Lecture 3:

Analytical Solution
of Linear Systems
Dominik Bongartz

Systems Analysis of Chemical Processes – H06T4A

 Linear Time-Invariant (LTI) Systems

Non-autonomous Autonomous

𝑢 𝑡 𝑥̇ 𝑡 = 𝐴 𝑥 𝑡 + 𝐵 𝑢(𝑡) 𝑦 𝑡 𝑥̇ 𝑡 = 𝐴 𝑥 𝑡 𝑦 𝑡
𝑦 𝑡 = 𝐶 𝑥 𝑡 + 𝐷 𝑢(𝑡) 𝑦 𝑡 =𝐶𝑥 𝑡
𝑥 0 = 𝑥! 𝑥 0 = 𝑥!

à System without inputs

(or 𝑢 𝑡 ≡ 0)

• 𝐴: system matrix / state-dynamics matrix, 𝑛" ×𝑛"

• 𝐵: input matrix, 𝑛" ×𝑛#
• 𝐶: output matrix, 𝑛$ ×𝑛"
• 𝐷: throughput matrix, 𝑛$ ×𝑛#

2
 Lecture Outline: Systems Analysis of Chemical Processes

Part 1: Time domain

L1: Modeling & state-space representation

L2: Linear algebra review 3.1: Autonomous linear time-invariant systems

L3: Analytical solution of linear systems

L4: Steady states & stability of linear systems

3.2 Non-autonomous linear time-invariant systems
L5: Nonlinear systems

Part 2: Frequency domain

L6: The Laplace transformation

L7: The transfer function

L8: Gain, poles, & zeros of transfer functions

L9: Graphical representation of transfer functions

L10: Common transfer functions

3
 Part 3.1:
Autonomous
Linear Time-Invariant Systems

4
 Intended Learning Outcomes

After this lecture part, you….

• know the formula for the analytical solution of autonomous LTI systems

• can compute the exponential of a matrix via diagonalization

• can describe the effect of distinct real and complex eigenvalues on the solution of LTI systems

5
 General Solution of Autonomous LTI Systems

𝑛" = 1
(reminder from calculus)

𝑥̇ 𝑡 = 𝑎 𝑥 𝑡
𝑎, 𝑥! , 𝑥 𝑡 ∈ ℝ
𝑥 0 = 𝑥!

General solution: 𝑥 𝑡 = 𝑒 %& 𝑥!

Verify:

𝑥̇ 𝑡 = 𝑎𝑒 %& 𝑥! = 𝑎𝑥(𝑡)

𝑥 0 = 𝑒 %⋅! 𝑥! = 𝑥!

6
7 Faculty, department, unit ...
 General Solution of Autonomous LTI Systems

𝑛" = 1
𝑛" > 1
(reminder from calculus)

𝑥̇ 𝑡 = 𝑎 𝑥 𝑡 𝑥̇ 𝑡 = 𝐴 𝑥 𝑡 𝑥! , 𝑥 𝑡 ∈ ℝ)!
𝑎, 𝑥! , 𝑥 𝑡 ∈ ℝ
𝑥 0 = 𝑥! 𝑥 0 = 𝑥! 𝐴 ∈ ℝ)!×)!

General solution: 𝑥 𝑡 = 𝑒 %& 𝑥! 𝑥 𝑡 = 𝑒 (& 𝑥! J

Exponential of a matrix?
Verify:

𝑥̇ 𝑡 = 𝑎𝑒 %& 𝑥! = 𝑎𝑥(𝑡)

𝑥 0 = 𝑒 %⋅! 𝑥! = 𝑥!

8
 The Matrix Exponential

Let A be an n×𝑛 matrix. Then,

-
1 + 1 . 1 /
𝑒( ≔ ; 𝐴 = 𝐼) + 𝐴 + 𝐴 + 𝐴 + ⋯ (analogous to the “normal” exponential function)
𝑘! 2! 3!
+,!

9
 The Matrix Exponential

Let A be an n×𝑛 matrix. Then,

-
1 + 1 . 1 /
𝑒( ≔ ; 𝐴 = 𝐼) + 𝐴 + 𝐴 + 𝐴 + ⋯ (analogous to the “normal” exponential function)
𝑘! 2! 3!
+,!

