SlideShare uma empresa Scribd logo
1 de 6
Baixar para ler offline
International Journal on Cybernetics & Informatics (IJCI) Vol. 3, No. 4, August 2014 
OPTIMAL CONTROL PROBLEM FOR PROCESSES 
REPRESENTED BY STOCHASTIC SEQUENTIAL MACHINE 
Yakup H. HACI and Muhammet CANDAN 
Department of Mathematics, Canakkale Onsekiz Mart University, Canakkale, Turkey 
ABSTRACT 
In this paper, optimal control problem for processes represented by stochastic sequential machine is 
analyzed. Principle of optimality is proven for the considered problem. Then by using method of dynamical 
programming, solution of optimal control problem is found. 
Keywords 
Optimal control problem, stochastic sequential machine, dynamical programming. 
1. INTRODUCTION 
Stochastic sequential machine (SSM) is the one of the most developing field of discrete system 
theory [2], [3]. It plays an important role in many areas such as construction of finite dynamical 
system, imitation modelling problem, coding of discrete systems and identification problems. 
Thus, it points out that it requires a comprehensive research. 
2. SSM 
SSM is generalization of multi-parametric finite sequential machine[3] but it contains probability 
variable. General form of this system is defined by[6]: 
K =< X , S,Y, s , p( ), F (.),G > 0 
n w v =1,2,...k (1) 
where X = [ GF(2) ] r , S = [ GF(2) ]m and Y = [ GF(2) ]q are input, state and output index 
(alphabet) respectively; s 0 is initial state vector, p(w) is determitistic discrete probability 
distribution ( W = { w , w ,..., w } is finite set, p ( w ) = { p ( w ) : w ÎW,  p 
( w ) = 1}) 
, 
1 2 p i i i w 
ÎW i 
characteristic Boolean vector functions [7] denoted by ( × ) = { ( × ),..., ( × 
)} 
F F Fn n n which are also 
1 
m 
known as transfer functions are nonlinear functions defined on the set 
[ ] [ ]k m r Z × GF(2) × GF(2) and G are an output characteristic functions defined on GF(2) 
where GF(2) is Galois field[8] and the symbol (.) denotes (c, s(c), x(c)) for simplicity. 
In addition to the definition, SSM is represented by: 
s(c e ) F (c, s(c), x(c),w(c)) n n + = v = 1,2,...k (2) 
DOI: 10.5121/ijci.2014.3403 21
International Journal on Cybernetics  Informatics (IJCI) Vol. 3, No. 4, August 2014 
)y(c) = G(s(c ) (3) 
where s(c), x(c), y(c) are m, r and q dimensional state, input and output vectors at the point 
c respectively, w(c) is a random variable [6], c {c c Z c c c c c c c Z} i 
22 
= Î k , £ £ L ,..., 0 
£ £ L 
k , Î k k k 
1 1 
0 
1 
1 
is point in k Z , determining position i L , ( i = 1,2,...k positive integer) is the duration of the stage 
n 
n e = . 
i of this process. Z is set of integers and (0,...,0,1,0,...,0) 
In SSM, each random variable w has a special case. For instance, w is an input variable in the 
identification problem of the SSM. However, w is a set of all possible states in the synthesis of 
optimal sequential machine 
Moreover, the state of the system depends on the random variable w which affects not only 
parameters of the SSM but also the input variable. 
Finally, equation (2) is transformed to: 
s(c e ) F (c, s(c), x(c) w(c)) n n + = Å v = 1,2,..., k (4) 
where symbol Åmeans that x Å w is always in input alphabet X . 
