Feedback Control System

Lecture 21
Chapter 9: Design via Root Locus

Prepared By: Muhammad Abdullah

[email protected] , [email protected]

National University FAST (CFD Campus)

My YouTube Channel:
https://www.youtube.com/@BlueFish506

1|Page
PD Controller Design:

Consider the following unity feedback system;

Figure 1

Where;
𝑠+2
𝐺(𝑠) = ; 𝐶(𝑠) = 𝐾(𝑠 + 𝑧1 )
(𝑠 2 + 𝑠 + 7.5)(𝑠 + 5)
Design PD controller using root locus method to meet the following design
specifications;

1. Percentage Overshoot: P.O. = 4.32%

2. Settling Time: 𝑇𝑠 = 0.8 𝑠𝑒𝑐

Solution:

Method-1: Graphical Method for Calculations

Step-1: Calculating the desired dominant poles

Required damping ratio is;
𝑃. 𝑂.
−𝑙𝑛 (
𝜁= 100 ) =
−𝑙𝑛(0.0432)
= 0.707
2 2
√𝜋 + [𝑙𝑛(0.0432)] 2
√𝜋 2 + [𝑙𝑛 (𝑃. 𝑂.)]
100
Required natural frequency required is;

2|Page
 4 4 4
𝑇𝑠 =  𝜔𝑛 = = = 7.07 𝑟𝑎𝑑/𝑠𝑒𝑐
𝜁𝜔𝑛 𝑇𝑠 𝜁 0.8 × 0.707
Close loop characteristic equation of the system is;
𝑠 2 + 2𝜁𝜔𝑛 𝑠 + 𝜔𝑛 2 = 0
𝑠 2 + 2 × 0.707 × 7.07𝑠 + (7.07)2 = 0
𝑠 2 + 10𝑠 + 50 = 0
Desired close loop dominant poles of the system are;
𝑠𝑑 = −5 ± 5𝑗

Step-2: Calculate Required Angle Contribution from zero of PD Controller

Pole zero map is shown below;

Figure 2

Angle contribution made by each pole and zero at -5+5j location is calculated as;
5 − 2.69 2.31
𝜃1 = 180° − 𝑡𝑎𝑛−1 ( ) = 180° − 𝑡𝑎𝑛−1 ( ) = 152.9°
5 − 0.5 4.5

3|Page
 5 + 2.69 7.69
𝜃2 = 180° − 𝑡𝑎𝑛−1 ( ) = 180° − 𝑡𝑎𝑛−1 ( ) = 120.3°
5 − 0.5 4.5
𝜃3 = 90°
5−0 5
𝜃4 = 180° − 𝑡𝑎𝑛−1 ( ) = 180° − 𝑡𝑎𝑛−1 ( ) = 121°
5−2 3
Now, using angle condition of root locus is given as;

∑(angles from zeros) − ∑(angles from poles) = (2𝑘 + 1)180∘

(𝜃𝑧1 + 𝜃4 ) − (𝜃1 + 𝜃2 + 𝜃3 ) = −180°

(𝜃𝑧1 + 121°) − (152.9° + 120.3° + 90°) = −180°
(𝜃𝑧1 + 121°) − (363.2°) = −180°
𝜃𝑧1 = −180° + 363.2° − 121° = 62.2°

Step-3: Calculate Location of zero

PD zero location is calculated as;

Figure 3

4|Page
 5
= 𝑡𝑎𝑛(62.2°)
𝑧1 − 5
5 5
𝑧1 = +5= + 5 = 7.6
𝑡𝑎𝑛(62.2°) 1.9

Step-4: Calculate Gain K using Magnitude Criteria

Length of poles and zeros is calculated as;
𝑑1 = √(5 − 0.5)2 + (5 − 2.69)2 = 5.1

𝑑2 = √(5 − 0.5)2 + (5 + 2.69)2 = 8.9

𝑑3 = √(5 − 5)2 + (5 − 0)2 = 5

𝑑4 = √(5 − 0) + (5 − 2)2 = 5.8

𝑑𝑧1 = √(7.64 − 5)2 + (5)2 = 5.7

Using magnitude condition of root locus;
𝑑1 𝑑2 𝑑3 5.1 × 8.9 × 5
𝐾= = = 6.84
𝑑4 𝑑𝑧1 5.8 × 5.7
Transfer function of PD controller is given as;
𝐶(𝑠) = 𝐾(𝑠 + 𝑧1 ) = 6.84(𝑠 + 7.6)

