Magnetic Levitation System

Modern Control Systems

Phase 1
Prof. Kebriaei
December 24, 2022

Shiva Shakeri

1
 Contents

1. Introduction 3

2. Magnetic Levitation System 4

3. State-Space and Input 5

4. Equilibrium Points 5

5. Linearization 6

6. State-Space Analysis 7

7. Transfer Function 10

8. PID Controller 11

9. Non-Linear Model Results in Simulink 17

Appendix A. 20

2
 1. Introduction
Magnetic levitation system is known as electro-mechanical systems in which an object is
levitate in speci c region without using any support. Now a days this technology is spreading
over a wide range because of its contact-less and friction-less properties as it removes energy
losses which occur due to friction. Since this technology was developed, researchers have
analyzed a large number of such systems. In the analysis of such systems most of the work has
been done on designing of controller as they require appropriate control action and are more
complex, such as State space controller, feedback linearization is used to control the position of
ball. However, there are mostly highly non-linear systems and open loop unstable. This
nonlinearity and unstability feature of magnetic levitation systems makes their control and
modeling very challenging.

The problem of controlling the magnetic eld by control method is taken up to levitate a metal
hollow sphere, here. The control problem is to supply controlled position of the ball such that the
magnetic force on the levitated body and gravitational force acting on it are exactly equal. Thus
the magnetic levitation system is inherently unstable without any control action.

The objective of the work can be directed to design controller considering the linear model as
well as nonlinear model. The controller designed is validated by simulating and implementing on
real time system.

3
fi
fi
 2. Magnetic Levitation System
2.1. Model Diagram
An overview of the general structure of a magnetic levitation system can be found in
Figure 1. Basically, an electromagnet and a metal ball make up the core of this system.

Figure 1. Model of Magnetic Levitation System

2.2. Model Parameters

The circuit’s resistance (R) and inductance (L) were measured with ohms meter and
inductance meter respectively. In this experiment, the object’s mass (M) is measured, placed
under the electromagnetic pole on a known distance (y) and current through the electromagnet
(I(t)) is gradually increased up to the point when it picks up the object. According to table 1, the
parameters which were obtained are as follows.

Table 1. MagLev System Model Parameters

R Circuit Resistance 50Ω

L Winding Inductance 0.2H

m
g Gravitational Constant 9.8 2
s
M Metal Ball Mass 0.491K g
N ×m
c Electromagnetic Force Constant 0.3
A2

4
 N ×s
fv Air Resistance(Friction) 0.04
m
y⋆ Output Operating Point 0.06m

3. State-Space and Input

In this system, there are three states, one controlling input, and one output. The
parameters x1 as the position of the metal ball, x 2 as the velocity of the metal ball, and x3 as
winding’s current are state parameters. The input is shown by parameter u, and presents the input
voltage. The input is shown by parameter 'u' which presents the input voltage. The 'y' parameter
shows the position of the metal ball as the output of this system.

As a result of writing the dynamic and physical equations for this system, we can nd the
state equations as follows.

x·1 = x 2
x 32 fv x 2
x·2 = − g + c
M 0.1 − x1
− M {y = [1,0,0]x
x·3 = L1 (−R x3 + u)

4. Equilibrium Points
At equilibrium points, the system doesn't change at all, and that means that the
parameters of the system should be equal to zero at that point. In order to nd the equilibrium
points of the system, we set the state equations to zero. Appendix A contains the Matlab code for
this part.

x2 = 0
c x 32 fv x 2
−g + M 0.1 − x1
− M
=0
1
L
(−R x3 + u) = 0

Then, by rewriting the equations using parameters:

5
fi
fi
 x2 = 0
0.3 x 32 0.04x 2
−9.8 + 0.491 0.1 − 0.06
− 0.491
=0
1
0.2
(−50x3 + u) = 0
By solving the equations, x1, x 2, and x3 are:

x1 = 0.06 x1 = 0.06
x2 = 0 ,and x2 = 0
x3 = − 0.800983 x3 = 0.800983

There are two values of x3, which show the current, meaning that the current can flow
clockwise or counterclockwise. The next section of the problem will be solved using the positive
value of the current flow.