The discrete optimal processes given by SSM are characterized by functional: 
( ) { ( ( ))} 
__ 
L J x M f s c w = (5) 
where {.} w M is a mathematical expected value of w . 
3. OPTIMAL CONTROL PROBLEM AND PRINCIPLE OF OPTIMALITY 
We can state optimal control problem [2] for processes represented by SSM as below: In order 
for given SSM to start from the known inital state 0 s to go any desired state ( ) * L s c , to which 
we expect to access in L steps, a control x(c)Î X [4] must exist such that the functional in (1) 
has a minimal value: 
d s(c + e ) = Fˆ (c, s(c), x(c),w(c)),cÎG n n v =1,2,...k (6) 
0 0 s(c ) = s , x(c)Î X , d cÎG (7) 
ˆ ( , ˆ ( , ( ), ( )), ( ), ( )) ˆ ( ˆ ( , ( ), ( )), ( ) ( ) 
n μ μ μ μ μ n n n n F c + e F c s c x c x c + e w c + e = F c + e F c s c x c x c + e w c + e 
(8) 
( ) { ( ( )} min 
__ 
= ® L J x M f s c w (9) 
where ˆ (.) 
n F (n = 1,..., k) denotes the pseudo Boolean expression of the Boolean vector 
^ 
F n = k and k L = L + L +...+ L 1 2 is the time duration of this process. 
function[7] (.)( 1,..., ) 
It is well-known that method of dynamic programming [5] is used for solution of optimal control 
problem. If we make use of this method to solve the considered problem then, (6)-(9) can be 
formulate as an optimal problem: 
( ) ˆ ( , ( ), ( ),w( )), (s ) n n d s c + e = F c s c x c c cÎG (10)
International Journal on Cybernetics  Informatics (IJCI) Vol. 3, No. 4, August 2014 
23 
s(s ) =À (11) 
( ) , (s ) d x c Î X cÎG (12) 
ˆ ( , ˆ ( , ( ), ( ), ( )), ( ), ( )) ˆ ( , ˆ ( , ( ), ( ), ( )), ( ), ( )) 
n μ μ μ μ μ n n n n F c+e F c s c x c w c x c+e w c+e =F c+e F c s c x c w c x c+e w c+e 
(13) 
( ) { ( ( ))} min 
__ 
= ® L J x M f s c w (14) 
where À is an arbitrary element in S . As it can be seen from (10)-(14), if we substitute 0 s = c 
or 0 À= s into problem (10)-(14) we obtain first problem stated above. If the conditions for 
existence of unique solution are satisfied then for the given initial condition s(s ) =Àand given 
( )( (s )), d x c cÎG we find a unique s(c) . That is, the functional (14) is the function of the 
parameters À and ( )( (s )) d x c cÎG : 
__ __ 
J ( x ) = J ( À , x ( G ( s )) 
(15) 
d where ( (s )) d x G denotes the range of the control x(c) on the points (s ) d cÎG : 
( (s )) { ( ), (s )} d d x G = x c cÎG (16) 
from the unique solution condition of the system (6), we find that the stochastic process can be 
investigated in the set (s ) d G and also in the set 
{ } d k k k G s = c c £ c s c £ c s 0 
( ) ; 0 
,..., 
1 1 1 
1 
(17) 
Definition. We say that the control ( ), ( (s )) d x c cÎG which minimizes the functional (5) in the 
problem (10)-(14) is optimal control with respect to the initial pair (s ,À) on the region (s ) d G 
Suppose that ( ) 0 x c is an optimal control with respect to the initial pair ( , ) 0 0 c s on the region d G 
and ( ) 0 s c is admissible optimal trajectory. Then ( ) 0 x c is an optimal control with respect to the 
initial pair ( , ( )) 0 s s s on the region (s ) d G for every d G . 
Proof. Suppose the contrary. Then there exists ( ) , (s ) d x c Î X cÎG such that we have 
__ __ 
( , ( ( )) ( , ( ( )) 0 
s s d d J À x G  J À x G (18) 
We choose a new control process d ~x(c), cÎG as follows: 
   