Verification using MATLAB:

Root locus plot with above designed PD controller is given as;

5|Page
 Figure 4

Above root locus plot verifies that root locus plot passes through −5 ± 5𝑗 point
approximately at gain of 𝐾 = 6.84

Method-2: Analytical Method for Calculations

It is given that:
𝑠+2
𝐺(𝑠) =
(𝑠 2 + 𝑠 + 7.5)(𝑠 + 5)
Put 𝑠 = 𝑠𝑑 = −5 + 5𝑗
−5 + 5𝑗 + 2
𝐺(𝑠𝑑 ) =
((−5 + 5𝑗)2 − 5 + 5𝑗 + 7.5)(−5 + 5𝑗 + 5)
−3 + 5𝑗 −3 + 5𝑗
𝐺(𝑠𝑑 ) = =
(−50𝑗 + 2.5 + 5𝑗)(5𝑗) 225 + 12.5𝑗
𝐺(𝑠𝑑 ) = −0.0121 + 0.0229𝑗 = 0.0259∠117.784°
From above calculations, it is seen that; ∠𝐺(𝑠𝑑 ) = 117.784°
Now using phase condition, angle contribution of PD zero can be calculated as;
𝜃𝑧1 + ∠𝐺(𝑠𝑑 ) = 180°

6|Page
𝜃𝑧1 + 117.784° = 180°
𝜃𝑧1 = 180° − 117.784° = 62.216°
Now location PD controller zero location is calculated as;
5
= 𝑡𝑎𝑛(62.216°)
𝑧1 − 5
5
𝑧1 = + 5 = 7.632
𝑡𝑎𝑛(62.216°)
From above presented calculations, it is seen that;
𝐺(𝑠𝑑 ) = 0.0259∠117.784°
Length of zero w.r.t 𝑠𝑑 point is calculated as;

𝐿𝑧1 = √(7.632 − 5)2 + (5)2 = 5.652

Using magnitude condition, we can write;
1 1
𝐾= = = 6.8312
𝐿𝑧1 |𝐺(𝑠𝑑 )| 5.652 × 0.0259

PID Controller Design:

Now design PID controller for same above system assuming that we still have
desired close loop dominant poles given as: 𝑠𝑑 = −5 ± 5𝑗
Transfer function of PID controller is given as;
𝐾(𝑠 + 𝑧1 )(𝑠 + 𝑧2 )
𝐶(𝑠) =
𝑠
For the sake of simplicity, you can assume 𝑧2 = 2.
Solution:
Given PID controller has zero at 𝑠 = −2 and a pole at 𝑠 = 0. So, we need to add the
effect of this zero and pole in already calculated angles and gain in PD controller
design example, to find new zero position.
Let angle of 𝑧1 calculated above in PD controller design example is represented by;

7|Page
 𝜃𝑧1𝑜𝑙𝑑 = 62.2°
Let gain of PD controller calculated is represented by;
𝑘𝑜𝑙𝑑 = 6.84

Figure 5

Angle of newly added pole at origin is;

5
𝜃0 = 180° − 𝑡𝑎𝑛−1 ( ) = 135°
5
Length of newly added pole at origin is;
𝑑0 = √(5)2 + (5)2 = 7.07
Let angle of 𝑧1 for PID controller is 𝜃𝑧1𝑛𝑒𝑤 , then we can write;
𝜃𝑧1𝑛𝑒𝑤 = 𝜃𝑧1𝑜𝑙𝑑 − 𝜃4 + 𝜃0 = 62.2° − 121° + 135° = 76.2°
Now POD zero location 𝑧1 is calculated as;
5
= 𝑡𝑎𝑛(𝜃𝑧1𝑛𝑒𝑤 ) = 𝑡𝑎𝑛(76.2°)
𝑧1 − 5
5 5
𝑧1 = +5= + 5 = 6.2
𝑡𝑎𝑛(76.2°) 4.07