5. Linearization
In order to determine a state-space equation for the system, the system has to be
linearized. Assume that x·1 = f (x1, x 2, x3), x·2 = g(x1, x 2, x3), and x·3 = h(x1, x 2, x3). Appendix A
contains the Matlab code for this part. The Jacobian Matrix is:

∂f ∂f ∂f
∂x1 ∂x 2 ∂x3 0 1 0
∂g ∂g ∂g 0.611x 32 1.222x3
A= ∂x1 ∂x 2 ∂x3
→A= −0.0814
(0.1 − x1) 2 −0.1 − x1
∂h ∂h ∂h 0 0 −250
∂x1 ∂x 2 ∂x3

In order to linearize the system, the equilibrium points take place in the Jacobian Matrix.

0 1 0 0 1 0
A1 = 245 −0.0814 −24.47 , A2 = 245 −0.0814 24.47
0 0 −250 0 0 −250

Therefore the linearized system is:

0 1 0 0 0 1 0 0
·x = 245 −0.0814 −24.47 x + 0 u, x· = 245 −0.0814 24.47 x + 0 u
1 1
0 0 −250 L
0 0 −250 L

In the next step, the eigenvalues of Jacobin Matrices are calculated.

6
 λ1 = − 250 λ1 = − 250
eig(A1) = λ 2 = − 15.7, eig(A2 ) = λ 2 = − 15.7
λ3 = 15.61 λ3 = 15.61

The eigenvalues of the linearized matrix are both real positive and negative values.
Therefore, the linearized system is unstable and the equilibrium points for the non-linear system
is also unstable.

6. State-Space Analysis
Controllability and observability of control systems are two of the major concepts of
modern control system theory. These concepts were introduced by R. Kalman in 1960. The
following are some de nitions that can be used to describe them. Appendix A contains the
Matlab code for this part.

- Controllability: In order to be able to do whatever we want with the given dynamic

system under control input, the system must be controllable.
- Observability: In order to see what is going on inside the system under observation, the
system must be observable.

6.1. Controllability
As a result of applying the theorem 1, we can determine whether the system is
controllable.

Theorem 1. A n-dimensions time-Invariant system,

x· (t) = A x (t) + Bu(t),
is totally controllable, if and only if the column vectors of the controllability matrix,
Φc = [B, A B, …, A n−1B],
span the n-dimensional space.

As a result, the controllability matrix of the system should be constructed as follows:

Φc = [B, A B, A 2 B] where the matrix B = [0,0,5]T.

0 1 0 0 0
A1B = 245 −0.0814 −24.47 0 = −122.35
0 0 −250 5 −1250
0 1 0 0 0
A2 B = 245 −0.0814 24.47 0 = 122.35
0 0 −250 5 −1250

7
fi
 245 −0.0814 −24.47 0 −122.35
A12 B = −19.943 245 6119.5 0 = 30597.45
0 0 62500 5 312500

245 −0.0814 24.47 0 122.35

A22 B = −19.943 245 −6119.5 0 = −30597.45
0 0 62500 5 312500
0 0 −122.35 0 0 122.35
→ Φc1 = 0 −122.35 30597.45 , Φc2 = 0 122.35 −30597.45
5 −1250 312500 5 −1250 312500

By calculating the determinant of the controllability matrices:

| Φc1 | = | Φc2 | = − 74847.61 ≠ 0

As a result of the theorem 1, the system is controllable if and only if | Φc | ≠ 0. Therefore

since the value of L = 0.2H, the system is controllable.

6.2. Observability
By using the theorem 2, one can determine the observability of a system.

Theorem 2. A n-dimensions time-Invariant system,

x· (t) = A x (t) + Bu(t),
is totally observable if and only if the column vectors of the observability matrix,
Φo = [C, CA, …, CA n−1],
span the n-dimensional space.

As a result, the controllability matrix of the system should be constructed as follows:

Φo = [C, CA, CA 2] where C = [1 0 0].

0 1 0
CA1 = [1 0 0] 245 −0.0814 −24.47 = [0 1 0]
0 0 −250
0 1 0
CA2 = [1 0 0] 245 −0.0814 24.47 = [0 1 0]
0 0 −250

245 −0.0814 −24.47

CA12 = [1 0 0] −19.943 245 6119.5 = [245 −0.0814 −24.47]
0 0 62500

8
 245 −0.0814 24.47
CA22 = [1 0 0] −19.943 245 6119.5 = [245 −0.0814 24.47]
0 0 62500

1 0 0 1 0 0
→ Φo1 = 0 1 0 , Φo2 = 0 1 0
245 −0.0814 −24.47 245 −0.0814 24.47

The observability matrix has the rank 3 and its vectors are linearly independent. Therefore
the matrix Φo spans the n-dimensional space and the system is always observable.

