an introduction to optimal control applied to...
TRANSCRIPT
![Page 1: An Introduction to Optimal Control Applied to …dimacs.rutgers.edu/Workshops/AIMS/slides/SLafrica1.pdf · An Introduction to Optimal Control Applied to Disease Models Suzanne Lenhart](https://reader031.vdocuments.mx/reader031/viewer/2022021901/5b89fbf27f8b9a287e8d7a56/html5/thumbnails/1.jpg)
An Introduction to OptimalControl Applied to Disease
ModelsSuzanne Lenhart
University of Tennessee, Knoxville
Departments of Mathematics
Lecture1 – p.1/37
![Page 2: An Introduction to Optimal Control Applied to …dimacs.rutgers.edu/Workshops/AIMS/slides/SLafrica1.pdf · An Introduction to Optimal Control Applied to Disease Models Suzanne Lenhart](https://reader031.vdocuments.mx/reader031/viewer/2022021901/5b89fbf27f8b9a287e8d7a56/html5/thumbnails/2.jpg)
Example
� ��� �
Number of cancer cells at time
�(exponential growth) State� ��� �
Drug concentration Control
� ��� � � � ��� � � ��� �
� �� � � � � known initial data
minimize � � � �� � � ��� � ��
where the first term represents number of cancercells and the second term represents harmfuleffects of drug on body.
Lecture1 – p.2/37
![Page 3: An Introduction to Optimal Control Applied to …dimacs.rutgers.edu/Workshops/AIMS/slides/SLafrica1.pdf · An Introduction to Optimal Control Applied to Disease Models Suzanne Lenhart](https://reader031.vdocuments.mx/reader031/viewer/2022021901/5b89fbf27f8b9a287e8d7a56/html5/thumbnails/3.jpg)
Optimal Control
Adjust controls in a system to achieve a goalSystem:
Ordinary differential equations
Partial differential equations
Discrete equations
Stochastic differential equations
Integro-difference equations
Lecture1 – p.3/37
![Page 4: An Introduction to Optimal Control Applied to …dimacs.rutgers.edu/Workshops/AIMS/slides/SLafrica1.pdf · An Introduction to Optimal Control Applied to Disease Models Suzanne Lenhart](https://reader031.vdocuments.mx/reader031/viewer/2022021901/5b89fbf27f8b9a287e8d7a56/html5/thumbnails/4.jpg)
Deterministic Optimal Control
Control of Ordinary Differential Equations (DE)� ��� �
control� ��� �
stateState function satisfies DEControl affects DE
� � ��� � � � ��� � ��� �� � �� � �
� ��� �� � ��� �
Goal (objective functional)
Lecture1 – p.4/37
![Page 5: An Introduction to Optimal Control Applied to …dimacs.rutgers.edu/Workshops/AIMS/slides/SLafrica1.pdf · An Introduction to Optimal Control Applied to Disease Models Suzanne Lenhart](https://reader031.vdocuments.mx/reader031/viewer/2022021901/5b89fbf27f8b9a287e8d7a56/html5/thumbnails/5.jpg)
Basic Idea
System of ODEs modeling situationDecide on format and bounds on the controlsDesign an appropriate objective functionalDerive necessary conditions for the optimal controlCompute the optimal control numerically
Lecture1 – p.5/37
![Page 6: An Introduction to Optimal Control Applied to …dimacs.rutgers.edu/Workshops/AIMS/slides/SLafrica1.pdf · An Introduction to Optimal Control Applied to Disease Models Suzanne Lenhart](https://reader031.vdocuments.mx/reader031/viewer/2022021901/5b89fbf27f8b9a287e8d7a56/html5/thumbnails/6.jpg)
Design an appropriate objective functional–balancing opposing factors in functional–include (or not) terms at the final time
Lecture1 – p.6/37
![Page 7: An Introduction to Optimal Control Applied to …dimacs.rutgers.edu/Workshops/AIMS/slides/SLafrica1.pdf · An Introduction to Optimal Control Applied to Disease Models Suzanne Lenhart](https://reader031.vdocuments.mx/reader031/viewer/2022021901/5b89fbf27f8b9a287e8d7a56/html5/thumbnails/7.jpg)
Big Idea
In optimal control theory, after formulating aproblem appropriate to the scenario, there areseveral basic problems :
(a) to prove the existence of an optimal control,
(b) to characterize the optimal control,
(c) to prove the uniqueness of the control,
(d) to compute the optimal control numerically,
(e) to investigate how the optimal controldepends on various parameters in the model.
