Introduction Modeling Lemmas Control without delay Proposed controller Simulation results Conclusion and Perspectives

.. .. .. .. .. . .. .. .. .. .. . ..

Control and stabilization of delayed systems

aft
Xxxx YYYYYY
Master Automatic control and systems
Dr ———
Faculty Sciences and Technology
———
University of Djelfa

August 20, 2023

1 / 14
Introduction Modeling Lemmas Control without delay Proposed controller Simulation results Conclusion and Perspectives
.. .. .. .. .. . .. .. .. .. .. . ..

Summery

aft
1 Introduction

2 Modeling

3 Lemmas

4 Control without delay

Dr
5 Proposed controller

6 Simulation results

7 Conclusion and Perspectives

2 / 14
Introduction Modeling Lemmas Control without delay Proposed controller Simulation results Conclusion and Perspectives
.. .. .. .. .. . .. .. .. .. .. . ..

aft
Motivations:
◀ Most systems are nonlinear
◀ Delay complicates the system analysis
◀ Delay can lead to the system instability
◀ ...
Dr
3 / 14
Introduction Modeling Lemmas Control without delay Proposed controller Simulation results Conclusion and Perspectives
.. .. .. .. .. . .. .. .. .. .. . ..

The TS fuzzy model can be justified by:

aft
◀ Its simplicity
◀ Uncertainties
◀ Its acceptable accuracy
◀ ....

Dr
4 / 14
Introduction Modeling Lemmas Control without delay Proposed controller Simulation results Conclusion and Perspectives
.. .. .. .. .. . .. .. .. .. .. . ..

The TS fuzzy model can be justified by:

aft
◀ Its simplicity
◀ Uncertainties
◀ Its acceptable accuracy
◀ ....

The system can be modeled as:

{ C α
D x(t) = f(x(t), x(t − τ (t)), u(t)), t ≥ 0,
Dr
x(s) = φ(s), s ∈ [−τ, 0]

x(t) ∈ ℜn the system state

u(t) ∈ ℜm the control vector
τ : the delay

4 / 14
Introduction Modeling Lemmas Control without delay Proposed controller Simulation results Conclusion and Perspectives
.. .. .. .. .. . .. .. .. .. .. . ..

Lemma 1
∫ b(t)

aft
I(x, t) = f(x, t) dx , a(t) & b(t) < ∞
a(t)

with a(t) � b(t) and f(x, t)

∫ b(t)
d(I(x, t)) db(t) da(t) ∂f(x, t)
= f(b(t), t) − f(a(t), t) + dx
dt dt dt a(t) ∂t

Lemma 2 Dr ∫ b(t)
I(x, t) = f(x, t) dx , a(t) & b(t) < ∞
a(t)

with a(t) � b(t) and f(x, t)

∫ b(t)
d(I(x, t)) db(t) da(t) ∂f(x, t)
= f(b(t), t) − f(a(t), t) + dx
dt dt dt a(t) ∂t

5 / 14
Introduction Modeling Lemmas Control without delay Proposed controller Simulation results Conclusion and Perspectives
.. .. .. .. .. . .. .. .. .. .. . ..

aft
Theorem
Let’s consider: {
Ωii < 0, (i = 1, 2, ...r),
Ωij + Ωji < 0, i < j, i, j = 1, 2, ...r
with:
 
PAi + AT T T T T
i P + PBi Kj + Kj Bi P + Q PAdi AT T T T T
i N + Kj Bi N 0
 Dr 

Ωij =  ∗ −(1 − µ)Q ATdi N
T
0 

 ∗ ∗ τ 2 R − N − NT 0 
∗ ∗ ∗ −R

6 / 14
Introduction Modeling Lemmas Control without delay Proposed controller Simulation results Conclusion and Perspectives
.. .. .. .. .. . .. .. .. .. .. . ..

