Introduction

Control systems is an interdisciplinary field that encompasses various engineering

specialties and finds application in everything from large industrial robots to tiny
electronic circuits.
What are Control systems:
Let's first define what a system is, before learning what control systems are. A system
is an arrangement of components designed to carry out a particular function. A
system responds in a certain way when an input is used to stimulate it. Examples of
systems include an air conditioner, a car, a resistor, and a capacitor. A system may
occasionally be referred to as a "plant." A control system is a device that controls the
output of the systems and directs the input that passes through it.

The diagram above illustrates the structure of a control system, which is designed to
modify the response of a plant or system as needed. To elaborate, consider a scenario
where we have a system to regulate, such as a motor requiring position control. In
this case, we employ a servomechanism, acting as the control system or controller,
to provide specific input to the motor regarding the desired rotation. The ultimate
aim is for the entire system, comprising both the plant and the controller, to achieve
the intended objective.
There are mainly two types of Control system:
1. Open loop control system:
A control system with an open loop is simple to operate. In this case, whether or not
we obtain the intended result has no bearing on the control action. To put it briefly,
no feedback is required.
We are unsure if the intended result is achieved in this instance because there may
be more disruptions while the control process is underway. These control systems
are less dependable because their faithfulness is largely dependent on how accurate
the initial calibration was. We should use open loop control since it would be far less
expensive in some circumstances where output accuracy is not a concern.
Controlling basic traffic lights, where each light's operation is solely dependent on a
set time, would be a good example.
2. Closed Loop Control system:
Closed-loop control systems, also referred to as feedback control systems, operate
based on the desired output. In these systems, the control action is influenced by the
feedback received from the output. It's important to note that closed-loop control
systems can incorporate one or multiple feedback paths to enhance their
functionality and responsiveness.

In a closed-loop system, the output is continuously compared with the reference

input, resulting in the generation of an error signal. This error signal is then fed back
to the controller, which takes corrective actions to minimize the error and achieve
the desired output. This iterative process contributes to the system's reliability and
ensures high output accuracy. While closed-loop control systems are known for their
dependability and precision, it's important to consider that the associated costs and
system complexity may increase. However, the benefits often outweigh these
drawbacks in many practical applications.
Controller in the Controller System:
A lot of controllers are used in the control system but the most famous and mostly
used controller are the PID and I-PD controller, because of its flexible nature and
easy to implement and easy to design,
Here, we discuss both these controllers and then design our electromechanical
system on the bases of these two controllers.
a) Proportional Integrator derivative PID:
Here I will explain each of the controller P, I and D individually:

1. Proportional:
The proportional component of a control system relies solely on the disparity
between the set point and the process variable, commonly referred to as the error
term. The proportional gain (Kc) defines the relationship between the output
response and the error signal. For example, if the error term has a magnitude of 10,
a proportional gain of 5 would result in a proportional response of 50. Generally,
elevating the proportional gain enhances the speed of the control system response.
However, excessive proportional gain may lead to oscillations in the process
variable. Further increasing Kc can intensify these oscillations, ultimately causing
system instability and potential loss of control.
 2. Integrator Response:
The integral component of a control system involves accumulating the error term
over time. This means that a small error term will lead to a gradual increase in the
integral component. The integral response will persistently grow over time unless
the error becomes zero. Consequently, the integral component aims to minimize
steady-state error, which is the ultimate difference between the process variable and
the set point. However, a situation known as integral windup may occur when
integral action saturates a controller without driving the error signal toward zero.
3. Derivative Response:
The derivative component of a control system is responsible for reducing the output
when the process variable experiences rapid increases. The derivative response is
directly proportional to the rate of change of the process variable. Adjusting the
derivative time (Td) parameter influences how the control system reacts to changes
in the error term, impacting the overall speed of the control system response. In
practice, most control systems employ very small derivative time values (Td)
because the Derivative Response is highly sensitive to noise in the process variable
signal. If the sensor feedback signal exhibits noise or if the control loop rate is too
slow, the derivative response can potentially destabilize the control system.
b) Integral – Proportional derivative I-PD:
The I-PD control algorithm is often overlooked and not fully utilized. In contrast to
the PI-D, it employs proportional action based on the process value (PV) rather than
the error. Despite common misconceptions, it is not limited to providing a slow
response to set point (SP) changes; any control algorithm can achieve this by
adjusting the tuning. Figure 1 illustrates a well-tuned controller's response to an SP
change (Curve A), while switching to I-PD yields a slower response (Curve B). The
I-PD algorithm no longer generates the proportional kick, relying solely on integral
action. The perception of I-PD as slow stems from using tuning parameters designed
for the PI-D algorithm, highlighting the importance of appropriate tuning for
different algorithms.
Calculations
Now we will design our Electromechanical system first with doing some
calculations then verify the result with Simulink simulation.
1. System modeling

