Feedback Control Systems

Dr. Basil Hamed

Electrical & Computer Engineering
Islamic University of Gaza
STABILITY OF LINEAR FEEDBACK
 PROBLEM DEFINITION
For people paralyzed from the neck down, the ability
to drive themselves around in motorized wheelchairs
is highly desirable.   A proposed system uses velocity
sensors mounted in the headgear at 900 intervals, so
that forward, left, right, or reverse directions can be
commanded.   Output of the headgear sensor is
proportional to the magnitude of the head movements.
  The block diagram for this system is shown in figure
1.   Here, typical values for the time constants are 1 =
0.5 s, 3 = 1 s, and 4 = 1/4 s.
Block Diagram
 Using MATLAB do the following
1) Determine the limiting gain K = K1K2K3 for a stable
system.
2) When the gain K is set equal to 1/3 of the limiting
value, determine if the settling time to within 2% of
the final value of the system is less than 4 s.
3) Determine the value of gain that results in a system
with a settling time of 4 s.  Also, obtain the value of the
roots of the characteristic equation when the settling
time is equal to 4 s.
Part 1)
 Routh-Hurwitz Table
Routh-Hurwitz Table:
s3          1          14         0
s2          7        8+8K       0
s1          A          0          0
s0        8+8K       0          0
A = -(8+8K-98) = 98-8-8K
              7               7
For stability, A > 0, therefore 0 < 98 - 8 - 8K
Therefore range for stability is given by:
0 < K < 11.25
 Part 2)
the closed-loop transfer function was found to be:

                  T(s) =                 8K                 
                            s3 + 7s2 + 14s + 8 + 8K.
In calculating the settling time, we assume the validity of a
second order approximation, allowing the use of the
dominant pole pair to find settling time as:

                  Ts =    4   
                          ζwn
where ζwn = σd = -1 * real part of dominant poles.
.
 Part 2)
K = 3.7333
Open-loop system Transfer function:
3.733
---------------------------------------------------
0.125 s^3 + 0.875 s^2 + 1.75 s + 1
Closed-loop system Transfer function:
3.733
----------------------------------------------------------
0.125 s^3 + 0.875 s^2 + 1.75 s + 4.733
P=
         -5.70955929441205
        -0.645220352793972 +     2.4931592351599i
        -0.645220352793972  -     2.4931592351599i
settling_time =   6.19943246160627
Part 2)
 Part 3)

For the system to be practical, a settling time of 4

seconds is required. Making use of the second order
equation, settling time Ts = 4 / ζwn where ζwn = - real
part of the dominant closed-loop pole pair, the settling
time is calculated for each K value from 0.1 to the
limiting gain or until one yielding a result of 4 seconds
is found. The step response, transfer functions and the
roots of the characteristic equation are displayed for
this value of K.
 Part 3)
For K = 1.5 settling time is approximated as 4 seconds.
Open-loop transfer function with K = 1.5
Transfer function:
               1.5
----------------------------------------------------
0.125 s^3 + 0.875 s^2 + 1.75 s + 1
Closed-loop transfer function with K = 1.5
Transfer function:
               1.5
------------------------------------------------------
0.125 s^3 + 0.875 s^2 + 1.75 s + 2.5
Roots of the characteristic equation for K = 1.5
P=
              -4.99999999999999
              -1 +      1.73205080756888i
              -1  -      1.73205080756888i
Settling time for K = 1.5
settling_time =   3.9999
Part 3)
Animation
Problem 2
 PROBLEM DEFINITION

The goal of vertical takeoff and landing (VTOL)

aircraft is to achieve operation from relatively
small airport and yet operate as normal aircraft
in level flight. An aircraft taking off in a form
similar to a missile (on end) is inherently
unstable. A control system using adjustable jets
can control the vehicle
Block Diagram
 Use MATLAB
a) Find and plot closed loop poles in s-plane and
discuss their location for K=100.
b) Determine the range of gain K for which the system
is stable, marginally stable and unstable.
c) Determine and plot the roots of the characteristic
equation for gain K obtained in part "b", which
makes the system to be marginally stable and for
selected gain K that makes the system unstable
including poles locations from part "a" giving full
comment.
d) Plot step responses of the system for K=100, selected
system gain, which makes the system to be unstable
and the obtained gain K in part "b" which makes
the system to be marginally stable, giving comments
on the obtained results.
 Part a)