Lecture1 – p.7/37
![Page 8: An Introduction to Optimal Control Applied to …dimacs.rutgers.edu/Workshops/AIMS/slides/SLafrica1.pdf · An Introduction to Optimal Control Applied to Disease Models Suzanne Lenhart](https://reader031.vdocuments.mx/reader031/viewer/2022021901/5b89fbf27f8b9a287e8d7a56/html5/thumbnails/8.jpg)
Deterministic Optimal Control- ODEs
Find piecewise continuous control � ��� �and
associated state variable � �� �
to maximize
� � !"
���# � ��� �# � ��� � �$ �
subject to
� % ��� � & ' ���# � �� �# � ��� � �
� ��( � & � "and � � � )* *
Lecture1 – p.8/37
![Page 9: An Introduction to Optimal Control Applied to …dimacs.rutgers.edu/Workshops/AIMS/slides/SLafrica1.pdf · An Introduction to Optimal Control Applied to Disease Models Suzanne Lenhart](https://reader031.vdocuments.mx/reader031/viewer/2022021901/5b89fbf27f8b9a287e8d7a56/html5/thumbnails/9.jpg)
Contd.
Optimal Control + , -�. /
achieves the maximum
Put + , -�. /
into state DE and obtain 0 , -�. /
0 , -�. /
corresponding optimal state+ , -�. /
, 0 , -�. /
optimal pair
Lecture1 – p.9/37
![Page 10: An Introduction to Optimal Control Applied to …dimacs.rutgers.edu/Workshops/AIMS/slides/SLafrica1.pdf · An Introduction to Optimal Control Applied to Disease Models Suzanne Lenhart](https://reader031.vdocuments.mx/reader031/viewer/2022021901/5b89fbf27f8b9a287e8d7a56/html5/thumbnails/10.jpg)
Necessary, Sufficient Conditions
Necessary ConditionsIf 1 2 3�4 5
, 6 2 34 5
are optimal, then the followingconditions hold...
Sufficient ConditionsIf 1 2 3�4 5
, 6 2 3�4 5
and
7
(adjoint) satisfy the conditions...then 1 2 3�4 5
, 6 2 3�4 5are optimal.
Lecture1 – p.10/37
![Page 11: An Introduction to Optimal Control Applied to …dimacs.rutgers.edu/Workshops/AIMS/slides/SLafrica1.pdf · An Introduction to Optimal Control Applied to Disease Models Suzanne Lenhart](https://reader031.vdocuments.mx/reader031/viewer/2022021901/5b89fbf27f8b9a287e8d7a56/html5/thumbnails/11.jpg)
Adjoint
like Lagrange multipliers to attach DE to objective func-
tional.Lecture1 – p.11/37
![Page 12: An Introduction to Optimal Control Applied to …dimacs.rutgers.edu/Workshops/AIMS/slides/SLafrica1.pdf · An Introduction to Optimal Control Applied to Disease Models Suzanne Lenhart](https://reader031.vdocuments.mx/reader031/viewer/2022021901/5b89fbf27f8b9a287e8d7a56/html5/thumbnails/12.jpg)
Deterministic Optimal Control- ODEs
Find piecewise continuous control 8 9�: ;and
associated state variable < 9: ;
to maximize
= >? @A
9�:B < 9�: ;B 8 9�: ; ;C :
subject to
< D 9�: ; E F 9�:B < 9: ;B 8 9�: ; ;
< 9�G ; E < Aand < 9 ; HI I
Lecture1 – p.12/37
![Page 13: An Introduction to Optimal Control Applied to …dimacs.rutgers.edu/Workshops/AIMS/slides/SLafrica1.pdf · An Introduction to Optimal Control Applied to Disease Models Suzanne Lenhart](https://reader031.vdocuments.mx/reader031/viewer/2022021901/5b89fbf27f8b9a287e8d7a56/html5/thumbnails/13.jpg)
Quick Derivation of Necessary Condition
Suppose J K is an optimal control and L Kcorresponding state.