6.3. Minimal Realization

The following theorem which is due to Kalman gives a characterization of minimal state
space realizations:

Theorem 3. A realization (A,B,C,D) is minimal if and only if it is controllable and

observable.

As a result of this theorem, the presented state-space is minimal.

6.4. State Transition Matrix

The state-transition matrix is a matrix whose product with the state vector x at an initial
time t0 gives x at a later time t. The state-transition matrix can be used to obtain the general
solution of linear dynamical systems.
First, by solving the homogeneous state equations using u(t) = 0: x· (t) = A x (t). The final
solution of this equation is x (t) = exp[A(t − t0 )]x (t0 ); where t0 is the initial time and x (t0 ) is the
initial state vector. The exp[A(t − t0 )] can be written as follows:

A(t − t0 ) A 2(t − t0 )2 A k (t − t0 )k
e A(t−t0 ) = exp[A(t − t0 )] = I + + +…+
1! 2! k!

Therefore, the matrix exp[A(t − t0 )] is squared and its dimensions equals to the dimensions
of matrix A. This matrix is called “state-transition matrix,” and is shown by:

Φ(t) = e At = exp[At]

The state transition matrix can also be calculated by this formula:

e At = ℒ−1[(sI − A)−1]

9
 By applying this formula to the system:

−1
s −1 0
e A1t −1
= ℒ ( −245 s + 0.0814 24.47 )
0 0 s + 250

By using the Sylvester interpolation method to calculate e At:

0 1 0 λ1 = − 250
A1 = 245 −0.0814 −24.47 , eig(A1) = λ 2 = − 15.7
0 0 −250 λ3 = 15.61

1 λ1 λ12 e λ1t 1 −250 62500 e −250t

1 λ 2 λ 22 e λ 2 t 1 −15.7 246.49 e −15.7t
=0→ =0
1 λ3 λ32 e λ3t 1 15.61 243.67 e 15.61t
I A A 2 e At I A A2 e At

1948496.6721e (At) + (265.61A 2 e −15.7t − 31.31A 2 e −250t − 234.3A 2 e 15.61t + 62256.33Ae −15.7t − 2.82Ae −250t−
(62253.51A e 15.61t − 1036542.5Ie −15.7t + 76733.279Ie −250t − 919627.5Ie 15.61t = 0

This equation can be solved for e At.

7. Transfer Function
The transfer function of an linearized system x· = A x + Bu, and y = Cx is

g(s) = c(sI − A)−1b

−1
s −1 0 0 −60073.7
g1(s) = [1 0 0] −245 s + 0.0814 24.47 0 =
(491s 3 + 122790s 2 − 110295s − 30073750)
0 0 s + 250 5

−1
s −1 0 0 60073.7
g1(s) = [1 0 0] −245 s + 0.0814 −24.47 0 =
(491s 3 + 122790s 2 − 110295s − 30073750)
0 0 s + 250 5

The system has no zeros, and the poles of the both transfer functions are −250, −15.69,
and 15.69. Appendix A contains the Matlab code for this part.

10
 8. PID Controller
8.1. Introduction
A proportional–integral–derivative controller (PID controller or three-term controller) is a
control loop mechanism employing feedback that is widely used in industrial control systems
and a variety of other applications requiring continuously modulated control. A PID controller
continuously calculates an error value as the difference between a desired setpoint (SP) and a
measured process variable (PV) and applies a correction based on proportional, integral, and
derivative terms (denoted P, I, and D respectively), hence the name.

Figure 2. A Block Diagram of a PID Controller

In practical terms, PID automatically applies an accurate and responsive correction to a

control function. An everyday example is the cruise control on a car, where ascending a hill
would lower speed if constant engine power were applied. The controller's PID algorithm
restores the measured speed to the desired speed with minimal delay and overshoot by increasing
the power output of the engine in a controlled manner.
The distinguishing feature of the PID controller is the ability to use the three control terms
of proportional, integral and derivative influence on the controller output to apply accurate and
optimal control. The block diagram in Fig. 2 shows the principles of how these terms are
generated and applied.