Î 
s 
( ), ( ) 
0 
x c c G 
~( ) 1 
x c (19) 
Î 
= 
s 
d 
( ), ( ) 
x c c G 
d 
As it can be seen, (19) is an admissible control such that
International Journal on Cybernetics  Informatics (IJCI) Vol. 3, No. 4, August 2014 
24 
( , ~( )) ( , ~( ( ) ( )). 
1 
0 
__ 
0 
__ 
s s d d d J s x G = J s x G ÈG (20) 
According to the condition, ( ) =À 0 s s . Thus we have 
__ 
( , ~( ( ) È ( )) = ( ( ), ~( ( )) = 0 
J s 0 
x G s G s J s s x G s = (À, ( ( )))  (À, 0 
( ( ))) = d d d __ 
1 
__ __ 
s s d d J x G J x G 
__ 
( ( ), 0 
( ( ))) ( 0 , 0 
( )) __ 
d d = J s s x G s = J s x G (21) 
and by virtue of (2.10) and (2.11) we can obtain 
__ 
( 0 
, ~( )) ( 0 , 0 
( )) __ 
d d J s x G  J s x G (22) 
Hence, (22) contradicts the hypothesis that the control d x (c),cÎG 0 is optimal. This completes 
the proof of the theorem. 
Let a function (for every fixed s and À) be corresponding to the optimal value of pseudo 
Boolean functional in the problem (10)-(14): 
( , ) min { ( ( ))} L B s M f s c w À = . (23) 
where minimization on the set of admissible control ( ), (s ) d x c cÎG . 
Now, we determine method of dynamical programming(It is known as Bellman equation) [5] for 
B(s ,À) . Suppose that d x (c),cÎG 0 is the admissible control corresponding to (10)-(14) with 
initial condition and ( ), ( ) 0 s d s c cÎG is also the optimal trajectory. Let the point 
s G ( )(v 1,2,...,k) d x Î s = n and any element y(c)Î X be specified. If x(s ) = y(c) , then the 
state of the system in the point x s n is determined by 
(x s ) ˆ (s , , ,w(s )) n n s = F À y (24) 
We consider the following problem: 
x ( ) ˆ ( , ( ), ( ),w( )), (x s ) n n d n s c = F c s c x c c cÎG (25) 
s(x s ) = Fˆ (s ,À, y) n n (26) 
( ) ( ), (x s ) d n x c Î X c cÎG (27) 
( ) { ( ( ))} min 
__ 
= ® L J x M f s c w (28) 
If ˆ( ), (x s ) d n y c cÎG and ˆ( ), (x s ) d n s c cÎG are optimal control and optimal trajectory 
respectively, then
International Journal on Cybernetics  Informatics (IJCI) Vol. 3, No. 4, August 2014 
{f ( ( ))} (x s , ˆ (s , , ˆ,w(s ))) w n n M s c B F y L = À (29) 
x c (30) 
25 
can be found. 
For (10)-(14), let ~x (c) be an admissible control below. 
   
