rise time, settling time and overshoot.

• In uncompensated system rise time, settling time and overshoot was 20.21 sec, 36.40 sec & 0
respectively.

• In compensated system 01 i-e Sisotool, the rise time, settling time and overshoot is 0.508 sec, 1.78
sec and 8.85% respectively and most importantly it fulfils our requirement in time domain.

• In compensated system 02 i-e PID controller, the rise time, settling time and overshoot is 0.165 sec,
1.9 sec and 8.34% respectively, this also fulfills our requirements.
• The last thing for comparison is steady state error (ess), which is observed crucially:

• In uncompensated system the value of steady state error is 0.0032 which is approximately equal to
zero.
• In compensated systems the error values are 0.0055 & 0.0017 respectively, both are satisfying.
• In terms of stability we say that uncompensated system is not stable in other hand compensated
system is stable and efficient. Compensator shifts the root locus to the left, which enhances the
responsiveness and stability of the system.
• Hence the two most optimized systems are shown with comparative analysis i-e phase lead
compensator and the PID controller either of which can be used for antenna azimuth position
control system, but the best performance is achieved by Phase lead compensator as we have seen.
Conclusion
The antenna azimuth position control system project effectively demonstrated the practical application of
control system design principles in achieving desired system performance. Starting with the derivation of
open-loop and closed-loop transfer functions, the analysis showed the system's inherent instability in the
open-loop configuration, with improvements in the closed-loop system performance.

To meet the specified requirements, phase lead compensators and a PID controller were designed and
implemented. The analytical design of the phase lead compensator showed moderate success but failed to
fulfill all performance criteria. However, the design using MATLAB's Sisotool refined the compensator,
achieving desired specifications for overshoot, rise time, and settling time.

The PID controller, initially designed using the Ziegler-Nichols method, required further tuning to meet the
transient response specifications. Once tuned, the PID controller provided a stable and optimized response,
fulfilling all time-domain and frequency-domain requirements, such as overshoot less than 10% and settling
time within 2 seconds.

Comparative analysis revealed that the Sisotool-designed compensator offered a balance between time-
domain and frequency-domain performance, while the tuned PID controller ensured enhanced stability and
precision. Ultimately, the project highlighted the importance of iterative design, simulation, and
optimization in control system development, showcasing robust techniques and tools like MATLAB for
real-world engineering applications.