Let’s explain the system terminology and the give data:

1. R = winding resistance
2. L = winding inductance
3. e = back emf of the DC motor
4. v = supply voltage
5. TL = load torque
6. T = Motor torque
7. w = angular velocity of shaft
8. b = Motor viscous friction constant
Given equations:
𝒆 = 𝑲𝒆 𝝎, 𝑻 = 𝑲𝒕 𝒊 𝒂𝒏𝒅 𝑻𝑳 = 𝑲𝒕 𝒊𝑳
In the given context, Ke and Kt represent the back electromotive force (emf) and
torque constants of the motor, respectively. The variable iL represents the equivalent
load current reflected by the load torque TL to the motor armature. The figure
mentioned above represents a combination of both electrical and mechanical
systems, resulting in two equations: one for the mechanical system and another for
the electrical system. First, we derive the Electrical system equation:
Lets apply the KVL on the electrical system: so we get the equation.
𝑑𝑖
𝑣 − 𝑖𝑅 − 𝐿 ( )−𝑒 =0
𝑑𝑡
Simplifying this equation
𝑑𝑖
𝑣 = 𝑖𝑅 + 𝐿 ( )+𝑒
𝑑𝑡
laplace transform of this equation
𝑉(𝑠) = 𝐼(𝑆)𝑅 + 𝑠𝐿𝐼(𝑠) + 𝑒(𝑠) − − − − − 1
Note
Current related to the load is:
𝑇
𝑖=
𝐾𝑡
Laplace transform is
𝑇(𝑠)
𝐼(𝑠) =
𝐾𝑡
Now:
𝑒(𝑠) = 𝐾𝑒 𝑤(𝑠) = 𝐾𝑒 𝑠𝛳(𝑠)
We get the equation by Putting the value of e(s) and I(s) in the equation 1 ;
𝑇(𝑠) 𝑇(𝑠)
𝑉(𝑠) = ( ) 𝑅 + 𝑠2𝐿 ( ) + 𝐾𝑒 𝑠𝛳(𝑠) − − − −2
𝐾𝑡 𝐾𝑡
This electrical equation is in the term of angular displacement and torque that motor
applied.
Mechanical Equations:
Using the free body diagram of the mechanical load to get the mechanical equation:
The equation for the Free body Diagram:
𝑇(𝑠) = (𝐽𝑠 2 + 𝑏𝑠)𝛳(𝑠) − − − − − −3
Now, we have the both the equation so have to move for deriving the transfor
function for the required values.
Therefore, we put the equation 3 to equation 2 and we get the equation is
𝑅 𝑠𝐿
𝑉(𝑠) = [( + ) (𝐽𝑠 2 + 𝑏𝑠) + 𝐾𝑒 𝑠] 𝛳(𝑠)
𝐾𝑡 𝐾𝑡
Assuming that the armature inductance (La) is relatively small compared to the
armature resistance (Ra), as is often the case in DC motors, the equation transforms
into the following form:
𝑅
𝑉(𝑠) = [( ) (𝐽𝑠 2 + 𝑏𝑠) + 𝐾𝑒 𝑠] 𝛳(𝑠)
𝐾𝑡
Now after the solving the equation for the transfer function of 𝛳(𝑠) 𝑎𝑛𝑑 𝑉(𝑠) we
gets,
𝛳(𝑠) 1
=
𝑉(𝑠) 𝑅
[( ) (𝐽𝑠 2 + 𝑏𝑠) + 𝐾𝑒 𝑠]
𝐾𝑡
Solving this equation, we gets
 𝐾𝑡
𝛳(𝑠) 𝐽𝑅
= − − − − − −( 𝑎
𝑉(𝑠) 𝐾𝑡 𝐾𝑒
𝑠 [𝑠 + 𝐽(𝑏 + ]
𝑅
in the simplified for we can write this equation as
𝛳(𝑠) 𝐾
=
𝑉(𝑠) 𝑠[𝑠 + 𝛼]
For the transfer function of the w(s) and V(s) we get
As we know that ϴ(s) = w(s)/s