8.2. PID Controller Tuning

First the open-loop(Fig. 3) , using the transfer function g(s), where

11
 Y(s) −60073.7
g1(s) = =
X(s) (491s + 122790s 2 − 110295s − 30073750)
3

Y(s) 60073.7
g2(s) = =
X(s) (491s 3 + 122790s 2 − 110295s − 30073750)

The step responses of two systems (g1 and g2) is shown in Fig. 4 and Fig. 5 respectively.

Figure 3. Open-Loop System Diagram

Figure 4. Step Response of G1 Figure 5. Step Response of G2

8.2.1. Proportional Control

The proportional controller (Kp) reduces the rise time, increases the overshoot, and reduces
the steady-state error. The closed-loop transfer function of our unity-feedback system with a
proportional controller is the following:

12
 KpG (s) 60037.7 + Kp
T (s) = =
1 + KpG (s) (491s 3 + 122790s 2 − 110295s + (−30073750 + Kp)

Figure 6. Proportional Controller Diagram

The step response of this system when Kp = 500 is shown in Fig 7 and Fig 8.

Figure 7. Step Response of KpG1 Figure 8. Step Response of KpG2

The step response of this system when Kp = 1600 for G2 is shown in Fig 9.
In Fig. 9 it is shown that the range of step response’s amplitude is more bounded by
choosing the right proportional constant, but as the above figures show, the system is still
unstable.

13
 Figure 9. Step Response of
(Kp = 1600)G2

8.2.2. Proportional-Derivative Control

Now, let's take a look at PD control. The addition of derivative control (Kd) tends to reduce
both the overshoot and the settling time. The closed loop transfer function of the given system
with a PD controller is:

(Kd s + Kp)G (s) Kd s + (60073.7 + Kp)

T3(s) = =
1 + (Kd s + Kp)G (s) (491s 3 + 122790s 2 + (−11029 + Kd )s + (−30073750 + Kp)

The step response of this system when Kp = 1600 and Kd = 500 for G2 is shown in Fig. 10.
The stepinfo() in this system will be in table 2.

Table 2. Step Info T_3

RiseTime 0.5010

SettlingTime 1.2261

SettlingMin 1.3100

SettlingMax 1.4549

Overshoot 0

Undershoot 0

Peak 1.4549

PeakTime 3.0464

14
 Figure 10. Step Response of T3

8.2.4. Proportional-Integral-Derivative Control

Now, let's examine PID control. The closed-loop transfer function of the given system with
a PID controller is:

(Kd s + Kp)G (s) Kd s + (60073.7 + Kp)

T4(s) = =
1 + (Kd s + Kp)G (s) (491s 3 + 122790s 2 + (−11029 + Kd )s + (−30073750 + Kp)

By setting the PID constants Kp = 1600, Ki = 2600, and Kd = 245, the step response of the
system is shown in Fig. 11.
The stepinfo() in this system will be in table 3.

Table 3. Step Info T_4

RiseTime 0.0117

SettlingTime 1.1550

SettlingMin 0.9167

SettlingMax 1.3107

Overshoot 31.0718

Undershoot 0

Peak 1.3107

PeakTime 0.3212

15
 Figure 11. Step Response of T4

Rising time, settling time, and steady-state error are small, but there exists a large value of
overshoot.

16
 9. Non-Linear Model Results in Simulink
The non-linear model of the Magnetic Levitation System is designed in Simulink and it’s
shown in Fig. 12. The parameters are assigned by the values in Table. 1.

Figure 12. Non-Linear Model of the Magnetic Levitation System Diagram

The completed system is shown in Fig. 13. The input is considered as step and the 3D
simulation of ball transition in the system is in the output to show better vision of this system.
First, the ball’s position can be seen in Fig. 13, where the ball’s position is 0 based on the
definition of output in section 2.

Figure 13. Magnetic Levitation System Diagram

17
 Figure 14. Ball’s Position T=0.00 Figure 17. Ball’s Position T=10.00

In the next step the PID controller is added to the system(Fig. 15.) The PID block in Matlab
is added and is tuned as Fig. 16. By adding the PID controller to the system, the ball moves
upward and stays stable at the position shown in Fig. 17.

Figure 15. Magnetic Levitation System Diagram and PID Controller

Figure 16. PID Controller Parameters

18
 Next, the step responses of position, velocity, and current(states) are shown in Fig. 18, Fig.
19, and Fig. 20 respectively.