Relevant Properties:

0
• 𝑒 (& = 𝐴𝑒 (&
0&

• 𝑒 ( is invertible, and 𝑒 ( 12 = 𝑒 1(
𝜆2 ⋯ 0 𝑒 3" ⋯ 0 à If 𝐴 is diagonizable, exploit to compute 𝑒 (
• for diagonal 𝐴 = ⋮ ⋱ ⋮ : 𝑒 ( = ⋮ ⋱ ⋮ à If not, more complicated (see handout)
0 ⋯ 𝜆) 0 3
⋯ 𝑒 # à In practice numerical (e.g. MATLAB: expm)
$" (4
• 𝑒4 = 𝑇 12 𝑒 ( 𝑇

10
 Example 1: Diagonal System Matrix

𝜆2 0 3" ! 3" & !

𝑥̇ = 𝑥 1 1 3" & 1
= 𝑒 0
&
0 𝜆. ⇒ 𝑥 𝑡 = 𝑒 (& 𝑥! = 𝑒 ! 3% =𝑒 ! 3% &
1 3" &
1 1 0 𝑒 3% & 1
𝑥 0 = 𝑒
= 3&
1
𝑒 %

11
 Example 1: Diagonal System Matrix

𝜆2 0 3" ! 3" & !

𝑥̇ = 𝑥 1 1 3" & 1
= 𝑒 0
&
0 𝜆. ⇒ 𝑥 𝑡 = 𝑒 (& 𝑥! = 𝑒 ! 3% =𝑒 ! 3% &
1 3" &
1 1 0 𝑒 3% & 1
𝑥 0 = 𝑒
= 3&
1
𝑒 %

𝜆2 = −1, 𝜆. = −3: 𝜆2 = 0, 𝜆. = −3: 𝜆2 = 0.3, 𝜆. = −3:

𝑥2

𝑥2 𝑥2

𝑥. 𝑥.
𝑥.
𝑡 𝑡 𝑡

12
 Example 2: Non-Diagonal System Matrix With Distinct Real Eigenvalues

+0 +1 +1
𝑥̇ = 𝑥 𝜆2 = −1, 𝑣2 =
−3 −4 −1 +1 +1 +1.5 +0.5
1 ⇒ ⇒ 𝑇= , 𝑇 12 =
𝑥 0 = +1 −1 −3 −0.5 −0.5
1 𝜆. = −3, 𝑣. =
−3

𝜆2 0 𝜆2 0
𝑇 12 𝐴𝑇 = ⇔ 𝐴=𝑇 𝑇 12
0 𝜆. 0 𝜆.

13
 Example 2: Non-Diagonal System Matrix With Distinct Real Eigenvalues

+0 +1 +1
𝑥̇ = 𝑥 𝜆2 = −1, 𝑣2 =
−3 −4 −1 +1 +1 +1.5 +0.5
1 ⇒ ⇒ 𝑇= , 𝑇 12 =
𝑥 0 = +1 −1 −3 −0.5 −0.5
1 𝜆. = −3, 𝑣. =
−3

𝜆2 0 𝜆2 0
𝑇 12 𝐴𝑇 = ⇔ 𝐴=𝑇 𝑇 12
0 𝜆. 0 𝜆.

3" ! $" 3" & !

4 4 & 4 4 $"
⇒ 𝑥 𝑡 = 𝑒 (& 𝑥! = 𝑒 ! 3% 𝑥! =𝑒 ! 3% & 𝑥!
3" & ! 𝑥2
= 𝑇𝑒 ! 3% & 𝑇 12 𝑥! = 𝑇 𝑒 3" & 0 𝑇 12 𝑥!
0 𝑒 3% &
2𝑒 1&
− 𝑒 1/
𝑡 𝑥.
= 1& 1/&
−2𝑒 + 3𝑒
à Linear combination of modes!
𝑡

14
 Example 3: System Matrix With Complex Eigenvalues

+0 +1 +1
𝑥̇ = 𝑥 𝜆2 = −1 + 2𝑗, 𝑣2 =
−5 −2 −1 + 2𝑗 +1 +1 2 2−𝑗 −𝑗
1 ⇒ ⇒𝑇= , 𝑇 12 =
𝑥 0 = +1 −1 + 2𝑗 −1 − 2𝑗 5 2+𝑗 +𝑗
1 𝜆. = −1 − 2j, 𝑣. =
−1 − 2𝑗

𝜆2 0 𝜆2 0
𝑇 12 𝐴𝑇 = ⇔ 𝐴=𝑇 𝑇 12
0 𝜆. 0 𝜆.