M N�O P
variation function,Q R .
t
u*(t)+ ah(t)
u*(t)
J K N�O P Q M N�O P
another control.S NOT Q P state corresponding to J K Q M
,
U S NOT Q PUO V W N OT S N�OT Q PT N J K Q M P N�O P P P
Lecture1 – p.13/37
![Page 14: An Introduction to Optimal Control Applied to …dimacs.rutgers.edu/Workshops/AIMS/slides/SLafrica1.pdf · An Introduction to Optimal Control Applied to Disease Models Suzanne Lenhart](https://reader031.vdocuments.mx/reader031/viewer/2022021901/5b89fbf27f8b9a287e8d7a56/html5/thumbnails/14.jpg)
Contd.
At
X Y Z
, [ \ Z] ^ _ Y ` a
t
x * (t)
y(t,a)
x 0
all trajectories start at same position
[ \ X] Z _ Y ` b \ X _ when ^ Y Z] control c b
d \ ^ _ Ye
a \ X] [ \ X] ^ _ ] c b \ X _ ^ f \ X _ _g X
Maximum ofd
w.r.t. ^ occurs at ^ Y Z
.Lecture1 – p.14/37
![Page 15: An Introduction to Optimal Control Applied to …dimacs.rutgers.edu/Workshops/AIMS/slides/SLafrica1.pdf · An Introduction to Optimal Control Applied to Disease Models Suzanne Lenhart](https://reader031.vdocuments.mx/reader031/viewer/2022021901/5b89fbf27f8b9a287e8d7a56/html5/thumbnails/15.jpg)
Contd.
hi jlk mhknononpnrqs t u v
wt
hhx jy j x mlz j x|{ k m m hx u y j} mlz j} { k m�~ y j v mlz j v{ k m
� wt
hhx jy j x mz j x|{ k m m hx � y j v mz j v{ k m ~ y j} mlz j} { k m u v��
Adding
v
to our
i jk mgives
Lecture1 – p.15/37
![Page 16: An Introduction to Optimal Control Applied to …dimacs.rutgers.edu/Workshops/AIMS/slides/SLafrica1.pdf · An Introduction to Optimal Control Applied to Disease Models Suzanne Lenhart](https://reader031.vdocuments.mx/reader031/viewer/2022021901/5b89fbf27f8b9a287e8d7a56/html5/thumbnails/16.jpg)
Contd.
� �l� ��� �� � ���|� � ��� � � �� � ��� � � � � ��� �� ��� � � ���|� � � � ��
� � �� � � �� � � ��� � �� � � �� � � �
� �� � � ��� � � ���|� � �� � ��� � � � � � � ��� � � ��� � � �
� � ��� �l� ���|� �� � �� � � �� �� � � �� �l �� � �� � � �� � � �
here we used product rule and � � � � ¡ ��
.
Lecture1 – p.16/37
![Page 17: An Introduction to Optimal Control Applied to …dimacs.rutgers.edu/Workshops/AIMS/slides/SLafrica1.pdf · An Introduction to Optimal Control Applied to Disease Models Suzanne Lenhart](https://reader031.vdocuments.mx/reader031/viewer/2022021901/5b89fbf27f8b9a287e8d7a56/html5/thumbnails/17.jpg)
Contd.
¢ £¢ ¤ ¥ ¦§ ¨ª© «¬« ¤ ® ¨ª¯ « °l± ² ® ¤ ³ ´« ¤ ® µ ¶ °�· ´ «¬« ¤
® µ °�· ´ ¸ © « ¬« ¤ ® ¸ ¯ « °± ² ® ¤ ³ ´« ¤ ¢ · ¹ µ °º ´ «« ¤ ¬ ° º¼» ¤ ´½
Arguments of
¨» ¸ terms are
°�·» ¬ °�·» ¤ ´» ± ² ® ¤ ³ °�· ´ ´
.