𝑤(𝑠) 𝐾
=
𝑉(𝑠) [𝑠 + 𝛼]
Now the equation become,
𝐾𝑡
𝑤(𝑠) 𝐽𝑅
= − − − − − − − (𝑎
𝑉(𝑠) 𝐾𝑡 𝐾𝑒
[𝑠 + 𝐽(𝑏 + ]
𝑅
Putting the value of the physical parameter we get the both equation as
The physical parameters are:
I. Moment of inertia of the rotor J=0.01 kg.m2
II. Motor viscous friction constant b=0.1 N.m.s
III. Electromotive force constant Ke=0.01 V/rad/sec
IV. Motor torque constant Kt=0.01 N.m/Amp
V. Electric resistance R=1 Ohm
Eq 1:
ϴ(s) 1
=
V(s) s[s + 0.001001]
Eq 2:
w(s) 1
=
V(s) [s + 0.001001]
Closed loop system:
Now the derive the equation for the above closed loop system
G(s) Kt
T(s) = ( + K p + K d s) ( )
s s
G(s)K t + K p K t s + K t K d s 2
T(s) =
s2
After putting the value of G(s) we get the simplified equation :
Kt KpKt
T(s) = + + KtKd
s 2 (s + 0.001) s

Determining gains
we determine the gain using the pole position methods:
here
𝑇(𝑠) = 𝐺(𝑠)𝐻(𝑠)
For closed loop system the characteristic equation will be like this
1 + 𝐺(𝑠)𝐻(𝑠) = 1 + 𝑇(𝑠) = 0
Let’s we want the pole at the position of -3 and -3, then the characteristic equation
become,
(𝑠 + 3)(𝑠 + 3) + 𝑇(𝑠) = 0
(𝑠 + 6𝑠 + 9) + 𝑇(𝑠) = 0
Now putting the equation of the T(s)
 𝐾𝑡 𝐾𝑝 𝐾𝑡
(𝑠 2 + 6𝑠 + 9) + + + 𝐾𝑡 𝐾𝑑 = 0
𝑠 2 (𝑠 + 0.001) 𝑠
Match Coefficients:
The I-PD Gains values are:

𝐾𝑝 = 6,
𝐾𝑡 = 1,
𝐾𝑑 = 9,

Now, we used this value in our simulation model and observed the system response.
Simulation of the Electromechanical system in the MATLAB Simulink
Simulation model of the system in the MATLAB Simulink with I-PD controller

The system response to the step input is given below:

The outcomes were derived based on the calculated values discussed earlier. In this
particular case, the system is characterized as a critically damped system, ensuring
there is no overshoot. However, it's noteworthy that such a system tends to have a
higher rise time and an extended settling time. To enhance the speed of the system
response, the Ki gain was increased. Unfortunately, this adjustment resulted in
overshooting in the system, as illustrated in the figure below.