15
 Example 3: System Matrix With Complex Eigenvalues

+0 +1 +1
𝑥̇ = 𝑥 𝜆2 = −1 + 2𝑗, 𝑣2 =
−5 −2 −1 + 2𝑗 +1 +1 2 2−𝑗 −𝑗
1 ⇒ ⇒𝑇= , 𝑇 12 =
𝑥 0 = +1 −1 + 2𝑗 −1 − 2𝑗 5 2+𝑗 +𝑗
1 𝜆. = −1 − 2j, 𝑣. =
−1 − 2𝑗

𝜆2 0 𝜆2 0
𝑇 12 𝐴𝑇 = ⇔ 𝐴=𝑇 𝑇 12
0 𝜆. 0 𝜆.
3" ! $" 3" &
⇒ 𝑥 𝑡 = 𝑒 (& 𝑥! =𝑒
4
! 3%
4 &
𝑥! =𝑇 𝑒 0 𝑇 12 𝑥!
0 𝑒 3% &
(127.8)&
=𝑇 𝑒 0
(121.8)&
𝑇 12 𝑥! Other suitable transformation matrix (real-valued)
0 𝑒

X 1& +cos(2𝑡) sin(2𝑡) X 12

8.&
= 𝑇𝑒 1& 𝑒 0 𝑇 12 𝑥! = 𝑇𝑒 𝑇 𝑥!
0 𝑒 1.8& −sin(2𝑡) cos(2𝑡)

Damping term Oscillation term

(recall: 𝑒 8.& = cos 2𝑡 + 𝑗 sin(2𝑡))

16
 Example 3: System Matrix With Complex Eigenvalues

+0 +1 +1
𝑥̇ = 𝑥 𝜆2 = −1 + 2𝑗, 𝑣2 =
−5 −2 −1 + 2𝑗 +1 +1 2 2−𝑗 −𝑗
1 ⇒ ⇒𝑇= , 𝑇 12 =
𝑥 0 = +1 −1 + 2𝑗 −1 − 2𝑗 5 2+𝑗 +𝑗
1 𝜆. = −1 − 2j, 𝑣. =
−1 − 2𝑗

𝜆2 0 𝜆2 0
𝑇 12 𝐴𝑇 = ⇔ 𝐴=𝑇 𝑇 12
0 𝜆. 0 𝜆.
3" ! $" 3" &
⇒ 𝑥 𝑡 = 𝑒 (& 𝑥! =𝑒
4
! 3%
4 &
𝑥! =𝑇 𝑒 0 𝑇 12 𝑥!
0 𝑒 3% &
(127.8)&
=𝑇 𝑒 0
(121.8)&
𝑇 12 𝑥! Other suitable transformation matrix (real-valued)
0 𝑒

X 1& +cos(2𝑡) sin(2𝑡) X 12

8.&
= 𝑇𝑒 1& 𝑒 0 𝑇 12 𝑥! = 𝑇𝑒 𝑇 𝑥!
0 𝑒 1.8& −sin(2𝑡) cos(2𝑡)

Damping term Oscillation term

(recall: 𝑒 8.& = cos 2𝑡 + 𝑗 sin(2𝑡))

17
 Example 3: System Matrix With Complex Eigenvalues – Visualization

𝑥(𝑡) = 𝑇𝑒 1& 𝑒 8.& 0 𝑇 12 𝑥!

0 𝑒 1.8&

𝑥2

𝑥.

18
 Example 3: System Matrix With Complex Eigenvalues – Visualization

𝑥(𝑡) = 𝑇𝑒 1& 𝑒 8.& 0 𝑇 12 𝑥!

0 𝑒 1.8&

Damping term Oscillation term

Oscillation term

𝑥2 Damping term

𝑥2

𝑥.

𝑡 𝑡

19
20 Faculty, department, unit ...
21 Faculty, department, unit ...
22 Faculty, department, unit ...
 Intended Learning Outcomes

After this lecture part, you….