¾ ¥¢ £¢ ¤ ° ¾ ´ ¥ ¦§ ° ¨ª© ® µ ¸ © ® µ ¶ ´¢ ¬¢ ¤
¿o¿o¿o¿ÁÀ § ® ° ¨ª¯ ® µ ¸ ¯ ´ ³ ¢ ·
¹ µ ° º ´ «¬« ¤ ° º¼» ¾ ´½
Arguments of¨» ¸ terms are
°�·» Ã ² °�· ´» ± ² °· ´ ´
.
Lecture1 – p.17/37
![Page 18: An Introduction to Optimal Control Applied to …dimacs.rutgers.edu/Workshops/AIMS/slides/SLafrica1.pdf · An Introduction to Optimal Control Applied to Disease Models Suzanne Lenhart](https://reader031.vdocuments.mx/reader031/viewer/2022021901/5b89fbf27f8b9a287e8d7a56/html5/thumbnails/18.jpg)
Contd.Choose
Ä Å�Æ Ç
s.t.
È ÉÊ�Ë Ì�Í Î ÏÐªÑ Ê�Ë|Ò Ó ÔÒ Õ Ô Ì×Ö ÈÊË ÌlØ Ñ Ê�Ë Ò Ó ÔÒ Õ Ô ÌÙ adjoint equationÈÊÚ ÌÍ Û
transversality condition
ÛÍ ÜÝ Ê ÐªÞ Ö ÈØ Þ Ìß Ê�Ë Ìà Ë
ß Ê�Ë Ì
arbitrary variation
á ÐªÞ ÊË Ò Ó ÔÒ Õ Ô Ì Ö ÈÊ�Ë ÌØ Þ Ê�Ë Ò Ó ÔÒ Õ Ô ÌÍ Û
for all
Ûâ Ë â Ú ã
Optimality condition.
Lecture1 – p.18/37
![Page 19: An Introduction to Optimal Control Applied to …dimacs.rutgers.edu/Workshops/AIMS/slides/SLafrica1.pdf · An Introduction to Optimal Control Applied to Disease Models Suzanne Lenhart](https://reader031.vdocuments.mx/reader031/viewer/2022021901/5b89fbf27f8b9a287e8d7a56/html5/thumbnails/19.jpg)
Using Hamiltonian
Generate these Necessary conditions fromHamiltonian
ä�åæ çæ èæ é ê ë ä åæ çæ è ê éíì ä�åæ çæ è êintegrand (adjoint) (RHS of DE)
maximize w.r.t. è at è î
ïï è ë ð ñ éì ñ ë ðoptimality eq.
é ò ë ó ïï ç é ò ë ó äô éíìô ê
adjoint eq.é ä ê ë ðtransversality condition
Lecture1 – p.19/37
![Page 20: An Introduction to Optimal Control Applied to …dimacs.rutgers.edu/Workshops/AIMS/slides/SLafrica1.pdf · An Introduction to Optimal Control Applied to Disease Models Suzanne Lenhart](https://reader031.vdocuments.mx/reader031/viewer/2022021901/5b89fbf27f8b9a287e8d7a56/html5/thumbnails/20.jpg)
Converted problem of finding control to maximizeobjective functional subject to DE, IC to usingHamiltonian pointwise.For maximization
õöõø÷ ö ù
at ÷ú û ÷ü as a function of ÷
For minimizationõöõø÷ ö ù
at ÷ú û ÷ü as a function of ÷
Lecture1 – p.20/37
![Page 21: An Introduction to Optimal Control Applied to …dimacs.rutgers.edu/Workshops/AIMS/slides/SLafrica1.pdf · An Introduction to Optimal Control Applied to Disease Models Suzanne Lenhart](https://reader031.vdocuments.mx/reader031/viewer/2022021901/5b89fbf27f8b9a287e8d7a56/html5/thumbnails/21.jpg)
Two unknowns ý þ and ÿ þ
introduce adjoint
�
(like a Lagrange multiplier)
Three unknowns ý þ , ÿ þ and
�nonlinear w.r.t. ý
Eliminate ý þ by setting � � �and solve for ý þ in terms of ÿ þ and
�
Two unknowns ÿ þ and�
with 2 ODEs (2 point BVP)+ 2 boundary conditions.