Figure 18. Ball’s Position Step Response Figure 19. Ball’s Speed Step Response

Figure 20. Winding’s Current Step Response

In section 4, the equilibrium points were calculated. The stable values of x1 as the ball’s
position was 0.06, x 2 was 0, and x3 as the winding’s current was 0.8009. The above figures
confirm that the system with the designed PID controller will have those values when it’s stable.

19
 Appendix A.
Matlab codes of project sections exist in this appendix.

1. Equilibrium Points (4)

syms x1 x2 x3 u

R = 50;
L = 0.2;
g = 9.8;
M = 0.491;
c = 0.3;
fv = 0.04;

eq1 = x2 == 0;
eq2 = x1 == 0.06;
eq3 = -g+(c/M)*(x3^2/(0.1-x1))-(fv*x2/M) == 0;
eq4 = (1/L)*(-R*x3+u) == 0;

stateEqs = [eq1, eq2, eq3, eq4];

sol = solve(stateEqs);

sol.x1
sol.x2
sol.x3

ans =

3/50
3/50

ans =

0
0

ans =

-(7*7365^(1/2))/750
(7*7365^(1/2))/750

20
 2. Linearization (5)
f = x2;
g = -g+(c/M)*(x3^2/(0.1-x1))-(fv*x2/M);
h = (1/L)*(-R*x3+u);

A = jacobian([f, g, h], [x1, x2, x3]);

A

A_1 = subs(A, {x1, x2, x3}, {xe1(1), xe2(1),xe3(1)})

A_2 = subs(A, {x1, x2, x3}, {xe1(2), xe2(2),xe3(2)})

eigA1 = eig(A_1)
eigA2 = eig(A_2)

A=

[ 0, 1, 0]
[(300*x3^2)/(491*(x1 - 1/10)^2), -40/491, -(600*x3)/(491*(x1 - 1/10))]
[ 0, 0, -250]

A_1 =

[ 0, 1, 0]
[245, -40/491, -(140*7365^(1/2))/491]
[ 0, 0, -250]

A_2 =

[ 0, 1, 0]
[245, -40/491, (140*7365^(1/2))/491]
[ 0, 0, -250]

eigA1 =

-250
- (3*6562805^(1/2))/491 - 20/491
(3*6562805^(1/2))/491 - 20/491

eigA2 =

-250
- (3*6562805^(1/2))/491 - 20/491
(3*6562805^(1/2))/491 - 20/491

21
 3. State-Space Analysis (6)

% Controllability
% We can also use the command "ctrb(A, B)"
B = [0; 0; 5];

A_1B = A_1 * B;
A_2B = A_2 * B;

A_12B = A_1^2 * B;
A_22B = A_2^2 * B;

Phi_c1 = [B A_1B A_12B]

Phi_c2 = [B A_2B A_22B]

det_Phi_c1 = det(Phi_c1)
det_Phi_c2 = det(Phi_c2)

% Observability
% We can also use the command "obsv(A, C)"
C = [1 0 0];

A_1C = C * A_1;
A_2C = C * A_2;

A_12C = C * A_1^2;
A_22C = C * A_2^2;

Phi_o1 = [C; A_1C; A_12C]

Phi_o2 = [C; A_2C; A_22C]

rank_Phi_o1 = rank(Phi_o1)
rank_Phi_o2 = rank(Phi_o2)

% State Transition Matrix

syms s

eA1t = ilaplace(inv((s*eye(size(A_1,1))-A_1)))
eA2t = ilaplace(inv((s*eye(size(A_2,1))-A_2)))

22
Phi_c1 =

[0, 0, -(700*7365^(1/2))/491]
[0, -(700*7365^(1/2))/491, (85953000*7365^(1/2))/241081]
[5, -1250, 312500]

Phi_c2 =

[0, 0, (700*7365^(1/2))/491]
[0, (700*7365^(1/2))/491, -(85953000*7365^(1/2))/241081]
[5, -1250, 312500]

det_Phi_c1 =

-36750000/491

det_Phi_c2 =

-36750000/491

Phi_o1 =

[ 1, 0, 0]
[ 0, 1, 0]
[245, -40/491, -(140*7365^(1/2))/491]