ü know the formula for the analytical solution of autonomous LTI systems

ü can compute the exponential of a matrix via diagonalization

ü can describe the effect of distinct real and complex eigenvalues on the solution of LTI systems

23
 Part 3.1:
Non-Autonomous
Linear Time-Invariant Systems

24
 Intended Learning Outcomes

After this lecture part, you….

• know the formula for the general solution of non-autonomous LTI systems

• can explain the role of forced and free response

• can name and draw common test signals

• can compute step and impulse response of simple systems

• know the general response of stable LTI systems to oscillations

25
 Linear Time-Invariant (LTI) Models

Non-autonomous Autonomous

𝑢 𝑡 𝑥̇ 𝑡 = 𝐴 𝑥 𝑡 + 𝐵 𝑢(𝑡) 𝑦 𝑡 𝑥̇ 𝑡 = 𝐴 𝑥 𝑡 𝑦 𝑡
𝑦 𝑡 = 𝐶 𝑥 𝑡 + 𝐷 𝑢(𝑡) 𝑦 𝑡 =𝐶𝑥 𝑡
𝑥 0 = 𝑥! 𝑥 0 = 𝑥!

⇒ 𝑥 𝑡 =? ⇒ 𝑥 𝑡 = 𝐴𝑒 (& 𝑥!
𝑦 𝑡 =? 𝑦 𝑡 = 𝐶𝐴𝑒 (& 𝑥!

26
 General Solution of Non-Autonomous LTI Systems

𝑥̇ 𝑡 = 𝐴 𝑥 𝑡 + 𝐵 𝑢(𝑡)
𝑥 0 = 𝑥!

&

General solution: 𝑥 𝑡 = 𝑒 (& 𝑥! + a 𝑒 ( &1: 𝐵𝑢 𝜏 𝑑𝜏

!

Verify:
&

𝑥̇ 𝑡 = 𝐴𝑒 (& 𝑥! + a 𝐴𝑒 ( &1:
𝐵𝑢 𝜏 𝑑𝜏 + 𝑒 ( &1&
𝐵𝑢 𝑡 = 𝐴𝑥 𝑡 + 𝐵𝑢(𝑡)
!
!

𝑥 0 = 𝑒 (⋅! 𝑥! + a 𝑒 ( !1:
𝐵𝑢 𝜏 𝑑𝜏 = 𝑥!
!

27
 General Solution of Non-Autonomous LTI Systems

𝑥̇ 𝑡 = 𝐴 𝑥 𝑡 + 𝐵 𝑢(𝑡)
𝑥 0 = 𝑥!

&

General solution: 𝑥 𝑡 = 𝑒 (& 𝑥! + a 𝑒 ( &1: 𝐵𝑢 𝜏 𝑑𝜏

!

free response forced response

à Qualitative behavior: à Depends on 𝑢(𝑡) –
eigenvalues of 𝐴 which 𝑢(𝑡) to look at? à Common input signals
Verify:
& (“test signals”)
𝑥̇ 𝑡 = 𝐴𝑒 (& 𝑥! + a 𝐴𝑒 ( &1:
𝐵𝑢 𝜏 𝑑𝜏 + 𝑒 ( &1&
𝐵𝑢 𝑡 = 𝐴𝑥 𝑡 + 𝐵𝑢(𝑡)
!
!

𝑥 0 = 𝑒 (⋅! 𝑥! + a 𝑒 ( !1:
𝐵𝑢 𝜏 𝑑𝜏 = 𝑥!
!

28
 Common Input Signals: Step

𝑢
0, 𝑡<0
𝑢 𝑡 =d
1, 𝑡≥0

=: ℎ(𝑡), Heaviside function

29
 Common Input Signals: Ramp

0, 𝑡<0
𝑢 𝑡 =d
𝑢 𝑡, 𝑡≥0

Relation to step signal ℎ(𝑡):

&
𝑢 𝑡 = 𝑡 ℎ(𝑡) = a ℎ 𝜏 𝑑𝜏
1-

30
 Common Input Signals: Oscillation

𝑢
𝑢 𝑡 = 𝛼 sin(𝜔𝑡)

Amplitude (Angular)
𝛼
frequency
1 2𝜋
𝑇= 𝜔 = 2𝜋𝑓 =
𝑓 𝑇

31
 Common Input Signals: Pulse

0, 𝑡 < 0
𝑢 𝑡 = q𝛼, 0 ≤ 𝑡 < 1/𝛼
0, 𝑡 ≥ 1/𝛼
𝑢

Area under pulse stays the same for all 𝛼:

- 2/< 2/< 1
a 𝑢 𝑡 𝑑𝑡 = a 𝛼𝑑𝑡 = 𝛼𝑡r =𝛼 −0=1
1- ! ! 𝛼

32
 Common Input Signals: Impulse

𝑢 𝑡 =𝛿 𝑡 , Dirac impulse

Definition 1: 𝛿 𝑡 = lim 𝑢>?@AB (𝑡)

<→-
“infinitely short and high pulse”

̇
Definition 2: 𝛿 𝑡 = ℎ(𝑡)
𝑢 “derivative of step signal”

33
 Common Input Signals: Impulse

𝑢 𝑡 =𝛿 𝑡 , Dirac impulse

Definition 1: 𝛿 𝑡 = lim 𝑢>?@AB (𝑡)

<→-
“infinitely short and high pulse”

̇
Definition 2: 𝛿 𝑡 = ℎ(𝑡)
𝑢 “derivative of step signal”
-
Property 1: ∫1- 𝛿 𝑡 𝑑𝑡 = ℎ(𝑡)|-
1- = 1 − 0 = 1

Property 2: for a continuous function 𝑓 𝑡 ,

-
⇒ a 𝑓 𝑡 𝛿 𝑡 𝑑𝑡 = 𝑓(0)
𝑡 1-
-
⇒ a 𝑓 𝑡 𝛿 𝑡 − 𝜏 𝑑𝑡 = 𝑓(𝜏)
1-

34
 Example 1: Two Tanks in Series

−1 +0 1
𝑥(𝑡)
̇ = 𝑥(𝑡) + 𝑢(𝑡)
+2 −2 0
𝑦 𝑡 = 0 1 𝑥(𝑡)
0
𝑥 0 =
0
(from Exercise 1-A)
Task: Compute the step and impulse responses of the system!

à 𝑥(𝑡) and 𝑦(𝑡) for 𝑢(𝑡) = ℎ(𝑡) and 𝑢(𝑡) = 𝛿(𝑡)

35
 Example 1: Two Tanks in Series

−1 +0 1
𝑥(𝑡)
̇ = 𝑥(𝑡) + 𝑢(𝑡)
+2 −2 0
𝑦 𝑡 = 0 1 𝑥(𝑡)
0
𝑥 0 =
0
(from Exercise 1-A)
Task: Compute the step and impulse responses of the system!

à 𝑥(𝑡) and 𝑦(𝑡) for 𝑢(𝑡) = ℎ(𝑡) and 𝑢(𝑡) = 𝛿(𝑡)

Solution Procedure:

1. Compute 𝑒 (& (via diagonalization, if possible)

&

2. For given 𝑢(𝑡) and 𝑥! , insert into 𝑥 𝑡 = 𝑒 (& 𝑥! + a 𝑒 ( &1:

𝐵𝑢 𝜏 𝑑𝜏 &
!
𝑦 𝑡 = 𝐶 𝑥 𝑡 + 𝐷 𝑢 𝑡 = 𝐶𝑒 (& 𝑥! + a 𝐶𝑒 ( &1:
𝐵𝑢 𝜏 𝑑𝜏 + 𝐷𝑢(𝑡)
!

36
 Example 1: Two Tanks in Series – Compute Matrix Exponential

−1 +0 1 1
𝑥(𝑡)
̇ = 𝑥(𝑡) + 𝑢(𝑡) 𝜆2 = −1, 𝑣2 = 1 0 +1 0
+2 −2 0 2 ⇒ 𝑇= , 𝑇 12 =
⇒ 2 1 −2 1
𝑦 𝑡 = 0 1 𝑥(𝑡) 0
𝜆. = −2, 𝑣. =
0 1
𝑥 0 =
0

37
 Example 1: Two Tanks in Series – Compute Matrix Exponential

−1 +0 1 1
𝑥(𝑡)
̇ = 𝑥(𝑡) + 𝑢(𝑡) 𝜆2 = −1, 𝑣2 = 1 0 +1 0
+2 −2 0 2 ⇒ 𝑇= , 𝑇 12 =
⇒ 2 1 −2 1
𝑦 𝑡 = 0 1 𝑥(𝑡) 0
𝜆. = −2, 𝑣. =
0 1
𝑥 0 =
0