8|Page
Let the gain of PID controller is 𝐾, then we can write;
𝑑0 7.07
𝐾= 𝑘𝑜𝑙𝑑 = × 6.84 = 8.34
𝑑4 5.8
Root locus plot for above calculated values for PID compensated system is shown
below;

Figure 6

Above root locus plot verifies that root locus plot passes through -5+5j point
approximately at gain of K = 8.34.
Transfer function of PID controller can be written as;
𝐾(𝑠 + 𝑧1 )(𝑠 + 𝑧2 ) 8.34(𝑠 + 6.2)(𝑠 + 2)
𝐶(𝑠) = =
𝑠 𝑠

9|Page
Lead Controller Design:

Transfer function of the mechanical system is given as;

500
𝐺(𝑠) =
𝑠(𝑠 + 9.7)(𝑠 + 51.2)
Use the root locus method to design a lead controller that achieves a closed-loop step
response with P.O. ≤ 3.0%, and a settling time 𝑇𝑠 ≤ 0.22𝑠 (using the 2% criterion).
In addition, the absolute steady-state error when tracking a unit ramp input should
be 𝑒𝑠𝑠 ≤ 0.11.

Solution:
As 𝑃. 𝑂. ≤ 3% and 𝑇𝑠 ≤ 0.22 sec, for the sake of design calculations we let 𝑃. 𝑂. =
2.5% and 𝑇𝑠 = 0.20 sec.
𝑃. 𝑂.
−𝑙𝑛 ( ) −𝑙𝑛(0.025)
𝜁= 100 = = 0.76
2
√𝜋 2 + [𝑙𝑛 (𝑃. 𝑂. √𝜋 2 + (𝑙𝑛(0.025))
2
100 )]
4 4
𝜔𝑛 = = = 26.2 𝑟𝑎𝑑/𝑠𝑒𝑐
𝑇𝑠 𝜁 0.2 × 0.76
Close loop characteristic equation of the system is;
𝑠 2 + 2𝜁𝜔𝑛 𝑠 + 𝜔𝑛 2 = 0
𝑠 2 + 2 × 0.76 × 26.2𝑠 + (26.2)2 = 0
𝑠 2 + 39.82𝑠 + 684.44 = 0
Desired close loop dominant poles of the system are;
𝑠𝑑 = −20 ± 17𝑗
Dominant pole locations and root locus plots are shown below;

10 | P a g e
 Figure 7

As stated above lead compensator is selected to meet the desired specifications;

𝑠+𝑧
𝐶(𝑠) = 𝐾 𝑤ℎ𝑒𝑟𝑒 𝑝 > 𝑧
𝑠+𝑝
Assuming 𝑧 = 8.97, with the application of the angle condition gives;
𝜃𝑑 = 𝜙 + 𝜃𝑧 − 180°
Where 𝜙 angle of the plant at 𝑠 = 𝑠𝑑 and angle of 𝑧 at 𝑠 = 𝑠𝑑 .
𝜃𝑑 = 70.6° + 122.9° − 180° = 13.5°
The value of the lead compensator pole 𝑝 is calculated as:
𝑏 17
𝑝= +𝑎 = + 20 = 90.6
𝑡𝑎𝑛𝜃𝑑 𝑡𝑎𝑛(13.5°)
𝑠+𝑝 𝑠 + 90.6 𝑠(𝑠 + 9.7)(𝑠 + 51.2)
𝐾=| | =| | = 133.2
(𝑠 + 𝑧)𝐺(𝑠) 𝑠=𝑠 𝑠 + 8.97 500
𝑑 𝑠=−20+17𝑖

Final transfer function of lead compensator is;

11 | P a g e
 𝐾(𝑠 + 𝑧) 133.2(𝑠 + 8.97)
𝐶(𝑠) = =
𝑠 + 𝑝 𝑠 + 90.6
Close loop step response and ramp response with above lead compensator is plotted
as;

Figure 8

12 | P a g e