Lecture1 – p.21/37
![Page 22: An Introduction to Optimal Control Applied to …dimacs.rutgers.edu/Workshops/AIMS/slides/SLafrica1.pdf · An Introduction to Optimal Control Applied to Disease Models Suzanne Lenhart](https://reader031.vdocuments.mx/reader031/viewer/2022021901/5b89fbf27f8b9a287e8d7a56/html5/thumbnails/22.jpg)
Pontryagin Maximum Principle
If � � ���
and � ���
are optimal for above problem, then thereexists adjoint variable
� ��
s.t.
� ���� � ��� � ��� � ��� � � ��� � ��� � � ��� � ���
at each time, where Hamiltonian�
is defined by
� ��� ��� � ��� � ��� � � ��� ��� � ��� � ��� ��� ��� � ��� �
and
� � ��� � � � � ��� ��� � ��� � ���
� � �� � �transversality condition
Lecture1 – p.22/37
![Page 23: An Introduction to Optimal Control Applied to …dimacs.rutgers.edu/Workshops/AIMS/slides/SLafrica1.pdf · An Introduction to Optimal Control Applied to Disease Models Suzanne Lenhart](https://reader031.vdocuments.mx/reader031/viewer/2022021901/5b89fbf27f8b9a287e8d7a56/html5/thumbnails/23.jpg)
Hamiltonian
��� � �� �! "! # $ % & � $ � � ! "! # $
# ' maximizes
�
w.r.t. #, �
is linear w.r.t. #
��� ( �� �! "! & $ # � $ % ) � ! "! & $
bounded controls, * + # �� $ + ,.
Bang-bang control or singular control
Example:
��� - # % & # % "/. & " 0
1 �1 # � - % & 2� 3
cannot solve for #
�
is nonlinear w.r.t. #, set
�54 � 3
and solve for # '
optimality equation.
Will discuss later
Lecture1 – p.23/37
![Page 24: An Introduction to Optimal Control Applied to …dimacs.rutgers.edu/Workshops/AIMS/slides/SLafrica1.pdf · An Introduction to Optimal Control Applied to Disease Models Suzanne Lenhart](https://reader031.vdocuments.mx/reader031/viewer/2022021901/5b89fbf27f8b9a287e8d7a56/html5/thumbnails/24.jpg)
Example 1
6 798:;
< = >@? A BDC ?subject to E F >? A G E >@? A = >@? AH E >JI A G K
What optimal control is expected?
Lecture1 – p.24/37
![Page 25: An Introduction to Optimal Control Applied to …dimacs.rutgers.edu/Workshops/AIMS/slides/SLafrica1.pdf · An Introduction to Optimal Control Applied to Disease Models Suzanne Lenhart](https://reader031.vdocuments.mx/reader031/viewer/2022021901/5b89fbf27f8b9a287e8d7a56/html5/thumbnails/25.jpg)
Example 1 worked
L MON PQ R SUT�V WX V
Y Z\[ Y ] R_^ YT` W [ a
b[ integrand ] c
RHS of DE [ R S ] cT Y ] R W
d bd R [ e R ] c[ ` f R g [ h ce at R g
d S bd R S [ e
c Z [ h d bd Y [ h c cT a W [ `
c[ c Q i j k l` [ c Q i j P f c Q[ `
cnm ` ^ R g m ` ^ Y g [ i k
Lecture1 – p.25/37
![Page 26: An Introduction to Optimal Control Applied to …dimacs.rutgers.edu/Workshops/AIMS/slides/SLafrica1.pdf · An Introduction to Optimal Control Applied to Disease Models Suzanne Lenhart](https://reader031.vdocuments.mx/reader031/viewer/2022021901/5b89fbf27f8b9a287e8d7a56/html5/thumbnails/26.jpg)
Example 2
o p9qrs
t u v@w x yz { v w x sD| w
subject to u } v@w x ~ u v w x { v@w x� u vJ� x ~ yz � s � y�
What optimal control is expected?