MATLAB results for K=100:

  P1= -3.1269+7.9403i    ;    zeros: z1= -2
  P2= -3.1269-7.9403i
  P3= -2.7463
Part a)
 Part b)
• K<12.857143 - unstable
• K=12.875143 - marginally stable
• K>12.875143 - stable
 Part C)
The roots for selected K=2 (which make the
system unstable) are:
P1=-9.8531
P2=0.4266+0.4733i
P3=0.4266-0.4733i
Part C)
 Part C)
The roots for K=12.857143 (from part b) are:
P1=-9
P2=0+1.6903i
P3=0-1.6903i
zeros: z1=-2
Part C)
 Part D)
Output step responses for K=100, K=12.8571 & K=2
Output step response of the aircraft control system
Output step response for K=100
Output step response for K=12.85714
Output step response for K=2
Problem 3
Consider the potential for a robot steering a
motorcycle.   The block diagram of the system
model is shown in Figure 1.   Determine the
range of K for stable operation of the
motorcycle when α1=g/h=9, α2=V2/hc=2.7, and
α3=V/hc=1.35 where g is the gravity, c is the
distance between the wheels of the motorcycle
and h is the height of the centre of gravity. We
assume the motorcycle is moving with a
constant velocity V=2m/s.   The time constant
of the controller is =0.2 s, and K>0.
Figure 1
 Part a

Using MATLAB do the following:

1) plot the Step Response for the Physical
System Dynamics only

2) plot the Root Locus and Step Response

for the Closed-Loop system with system
gain (K) is 1
Figure 2
 Part b
For α3=V/hc=3, using MATLAB do the
following:
1)plot the Root Locus and Step Response for
the Closed-Loop system with system gain,
K=0.1 and 57 degree (1 rad) setpoint

2) plot the Root Locus and Step Response for

this system with angle equal to 0 radian (0
degree setpoint)
 Part a 1)
In this part, the step response of just the
Physical System Dynamics is to be
displayed. From figure 1 of the problem
definition, the Physical system is given by
the transfer function:

     1      =      1        for α1 = 9.

 s2 - α1       s2 - 9
Transfer function:  1/s^2 - 9
 
 
 Part a 2)
The system defined by the block diagram in
figure 1 of the problem, has non-unity feedback.
Therefore the root locus is found using the
product of G1(s) and H(s), and the Matlab
function rlocus(GH). Here, G1(s) is the product
of the controller and dynamics transfer function.
The step response can then be found for the
closed loop transfer function T(s) =
feedback(G1(s),H(s)) with gain K=1.
 Part a 2)   RESULTS
• Controller Transfer function:
1.35 s + 2.7
  0.2 s + 1
• Dynamics Transfer function:
     1     
 s^2 - 9
• Closed-Loop System Transfer function:
                1.35 s + 2.7                
 1.55 s^3 + 6.4 s^2 + 4.95 s - 6.3
 Part b 1)

The system defined by the block diagram in

figure 2 of the problem, has non-unity
feedback. Therefore the root locus is found
using the product of G1(s) and H(s), and the
Matlab function rlocus(GH). Here, G1(s) is
the product of the controller and dynamics
transfer function. The step response can then
be found for the closed loop transfer function
T(s) = feedback(G1(s),H(s)) with gain K=0.1.
Controller Transfer function:
3 s^2 - 6 s - 9
   0.2 s + 1

Dynamics Transfer function:

     1     
 s^2 – 9

Closed-Loop System Transfer function:

               3 s^2 - 6 s - 9                        
 0.3 s^4 + 2.6 s^3 - 5.6 s^2 - 11.4 s - 9.9
 Part b 2)

For the system described by figure 2 of the

problem, the root locus and step response are
found for an angle of 0 radians (0 degrees).
This example corresponds to vertical travel
of the motorbike. The system gain is again
0.1, but this time the input is 0 degrees.
Controller Transfer function:
 3 s^2 - 6 s - 9 
     0.2 s + 1

Dynamics Transfer function:

     1     
 s^2 – 9

Closed-Loop System Transfer function:

                   3 s^2 - 6 s - 9                    
 0.3 s^4 + 2.6 s^3 - 5.6 s^2 - 11.4 s - 9.9