= 
Î 
= 
, 
y c 
ˆ( ), ( ) 
~( ) 
x s 
s 
d n y c c G 
Also, ~s (c) can be obtained by 
 
  
 
  
ˆ ( , , , ( )), 
s À w s = 
s 
F y c 
s c (31) 
Î 
= 
ˆ( ) , ( ) 
~( ) 
x s 
n 
n 
d s c c G 
__ 
It is evident that the value of ( ) { ( ( ))} 
L J x M f s c w = to control ~x (c) is determined by 
{ (~( ))}= { ( ˆ( ))}= L L M f s c M f s c w w (x s , ˆ (s , , ˆ,w(s ))) n n = B F À y . (32) 
Since ~( ), (s ) d x c cÎG is not largely optimal control, we can state 
{ (~( ))} { ( ( ))} ( , ) 0 f ³ f = s À w w M s c M s c B L L (33) 
Thus, we have 
(s , ) (x s , ˆ (s , , ˆ,w(s ))) n n B À £ B F À y (34) 
On the other hand, if ( ) ( ), 0 y c = x s then by the principle of optimality [2], 
ˆ( )( ( )) ( )( ( )) 0 x s x s d n d n y c cÎG = x c c ÎG (35) 
Therefore, 
( , ) ( , ˆ ( , , ( ), ( ))) 0 s x s s s w s n n B À = B F À x (36) 
By (34) and (35), Bellman equation[5] can be determined by 
( , ) min ( , ˆ ( , , ˆ, ( ))), 
s x s s w s n n 
B B F y 
S 
y X 
ÀÎ 
À = À 
Î 
s (37) 
( )
International Journal on Cybernetics  Informatics (IJCI) Vol. 3, No. 4, August 2014 
26 
4. CONCLUSIONS 
It is shown that Bellman equation for optimal processes with stochastic sequential machine is 
obtained and the principle of optimality is proven. 
REFERENCES 
[1] Anderson, J. A., (2004) Discrete Mathematics witk Combinatorics, Prentice Hall, New Jersey, p.45 
“This is my paper”, ABC Transactions on ECE, Vol. 10, No. 5, pp120-122. 
[2] Boltyanskii, V. G., (1978) Optimal Control of Discrete Systems, John Willey, New York, p.363. 
[3] Gaishun, I. V., (1983) Completely Solvable Multidimensional Differential Equations, Nauka and 
Tekhnika, Minsk, p.231. 
[4] Hacı, Y., Ozen, K., (2009) “Terminal Control Problem for Processes Represented by Nonlinear Multi 
Binary Dynamic System”, Control and Cybernetics, Vol. 38, No. 3, pp625-633. 
[5] Bellman, R., (1957) Dynamic Programming, Princeton University Press, Princeton, p.12 
[6] Yermolyev, Y. M., (1976) Stochastic Programming Methods, Nauka (in Russian), p.240. 
[7] Yablonsky, S. V., (1989) Introduction to Discrete Mathematics, Mir Publishers, Moscow, p.9. 
[8] Fraileigh, J. B., (1998) A First Course in Abstract Algebra, 6th ed. Addison--Wesley Publisher, 
p.419-423 
Authors 
Yakup H. HACI received Ph.D at physics and mathematics in Azerbaijan NationalAcademy of Sciences 
at the Institute of Cybernetics. He is currently working as professor and head of department of 
Mathematics, in Canakkale Onsekiz Mart University. His primary research interests are in the areas of the 
theory of multiparametric binary sequential machines,linear and nonlinear optimization and discrete 
mathematics. He is an author(co-author) of over 38 papers. 
Muhammet CANDAN received his M.Sc degree from Canakkale Onsekiz Mart University, Canakkale, 
Turkey, in 2012. He is currently Ph.D. candidate and working as a research assistant in Canakkale Onsekiz 
Mart University. His research interests cover applied mathematics, finite dynamical systems, graph theory.

Mais conteúdo relacionado

Mais procurados

Common derivatives integrals
Common derivatives integralsCommon derivatives integrals
Common derivatives integrals
Kavin Ruk
 
Calculo de integrais_indefinidos_com_aplicacao_das_proprie
Calculo de integrais_indefinidos_com_aplicacao_das_proprieCalculo de integrais_indefinidos_com_aplicacao_das_proprie
Calculo de integrais_indefinidos_com_aplicacao_das_proprie
Rigo Rodrigues
 
Solving the energy problem of helium final report
Solving the energy problem of helium final reportSolving the energy problem of helium final report
Solving the energy problem of helium final report
JamesMa54
 
Fast and efficient exact synthesis of single qubit unitaries generated by cli...
Fast and efficient exact synthesis of single qubit unitaries generated by cli...Fast and efficient exact synthesis of single qubit unitaries generated by cli...
Fast and efficient exact synthesis of single qubit unitaries generated by cli...
JamesMa54
 

Mais procurados (20)

SPSF04 - Euler and Runge-Kutta Methods
SPSF04 - Euler and Runge-Kutta MethodsSPSF04 - Euler and Runge-Kutta Methods
SPSF04 - Euler and Runge-Kutta Methods
 
Stochastic Hydrology Lecture 1: Introduction
Stochastic Hydrology Lecture 1: Introduction Stochastic Hydrology Lecture 1: Introduction
Stochastic Hydrology Lecture 1: Introduction
 
Calculus of variation problems
Calculus of variation   problemsCalculus of variation   problems
Calculus of variation problems
 
Singularities in the one control problem. S.I.S.S.A., Trieste August 16, 2007.
Singularities in the one control problem. S.I.S.S.A., Trieste August 16, 2007.Singularities in the one control problem. S.I.S.S.A., Trieste August 16, 2007.
Singularities in the one control problem. S.I.S.S.A., Trieste August 16, 2007.
 