Therefore, an increment in the Ki value introduces overshooting in the system, albeit

with an improvement in the system response. This scenario presents a trade-off
between the system response and overshoot. The decision on which characteristic to
prioritize depends on the specific requirements and considerations of the application
at hand.
Note:
By the consider the give value of the settling time rise time and system overshot then
the integrator gain should be increase to the achieve the given values. As shown in
the figures
Trial-error rules for PID tuning --- Ziegler-Nichols approaches:
The procedure for determining the optimal values of P, I, and D to achieve an ideal
response from a control system is known as tuning. Various methods exist for tuning,
and two commonly discussed approaches are the "guess and check" method and the
Ziegler-Nichols method.
Tuning the gains of a PID controller involves a trial-and-error method, which
becomes more straightforward as engineers grasp the significance of each gain
parameter. Initially, the I and D terms are set to zero, and the proportional gain is
incrementally increased until loop output oscillates. Care is taken to avoid system
instability. After setting P for a desired fast response, the integral term is raised to
eliminate oscillations, albeit at the cost of increased overshoot. Some overshoot is
essential for a swift response. Fine-tuning the integral term minimizes steady-state
error. Once P and I achieve the desired fast control system with minimal steady-state
error, the derivative term is increased for quicker convergence to the set point.
However, increasing the derivative term reduces overshoot, providing higher gain
with stability but increasing sensitivity to noise. Engineers often navigate trade-offs
between different control system characteristics to align with specific requirements.
The Ziegler-Nichols method, akin to trial and error, is a popular approach for tuning
a PID controller. It involves setting I and D to zero and incrementally raising P until
oscillations begin. Upon oscillation onset, critical gain (Kc) and oscillation period
(Pc) are recorded. Adjustments to P, I, and D are then made based on the
corresponding values in the table provided below
Calculation for PID gains:
Following the above procedure we calculated the Kp, Ki, and Kp value.
For stable oscillation of the system.

This is stable oscillation for the Kp value of 10, and the oscillation time period is 2
seconds.
Thus 𝐾𝑢 = 1 𝑎𝑛𝑑 𝑇𝑢 = 0.6𝑠𝑒𝑐
Now according to the above table we find the Kp, Ki and Kd.
𝐾𝑝 = 0.6𝐾𝑢 = 0.6 ∗ 1 = 0.6
Ki = ?
𝑇𝑢 0.6
𝑇𝑖 = = = 0.3
2 2
𝐾𝑝 1.2 ∗ 𝐾𝑢
𝐾𝑖 = = =2
𝑇𝑖 𝑇𝑖
Kd = ?
𝐾𝑑 = 0.075 ∗ 𝐾𝑢 ∗ 𝑇𝑢 = 0.045
Now the simulation Model for the PID controller:
Simulation results

Implementing the PID controller using the Ziegler-Nichols method yields optimal
system response. However, reducing the Ki gain introduces oscillations, while
increasing it leads to system overshot. In summary, for superior system response, the
PID controller excels. Conversely, if minimizing overshoot is crucial, the I-PD
controller emerges as the preferred choice.
Conclusion:
Utilizing MATLAB Simulink for implementing a control system is considered
highly effective. In this scenario, both PID and I-PD controllers were implemented,
and their results were verified against the calculated values. Notably, the I-PD
controller exhibited the lowest overshoot, but it also showed a somewhat lower
system response. In comparison, the PID controller demonstrated a rapid response;
however, increasing the Ki value led to overshooting in the system. This situation
presents a tradeoff between achieving a quick system response and minimizing
overshooting.
REFERENCE:
1. Kushal Gowda “Introduction to the control system 1.1”, Published Aug 19,
2020.
2. Classical PID Control
by Graham C. Goodwin, Stefan F. Graebe, Mario E. Salgado
Control System Design, Prentice Hall PTR
3. PID Control of Continuous Processes
by John W. Webb Ronald A. Reis
Programmable Logic Controllers, Fourth Edition, Prentice Hall PTR.
4. Norman S. NISE, “control system engineering” Book 6th edition.
5. Brian R Copeland, “The Design of PID Controllers using Ziegler Nichols
Tuning”, March 2008.
6. Ibrahim kaya, “I-PD Controller Design for Integrating Time Delay Processes
Based on Optimum Analytical Formulas”, IFAC-PapersOnLine, Volume 51,
Issue 4, 2018, Pages 575-580