Lecture1 – p.26/37
![Page 27: An Introduction to Optimal Control Applied to …dimacs.rutgers.edu/Workshops/AIMS/slides/SLafrica1.pdf · An Introduction to Optimal Control Applied to Disease Models Suzanne Lenhart](https://reader031.vdocuments.mx/reader031/viewer/2022021901/5b89fbf27f8b9a287e8d7a56/html5/thumbnails/27.jpg)
Hamiltonian
� � �� � � � � � � �
The adjoint equation and transversality conditiongive
� � �@� � � ��
� � � � � � �� � � � � � � � �@� � � � ��� � � � �
Lecture1 – p.27/37
![Page 28: An Introduction to Optimal Control Applied to …dimacs.rutgers.edu/Workshops/AIMS/slides/SLafrica1.pdf · An Introduction to Optimal Control Applied to Disease Models Suzanne Lenhart](https://reader031.vdocuments.mx/reader031/viewer/2022021901/5b89fbf27f8b9a287e8d7a56/html5/thumbnails/28.jpg)
Example 2 continued
and the optimality condition leads to
� ��
��� � � � � � �@� � � � � ¡ ¢ £�¤ ¥¦
The associated state is
§ � �@� � � ¡¨ ¢ £�¤ ¥ ¡ ¦
Lecture1 – p.28/37
![Page 29: An Introduction to Optimal Control Applied to …dimacs.rutgers.edu/Workshops/AIMS/slides/SLafrica1.pdf · An Introduction to Optimal Control Applied to Disease Models Suzanne Lenhart](https://reader031.vdocuments.mx/reader031/viewer/2022021901/5b89fbf27f8b9a287e8d7a56/html5/thumbnails/29.jpg)
Graphs, Example 2
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2−1
0
1
2
3
Time
Sta
te
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 20
2
4
6
8
Time
Adj
oint
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2−8
−6
−4
−2
0
Time
Con
trol
Example 2.2
Lecture1 – p.29/37
![Page 30: An Introduction to Optimal Control Applied to …dimacs.rutgers.edu/Workshops/AIMS/slides/SLafrica1.pdf · An Introduction to Optimal Control Applied to Disease Models Suzanne Lenhart](https://reader031.vdocuments.mx/reader031/viewer/2022021901/5b89fbf27f8b9a287e8d7a56/html5/thumbnails/30.jpg)
Example 3
©ª
«¬® ¯ ¬ ° ¯® °U±² ³
® ´\µ ® ¶ ¬_· ® «¸ ± µ ¹
ºµ ¬® ¯ ¬ ° ¯® ° ¶ » «® ¶ ¬ ±
¼ º¼¬ µ ® ¯ ¹¬ ¶ »µ ½
at ¬ ¾ ¿ ¬ ¾ µ ® ¶ »¹
» ´µ ¯ ¼ º¼® µ ¯ «¬ ¯ ¹® ¶ »±· » «À ± µ ½
» ´µ ¯ ® ¶ »¹ ¯ ¹® ¶ »
® ´Áµ ® ¶® ¶ »¹
Lecture1 – p.30/37
![Page 31: An Introduction to Optimal Control Applied to …dimacs.rutgers.edu/Workshops/AIMS/slides/SLafrica1.pdf · An Introduction to Optimal Control Applied to Disease Models Suzanne Lenhart](https://reader031.vdocuments.mx/reader031/viewer/2022021901/5b89fbf27f8b9a287e8d7a56/html5/thumbnails/31.jpg)
Contd.