Common derivatives integrals
Common derivatives integralsCommon derivatives integrals
Common derivatives integrals
 
Calculo de integrais_indefinidos_com_aplicacao_das_proprie
Calculo de integrais_indefinidos_com_aplicacao_das_proprieCalculo de integrais_indefinidos_com_aplicacao_das_proprie
Calculo de integrais_indefinidos_com_aplicacao_das_proprie
 
Mathcad explicit solution cubic equation examples
Mathcad   explicit solution cubic equation examplesMathcad   explicit solution cubic equation examples
Mathcad explicit solution cubic equation examples
 
Iit jam 2016 physics solutions BY Trajectoryeducation
Iit jam 2016 physics solutions BY TrajectoryeducationIit jam 2016 physics solutions BY Trajectoryeducation
Iit jam 2016 physics solutions BY Trajectoryeducation
 
IIT Jam math 2016 solutions BY Trajectoryeducation
IIT Jam math 2016 solutions BY TrajectoryeducationIIT Jam math 2016 solutions BY Trajectoryeducation
IIT Jam math 2016 solutions BY Trajectoryeducation
 
alt klausur
alt klausuralt klausur
alt klausur
 
Numerical Algorithm for a few Special Functions
Numerical Algorithm for a few Special FunctionsNumerical Algorithm for a few Special Functions
Numerical Algorithm for a few Special Functions
 
Formula m2
Formula m2Formula m2
Formula m2
 
MUMS Opening Workshop - Panel Discussion: Facts About Some Statisitcal Models...
MUMS Opening Workshop - Panel Discussion: Facts About Some Statisitcal Models...MUMS Opening Workshop - Panel Discussion: Facts About Some Statisitcal Models...
MUMS Opening Workshop - Panel Discussion: Facts About Some Statisitcal Models...
 
A Course in Fuzzy Systems and Control Matlab Chapter Three
A Course in Fuzzy Systems and Control Matlab Chapter ThreeA Course in Fuzzy Systems and Control Matlab Chapter Three
A Course in Fuzzy Systems and Control Matlab Chapter Three
 
Solving the energy problem of helium final report
Solving the energy problem of helium final reportSolving the energy problem of helium final report
Solving the energy problem of helium final report
 
Fast and efficient exact synthesis of single qubit unitaries generated by cli...
Fast and efficient exact synthesis of single qubit unitaries generated by cli...Fast and efficient exact synthesis of single qubit unitaries generated by cli...
Fast and efficient exact synthesis of single qubit unitaries generated by cli...
 
Convex Optimization Modelling with CVXOPT
Convex Optimization Modelling with CVXOPTConvex Optimization Modelling with CVXOPT
Convex Optimization Modelling with CVXOPT
 
Numerical integration based on the hyperfunction theory
Numerical integration based on the hyperfunction theoryNumerical integration based on the hyperfunction theory
Numerical integration based on the hyperfunction theory
 
Calculus of variations & solution manual russak
Calculus of variations & solution manual   russakCalculus of variations & solution manual   russak
Calculus of variations & solution manual russak
 
Solovay Kitaev theorem
Solovay Kitaev theoremSolovay Kitaev theorem
Solovay Kitaev theorem
 

Destaque

An intrusion detection algorithm for ami
An intrusion detection algorithm for amiAn intrusion detection algorithm for ami
An intrusion detection algorithm for ami
IJCI JOURNAL
 
Obamacare Health Spikes - Tallahassee Demcorat.PDF
Obamacare Health Spikes - Tallahassee Demcorat.PDFObamacare Health Spikes - Tallahassee Demcorat.PDF
Obamacare Health Spikes - Tallahassee Demcorat.PDF
Daniel Khan
 

Destaque (15)

Integral imaging three dimensional (3 d)
Integral imaging three dimensional (3 d)Integral imaging three dimensional (3 d)
Integral imaging three dimensional (3 d)
 
An intrusion detection algorithm for ami
An intrusion detection algorithm for amiAn intrusion detection algorithm for ami
An intrusion detection algorithm for ami
 