Phi_o2 =

[ 1, 0, 0]
[ 0, 1, 0]
[245, -40/491, (140*7365^(1/2))/491]

rank_Phi_o1 =

rank_Phi_o2 =

23
eA1t =

[exp(-(20*t)/491)*(cosh((3*6562805^(1/2)*t)/491) + (4*6562805^(1/2)*sinh((3*6562805^(1/2)*t)/
491))/3937683), (491*6562805^(1/2)*exp(-(20*t)/491)*sinh((3*6562805^(1/2)*t)/
491))/19688415, (4*7365^(1/2)*exp(-(20*t)/491)*(cosh((3*6562805^(1/2)*t)/491) -
(8182*6562805^(1/2)*sinh((3*6562805^(1/2)*t)/491))/1312561))/873063 - (4*7365^(1/2)*exp(-250*t))/
873063]
[ (24059*6562805^(1/2)*exp(-(20*t)/491)*sinh((3*6562805^(1/2)*t)/491))/3937683,
exp(-(20*t)/491)*(cosh((3*6562805^(1/2)*t)/491) - (4*6562805^(1/2)*sinh((3*6562805^(1/2)*t)/491))/
3937683), (1000*7365^(1/2)*exp(-250*t))/873063 - (1000*7365^(1/2)*exp(-(20*t)/
491)*(cosh((3*6562805^(1/2)*t)/491) - (8353*6562805^(1/2)*sinh((3*6562805^(1/2)*t)/491))/
328140250))/873063]
[ 0,
0,
exp(-250*t)]

eA2t =

[exp(-(20*t)/491)*(cosh((3*6562805^(1/2)*t)/491) + (4*6562805^(1/2)*sinh((3*6562805^(1/2)*t)/
491))/3937683), (491*6562805^(1/2)*exp(-(20*t)/491)*sinh((3*6562805^(1/2)*t)/
491))/19688415, (4*7365^(1/2)*exp(-250*t))/873063 - (4*7365^(1/2)*exp(-(20*t)/
491)*(cosh((3*6562805^(1/2)*t)/491) - (8182*6562805^(1/2)*sinh((3*6562805^(1/2)*t)/491))/
1312561))/873063]
[ (24059*6562805^(1/2)*exp(-(20*t)/491)*sinh((3*6562805^(1/2)*t)/491))/3937683,
exp(-(20*t)/491)*(cosh((3*6562805^(1/2)*t)/491) - (4*6562805^(1/2)*sinh((3*6562805^(1/2)*t)/491))/
3937683), (1000*7365^(1/2)*exp(-(20*t)/491)*(cosh((3*6562805^(1/2)*t)/491) -
(8353*6562805^(1/2)*sinh((3*6562805^(1/2)*t)/491))/328140250))/873063 -
(1000*7365^(1/2)*exp(-250*t))/873063]
[ 0,
0,
exp(-250*t)]

24
 4. Transfer Function (7)

g1 = C * (inv((s*eye(size(A_1,1))-A_1))) * B
g2 = C * (inv((s*eye(size(A_2,1))-A_2))) * B

poleq = 491*s^3 + 122790*s^2 - 110295*s - 30073750;

solve(poleq, s)

g1 =

(700*7365^(1/2))/(- 491*s^3 - 122790*s^2 + 110295*s + 30073750)

g2 =

-(700*7365^(1/2))/(- 491*s^3 - 122790*s^2 + 110295*s + 30073750)

ans =

-250
- (3*6562805^(1/2))/491 - 20/491
(3*6562805^(1/2))/491 - 20/491

25
 5. PID Controller

G1 = -60073.7 / (491*s^3 + 122790*s^2 - 110295*s - 30073750);

G2 = 60073.7 / (491*s^3 + 122790*s^2 - 110295*s - 30073750);

t = 0:0.01:2;
figure;
step(G1, t)
title('Step Response Open Loop G1')

figure;
step(G2, t)
title('Step Response Open Loop G2')
% P
Kp = 1600;
C = pid(Kp);

T1 = feedback(C*G1,1);
T2 = feedback(C*G2,1);

figure;
step(T1,t)
title('Step Response Closed Loop Kp.G1')
figure;
step(T2,t)
title('Step Response Closed Loop Kp.G2')
%PD
Kp = 1600;
Kd = 500;

C = pid(Kp,0,Kd);
T3 = feedback(C*G2,1);
figure;
stepinfo(T3)
step(T3,t)
title('Step Response Closed Loop Kd.Kp.G2')
%PID
Kp = 1600;
Ki = 2600;
Kd = 245;

C = pid(Kp,Ki,Kd);
T4 = feedback(C*G2,1);
figure;
stepinfo(T4)
step(T4,t)
title('Step Response Closed Loop Kd.Ki.Kp.G2')

26