3" ! $" 3" & ! 3" &

𝑒 (&
=𝑒
4
! 3%
4 &
= 𝑇𝑒 ! 3% & 𝑇 12
=𝑇 𝑒 0 T 12
0 𝑒 3% &
Mode 1
1 0 𝑒 1& +1 0 𝑒 1& 0
= 0 =
2 1 0 𝑒 1.& −2 1 2𝑒 1& − 2𝑒 1.& 𝑒 1.&
Mode 2

38
 Example 1: Two Tanks in Series – Step Response

−1 +0 1
𝑥(𝑡)
̇ = 𝑥(𝑡) + 𝑢(𝑡)
+2 −2 0
0, 𝑡<0
𝑦 𝑡 = 0 1 𝑥(𝑡) 𝑢 𝑡 = ℎ(𝑡) = d
1, 𝑡≥0
0
𝑥 0 =
0
& &
1(&1:) 1
(&
⇒ 𝑥 𝑡 = 𝑒 𝑥! + a 𝑒 ( &1:
𝐵𝑢 𝜏 𝑑𝜏 = a 𝑒 0 1𝑑𝜏
2𝑒 1(&1:) − 2𝑒 1.(&1:) 𝑒 1.(&1:) 0
! !
=0 &
&
=a 𝑒 1(&1:) 𝑑𝜏 𝑒 1(&1:)
= y
2𝑒 1(&1:) − 2𝑒 1.(&1:) 2𝑒 1(&1:)
−𝑒 1.(&1:)
! !
1&
= 1 − 𝑒
1 − 2𝑒 1& + 𝑒 1.&

⇒𝑦 𝑡 = 0 1 𝑥 𝑡 = 1 − 2𝑒 1& + 𝑒 1.&

39
 Example 1: Two Tanks in Series – Step Response

−1 +0 1
𝑥(𝑡)
̇ = 𝑥(𝑡) + 𝑢(𝑡)
+2 −2 0
0, 𝑡<0
𝑦 𝑡 = 0 1 𝑥(𝑡) 𝑢 𝑡 = ℎ(𝑡) = d
1, 𝑡≥0
0
𝑥 0 =
0
& &
1(&1:) 1
(&
⇒ 𝑥 𝑡 = 𝑒 𝑥! + a 𝑒 ( &1:
𝐵𝑢 𝜏 𝑑𝜏 = a 𝑒 0 1𝑑𝜏
2𝑒 1(&1:) − 2𝑒 1.(&1:) 𝑒 1.(&1:) 0
! !
=0 &
&
=a 𝑒 1(&1:) 𝑑𝜏 𝑒 1(&1:)
= y
2𝑒 1(&1:) − 2𝑒 1.(&1:) 2𝑒 1(&1:)
−𝑒 1.(&1:)
! !
Mode 1
1&
= 1 − 𝑒
1 − 2𝑒 1& + 𝑒 1.&
Mode 1 Mode 2 Mode 1 Mode 2
⇒𝑦 𝑡 = 0 1 𝑥 𝑡 = 1 − 2𝑒 1& + 𝑒 1.&

40
 Example 1: Two Tanks in Series – Step Response: Visualization

1&
𝑥(𝑡) = 1 − 𝑒
1 − 2𝑒 1& + 𝑒 1.&

𝑦 𝑡 = 1 − 2𝑒 1& + 𝑒 1.&

~𝑒 1& ~𝑒 1&
(≙ Mode 1) (≙ Mode 1)

𝑢
𝑥2 𝑦 = 𝑥.

𝑥̇ . 0 = 0
𝑡 𝑡 𝑡

41
 Example 1: Two Tanks in Series – Impulse Response

−1 +0 1
𝑥(𝑡)
̇ = 𝑥(𝑡) + 𝑢(𝑡)
+2 −2 0
𝑦 𝑡 = 0 1 𝑥(𝑡) 𝑢 𝑡 = 𝛿(𝑡)
0
𝑥 0 =
0
&
𝑒 1&
(& ( &1: (&
⇒ 𝑥 𝑡 = 𝑒 𝑥! + a 𝑒 𝐵𝛿 𝜏 𝑑𝜏 =𝑒 𝐵 =
2𝑒 1& − 2𝑒 1.&
!
=0