PREDICTING BANKRUPTCY USING MACHINE LEARNING ALGORITHMS
PREDICTING BANKRUPTCY USING MACHINE LEARNING ALGORITHMSPREDICTING BANKRUPTCY USING MACHINE LEARNING ALGORITHMS
PREDICTING BANKRUPTCY USING MACHINE LEARNING ALGORITHMS
 
Application of fuzzy logic for user
Application of fuzzy logic for userApplication of fuzzy logic for user
Application of fuzzy logic for user
 
Evaluation of models for predicting user’s next request in web usage mining
Evaluation of models for predicting user’s next request in web usage miningEvaluation of models for predicting user’s next request in web usage mining
Evaluation of models for predicting user’s next request in web usage mining
 
DATA MINING CLASSIFICATION ALGORITHMS FOR KIDNEY DISEASE PREDICTION
DATA MINING CLASSIFICATION ALGORITHMS FOR KIDNEY DISEASE PREDICTION DATA MINING CLASSIFICATION ALGORITHMS FOR KIDNEY DISEASE PREDICTION
DATA MINING CLASSIFICATION ALGORITHMS FOR KIDNEY DISEASE PREDICTION
 
DEVELOPMENT AND TESTING OF AN ENERGY HARVESTING TILE
DEVELOPMENT AND TESTING OF AN ENERGY HARVESTING TILEDEVELOPMENT AND TESTING OF AN ENERGY HARVESTING TILE
DEVELOPMENT AND TESTING OF AN ENERGY HARVESTING TILE
 
COMPARATIVE STUDY ON VEHICLE DETECTION TECHNIQUES IN AERIAL SURVEILLANCE
COMPARATIVE STUDY ON VEHICLE DETECTION TECHNIQUES IN AERIAL SURVEILLANCECOMPARATIVE STUDY ON VEHICLE DETECTION TECHNIQUES IN AERIAL SURVEILLANCE
COMPARATIVE STUDY ON VEHICLE DETECTION TECHNIQUES IN AERIAL SURVEILLANCE
 
Adopting level set theory based algorithms to segment human ear
Adopting level set theory based algorithms to segment human earAdopting level set theory based algorithms to segment human ear
Adopting level set theory based algorithms to segment human ear
 
Recommendation Letter
Recommendation LetterRecommendation Letter
Recommendation Letter
 
Steve
SteveSteve
Steve
 
Obamacare Health Spikes - Tallahassee Demcorat.PDF
Obamacare Health Spikes - Tallahassee Demcorat.PDFObamacare Health Spikes - Tallahassee Demcorat.PDF
Obamacare Health Spikes - Tallahassee Demcorat.PDF
 
RMD Profile_091016
RMD Profile_091016RMD Profile_091016
RMD Profile_091016
 
Creaciones técnicas leo2
Creaciones técnicas leo2Creaciones técnicas leo2
Creaciones técnicas leo2
 
excellqs
excellqsexcellqs
excellqs
 

Semelhante a Optimal control problem for processes

Natural and Clamped Cubic Splines
Natural and Clamped Cubic SplinesNatural and Clamped Cubic Splines
Natural and Clamped Cubic Splines
Mark Brandao
 

Semelhante a Optimal control problem for processes (20)

assignemts.pdf
assignemts.pdfassignemts.pdf
assignemts.pdf
 
Response Surface in Tensor Train format for Uncertainty Quantification
Response Surface in Tensor Train format for Uncertainty QuantificationResponse Surface in Tensor Train format for Uncertainty Quantification
Response Surface in Tensor Train format for Uncertainty Quantification
 
A common random fixed point theorem for rational inequality in hilbert space
A common random fixed point theorem for rational inequality in hilbert spaceA common random fixed point theorem for rational inequality in hilbert space
A common random fixed point theorem for rational inequality in hilbert space
 
On the Principle of Optimality for Linear Stochastic Dynamic System
On the Principle of Optimality for Linear Stochastic Dynamic System On the Principle of Optimality for Linear Stochastic Dynamic System
On the Principle of Optimality for Linear Stochastic Dynamic System
 