∑r

aft
We consider: u(t) = i=1 hi (θ(t))[Ki x(t) + Kdi x(t − τ (t)]

Theorem
{
Ω¯ii < 0, i = 1, 2, ...r
Ω¯ij + Ω¯ji < 0, i < j, i, j = 1, 2, ...r

where:
Ai X + XT AT T T
i + Bi Yj + Yj Bi + Q̄ Adi X + Bi Ydj ϵXT AT T T
i + ϵYj Bi 0

Ω¯ij = 

Dr ∗
∗
∗
Kj = Yj X−1 � Kdj = Ydj X−1 (j = 1, 2, ...r)
( 1 − µ)Q̄
∗
∗
ϵX Adi + ϵYT
T T

τ 2 R̄ − ϵX − ϵXT
∗
T
di Bi 0
0
−R̄

7 / 14
Introduction Modeling Lemmas Control without delay Proposed controller Simulation results Conclusion and Perspectives
.. .. .. .. .. . .. .. .. .. .. . ..

System behavior without controller

1051
0.5

aft
0

-0.5
States

-1

-1.5
Dr
-2

-2.5
0 5 10 15 20 25 30 35 40 45 50
Time(second)

Figure 1: System state

System unstable
8 / 14
Introduction Modeling Lemmas Control without delay Proposed controller Simulation results Conclusion and Perspectives
.. .. .. .. .. . .. .. .. .. .. . ..

System behavior with controller

aft
0

-5
Control input

-10

-15

Dr
-20

-25

-30
0 5 10 15 20 25 30 35 40 45 50
Time(second)

Figure 2: Control signal

9 / 14
Introduction Modeling Lemmas Control without delay Proposed controller Simulation results Conclusion and Perspectives
.. .. .. .. .. . .. .. .. .. .. . ..

System behavior with controller

aft
0.5
States

0
Dr
-0.5
0 5 10 15 20 25 30 35 40 45 50
Time(second)

Figure 3: System state

The system is stable but it needs more enhancement

10 / 14
Introduction Modeling Lemmas Control without delay Proposed controller Simulation results Conclusion and Perspectives
.. .. .. .. .. . .. .. .. .. .. . ..

System behavior with the proposed controller

aft
0

-5
Control input

-10

-15

Dr
-20

-25

-30
0 5 10 15 20 25 30 35 40 45 50
Time(second)

Figure 4: Proposed controller signal

11 / 14
Introduction Modeling Lemmas Control without delay Proposed controller Simulation results Conclusion and Perspectives
.. .. .. .. .. . .. .. .. .. .. . ..

System behavior with the proposed controller

0.8

aft
0.6

0.4

0.2

0
States

-0.2

-0.4
Dr
-0.6

-0.8

-1
0 5 10 15 20 25 30 35 40 45 50
Time(second)

Figure 5: System state with the proposed controller

Stability + better performance

12 / 14
Introduction Modeling Lemmas Control without delay Proposed controller Simulation results Conclusion and Perspectives
.. .. .. .. .. . .. .. .. .. .. . ..

Quantification of the comparative study

Classical controller Proposed controller Enhancement rate

aft
Settling time 26 20 23 %
Pic to pic x3 1.36 1.16 15 %
∫ts 2
(x3 )dt 2.5540 0.7771 70 %
0
∫ts
(u2 )dt 12.8476 7.1868 40 %
0

Table 1: Quantification of the comparative study

Dr
We remark that
◀ Enhancement of the settling time of 23%
◀ Reduction of the control energy by 40%
◀ Overall enhancement by 70%
◀ Pic to pic reduction by 15%

13 / 14
Introduction Modeling Lemmas Control without delay Proposed controller Simulation results Conclusion and Perspectives
.. .. .. .. .. . .. .. .. .. .. . ..

We conclude that:

aft
Conclusion
◀ Lyapunov method efficiency.
◀ Proposed controller leads to better performance.
◀ Delayed controller enhances the performance.
◀ Proposed approach allows reduction of the control energy.

As perspectives we propose:
Dr
perspectives
◀ Perspective 1.
◀ Perspective 2.
◀ Perspective 3.

14 / 14