 ÃÅÄ ÆÇ Â È ÉÇ
É ÃÅÄ ÆÇ Â/Ê ÆÇ É
 ËÌ Í Ä ÇÏÎ É ËÐ Í Ä ÑSolve for Â Ò Î É
and then get Ó Ò .Do numerically with Matlab or by hand
ÂÉ Ä ÔÖÕ ÌÇ ÆÊ Æ × ØÙÚ È ÔÖÛ ÌÊ Ç ÆÊ Æ × Ü ØÙ Ú
Lecture1 – p.31/37
![Page 32: An Introduction to Optimal Control Applied to …dimacs.rutgers.edu/Workshops/AIMS/slides/SLafrica1.pdf · An Introduction to Optimal Control Applied to Disease Models Suzanne Lenhart](https://reader031.vdocuments.mx/reader031/viewer/2022021901/5b89fbf27f8b9a287e8d7a56/html5/thumbnails/32.jpg)
Exercise
ÝÞßàá
âãåä æ çè é
ä ê ë ì í æ îï ä ãJð ç ë ì ï æ control ñ ë
òò æ ë ð
ó ê ë ó ë
ó ã ì ç ëæ ô ë ä ô ë
Lecture1 – p.32/37
![Page 33: An Introduction to Optimal Control Applied to …dimacs.rutgers.edu/Workshops/AIMS/slides/SLafrica1.pdf · An Introduction to Optimal Control Applied to Disease Models Suzanne Lenhart](https://reader031.vdocuments.mx/reader031/viewer/2022021901/5b89fbf27f8b9a287e8d7a56/html5/thumbnails/33.jpg)
Exercise completed
õ ö÷øù
úûü ýÿþ � ��
ü � � � � þ �� ü û � � � þ control
û � � ü � þ � � � � ü ý þ ý � û � � þ � �
� �þ � � � �þ � � þ � � � � ø ø � � �� �
� � � ��
�ü � � �� � û � � � � � � � � � �
ü � � � � þ � � � ��
� û � � � � �
ü � û � � � � � � � � û � � � � ý � � �� þ � û � � � � � û � � � ��
Lecture1 – p.33/37
![Page 34: An Introduction to Optimal Control Applied to …dimacs.rutgers.edu/Workshops/AIMS/slides/SLafrica1.pdf · An Introduction to Optimal Control Applied to Disease Models Suzanne Lenhart](https://reader031.vdocuments.mx/reader031/viewer/2022021901/5b89fbf27f8b9a287e8d7a56/html5/thumbnails/34.jpg)
Contd.
��
��� � ��� � � ��� �! "� #There is not an “Optimal Control" in this case.
Want finite maximum.
Here unbounded optimal stateunbounded OC
Lecture1 – p.34/37
![Page 35: An Introduction to Optimal Control Applied to …dimacs.rutgers.edu/Workshops/AIMS/slides/SLafrica1.pdf · An Introduction to Optimal Control Applied to Disease Models Suzanne Lenhart](https://reader031.vdocuments.mx/reader031/viewer/2022021901/5b89fbf27f8b9a287e8d7a56/html5/thumbnails/35.jpg)
Opening Example
$ %�& '
Number of cancer cells at time
&(exponential growth) State
( %�& '
Drug concentration Control
) $)& * + $ %�& ' , ( %�& '
$ %.- ' * $0/ known initial data
minimize $ % ' 1/ ( 2 %�& ' )&
See need for bounds on the control. See salvageterm.
Lecture1 – p.35/37
![Page 36: An Introduction to Optimal Control Applied to …dimacs.rutgers.edu/Workshops/AIMS/slides/SLafrica1.pdf · An Introduction to Optimal Control Applied to Disease Models Suzanne Lenhart](https://reader031.vdocuments.mx/reader031/viewer/2022021901/5b89fbf27f8b9a287e8d7a56/html5/thumbnails/36.jpg)
Further topics to be covered
Interpretation of the adjointSalvage termNumerical algorithmsSystems caseLinear in the control caseDiscrete models
Lecture1 – p.36/37
![Page 37: An Introduction to Optimal Control Applied to …dimacs.rutgers.edu/Workshops/AIMS/slides/SLafrica1.pdf · An Introduction to Optimal Control Applied to Disease Models Suzanne Lenhart](https://reader031.vdocuments.mx/reader031/viewer/2022021901/5b89fbf27f8b9a287e8d7a56/html5/thumbnails/37.jpg)
more info
See my homepage www.math.utk.edu 3 4lenhartOptimal Control Theory in Application to Biologyshort course lectures and lab notesBook: Optimal Control applied to Biological ModelsCRC Press, 2007, Lenhart and J. Workman
Lecture1 – p.37/37