Maneuvering target track prediction model
Maneuvering target track prediction modelManeuvering target track prediction model
Maneuvering target track prediction model
 
An Approach For Solving Nonlinear Programming Problems
An Approach For Solving Nonlinear Programming ProblemsAn Approach For Solving Nonlinear Programming Problems
An Approach For Solving Nonlinear Programming Problems
 
On the principle of optimality for linear stochastic dynamic system
On the principle of optimality for linear stochastic dynamic systemOn the principle of optimality for linear stochastic dynamic system
On the principle of optimality for linear stochastic dynamic system
 
Exact Quasi-Classical Asymptoticbeyond Maslov Canonical Operator and Quantum ...
Exact Quasi-Classical Asymptoticbeyond Maslov Canonical Operator and Quantum ...Exact Quasi-Classical Asymptoticbeyond Maslov Canonical Operator and Quantum ...
Exact Quasi-Classical Asymptoticbeyond Maslov Canonical Operator and Quantum ...
 
Adaptive dynamic programming algorithm for uncertain nonlinear switched systems
Adaptive dynamic programming algorithm for uncertain nonlinear switched systemsAdaptive dynamic programming algorithm for uncertain nonlinear switched systems
Adaptive dynamic programming algorithm for uncertain nonlinear switched systems
 
Notions of equivalence for linear multivariable systems
Notions of equivalence for linear multivariable systemsNotions of equivalence for linear multivariable systems
Notions of equivalence for linear multivariable systems
 
paper acii sm.pdf
paper acii sm.pdfpaper acii sm.pdf
paper acii sm.pdf
 
Accurate Numerical Method for Singular Initial-Value Problems
Accurate Numerical Method for Singular Initial-Value ProblemsAccurate Numerical Method for Singular Initial-Value Problems
Accurate Numerical Method for Singular Initial-Value Problems
 
ACCURATE NUMERICAL METHOD FOR SINGULAR INITIAL-VALUE PROBLEMS
ACCURATE NUMERICAL METHOD FOR SINGULAR INITIAL-VALUE PROBLEMSACCURATE NUMERICAL METHOD FOR SINGULAR INITIAL-VALUE PROBLEMS
ACCURATE NUMERICAL METHOD FOR SINGULAR INITIAL-VALUE PROBLEMS
 
ACCURATE NUMERICAL METHOD FOR SINGULAR INITIAL-VALUE PROBLEMS
ACCURATE NUMERICAL METHOD FOR SINGULAR INITIAL-VALUE PROBLEMSACCURATE NUMERICAL METHOD FOR SINGULAR INITIAL-VALUE PROBLEMS
ACCURATE NUMERICAL METHOD FOR SINGULAR INITIAL-VALUE PROBLEMS
 
Accurate Numerical Method for Singular Initial-Value Problems
Accurate Numerical Method for Singular Initial-Value ProblemsAccurate Numerical Method for Singular Initial-Value Problems
Accurate Numerical Method for Singular Initial-Value Problems
 
SOLVING BVPs OF SINGULARLY PERTURBED DISCRETE SYSTEMS
SOLVING BVPs OF SINGULARLY PERTURBED DISCRETE SYSTEMSSOLVING BVPs OF SINGULARLY PERTURBED DISCRETE SYSTEMS
SOLVING BVPs OF SINGULARLY PERTURBED DISCRETE SYSTEMS
 
Natural and Clamped Cubic Splines
Natural and Clamped Cubic SplinesNatural and Clamped Cubic Splines
Natural and Clamped Cubic Splines
 
cheb_conf_aksenov.pdf
cheb_conf_aksenov.pdfcheb_conf_aksenov.pdf
cheb_conf_aksenov.pdf
 
Talk at SciCADE2013 about "Accelerated Multiple Precision ODE solver base on ...
Talk at SciCADE2013 about "Accelerated Multiple Precision ODE solver base on ...Talk at SciCADE2013 about "Accelerated Multiple Precision ODE solver base on ...
Talk at SciCADE2013 about "Accelerated Multiple Precision ODE solver base on ...
 