𝑒 1! 1
Note: lim 𝑥(𝑡) = = ≠ 𝑥! à 𝑥2 „jumps“ at 𝑡 = 0
&→!& 2𝑒 1! − 2𝑒 1.⋅! 0

⇒𝑦 𝑡 = 0 1 𝑥 𝑡 = 2𝑒 1& − 2𝑒 1.&

42
 Example 1: Two Tanks in Series – Impulse Response: Visualization

1&
𝑥(𝑡) = 𝑒
2𝑒 1& − 2𝑒 1.&

𝑦 𝑡 = 2𝑒 1& − 2𝑒 1.&

𝑢 ~𝑒 1& 𝑥. = 𝑦 ~𝑒 1&
(≙ Mode 1) (≙ Mode 1)
𝑥2

𝑡 𝑡 𝑡

43
 Example 2: Response to Oscillation

𝑥̇ 𝑡 = −𝑥(𝑡) + 𝑢(𝑡)
𝑢 𝑡 = sin(𝜔𝑡) ⇒ 𝑒 (& = 𝑒 1&
𝑥 0 =1

free response forced response

& &
⇒ 𝑥 𝑡 = 𝑒 1& 𝑥! + a 𝑒 1(&1:) ⋅ 1 ⋅ sin(𝜔𝜏) 𝑑𝜏 = 𝑒 1& 𝑥! + 𝑒 1& a 𝑒 : sin(𝜔𝜏) 𝑑𝜏
! !
&
𝑒:
= 𝑒 1& 𝑥! + 𝑒 1& sin(𝜔𝜏) − 𝜔 cos(𝜔𝜏) •
1 + 𝜔.
!

C 2 C%
= 𝑒 1& 𝑥! + + sin 𝜔𝑡 − arcsin
27C% 27C% 27C%

44
 Example 2: Response to Oscillation

𝑥̇ 𝑡 = −𝑥(𝑡) + 𝑢(𝑡)
𝑢 𝑡 = sin(𝜔𝑡) ⇒ 𝑒 (& = 𝑒 1&
𝑥 0 =1

free response forced response

& &
⇒ 𝑥 𝑡 = 𝑒 1& 𝑥! + a 𝑒 1(&1:) ⋅ 1 ⋅ sin(𝜔𝜏) 𝑑𝜏 = 𝑒 1& 𝑥! + 𝑒 1& a 𝑒 : sin(𝜔𝜏) 𝑑𝜏
! !
&
𝑒:
= 𝑒 1& 𝑥! + 𝑒 1& sin(𝜔𝜏) − 𝜔 cos(𝜔𝜏) •
1 + 𝜔.
!
amplitude gain phase delay
C 2 C%
= 𝑒 1& 𝑥! + + sin 𝜔𝑡 − arcsin
27C% 27C% 27C%

transient response steady-state response

(contributions of forced + free!) (part of forced)

45
 General Response of (Stable) LTI Systems to Oscillations

for 𝑡 → ∞:
𝑢 𝑡 = 𝛼 sin(𝜔𝑡) 𝑦 𝑡 = A 𝜔 𝛼 sin(𝜔𝑡 − 𝜑 𝜔 )
LTI System

𝛼 𝐴(𝜔)𝛼

𝑢 𝑦

−𝛼 −𝐴(𝜔)𝛼
𝑡 𝑡

𝐴 𝜔 : (amplitude) gain 𝜑 𝜔 : phase delay

46
 Intended Learning Outcomes

After this lecture part, you….

ü know the formula for the general solution of non-autonomous LTI systems

ü can explain the role of forced and free response

ü can name and draw common test signals

ü can compute step and impulse response of simple systems

ü know the general response of stable LTI systems to oscillations

47
 Lecture Outline: Systems Analysis of Chemical Processes

Part 1: Time domain

L1: Modeling & state-space representation

L2: Linear algebra review 3.1: Autonomous linear time-invariant systems

L3: Analytical solution of linear systems

L4: Steady states & stability of linear systems

3.2 Non-autonomous linear time-invariant systems
L5: Nonlinear systems

Part 2: Frequency domain

L6: The Laplace transformation

L7: The transfer function

L8: Gain, poles, & zeros of transfer functions

L9: Graphical representation of transfer functions

L10: Common transfer functions

48
See you next week!