A New Nonlinear Reinforcement Scheme for Stochastic Learning Automata
A New Nonlinear Reinforcement Scheme for Stochastic Learning AutomataA New Nonlinear Reinforcement Scheme for Stochastic Learning Automata
A New Nonlinear Reinforcement Scheme for Stochastic Learning Automata
 

Último

IAC 2024 - IA Fast Track to Search Focused AI Solutions
IAC 2024 - IA Fast Track to Search Focused AI SolutionsIAC 2024 - IA Fast Track to Search Focused AI Solutions
IAC 2024 - IA Fast Track to Search Focused AI Solutions
Enterprise Knowledge
 
CNv6 Instructor Chapter 6 Quality of Service
CNv6 Instructor Chapter 6 Quality of ServiceCNv6 Instructor Chapter 6 Quality of Service
CNv6 Instructor Chapter 6 Quality of Service
giselly40
 

Último (20)

04-2024-HHUG-Sales-and-Marketing-Alignment.pptx
04-2024-HHUG-Sales-and-Marketing-Alignment.pptx04-2024-HHUG-Sales-and-Marketing-Alignment.pptx
04-2024-HHUG-Sales-and-Marketing-Alignment.pptx
 
Breaking the Kubernetes Kill Chain: Host Path Mount
Breaking the Kubernetes Kill Chain: Host Path MountBreaking the Kubernetes Kill Chain: Host Path Mount
Breaking the Kubernetes Kill Chain: Host Path Mount
 
Presentation on how to chat with PDF using ChatGPT code interpreter
Presentation on how to chat with PDF using ChatGPT code interpreterPresentation on how to chat with PDF using ChatGPT code interpreter
Presentation on how to chat with PDF using ChatGPT code interpreter
 
IAC 2024 - IA Fast Track to Search Focused AI Solutions
IAC 2024 - IA Fast Track to Search Focused AI SolutionsIAC 2024 - IA Fast Track to Search Focused AI Solutions
IAC 2024 - IA Fast Track to Search Focused AI Solutions
 
Bajaj Allianz Life Insurance Company - Insurer Innovation Award 2024
Bajaj Allianz Life Insurance Company - Insurer Innovation Award 2024Bajaj Allianz Life Insurance Company - Insurer Innovation Award 2024
Bajaj Allianz Life Insurance Company - Insurer Innovation Award 2024
 
Axa Assurance Maroc - Insurer Innovation Award 2024
Axa Assurance Maroc - Insurer Innovation Award 2024Axa Assurance Maroc - Insurer Innovation Award 2024
Axa Assurance Maroc - Insurer Innovation Award 2024
 
CNv6 Instructor Chapter 6 Quality of Service
CNv6 Instructor Chapter 6 Quality of ServiceCNv6 Instructor Chapter 6 Quality of Service
CNv6 Instructor Chapter 6 Quality of Service
 
What Are The Drone Anti-jamming Systems Technology?
What Are The Drone Anti-jamming Systems Technology?What Are The Drone Anti-jamming Systems Technology?
What Are The Drone Anti-jamming Systems Technology?
 
Real Time Object Detection Using Open CV
Real Time Object Detection Using Open CVReal Time Object Detection Using Open CV
Real Time Object Detection Using Open CV
 
Exploring the Future Potential of AI-Enabled Smartphone Processors
Exploring the Future Potential of AI-Enabled Smartphone ProcessorsExploring the Future Potential of AI-Enabled Smartphone Processors
Exploring the Future Potential of AI-Enabled Smartphone Processors
 
[2024]Digital Global Overview Report 2024 Meltwater.pdf
[2024]Digital Global Overview Report 2024 Meltwater.pdf[2024]Digital Global Overview Report 2024 Meltwater.pdf
[2024]Digital Global Overview Report 2024 Meltwater.pdf
 
Slack Application Development 101 Slides
Slack Application Development 101 SlidesSlack Application Development 101 Slides
Slack Application Development 101 Slides
 
08448380779 Call Girls In Friends Colony Women Seeking Men
08448380779 Call Girls In Friends Colony Women Seeking Men08448380779 Call Girls In Friends Colony Women Seeking Men
08448380779 Call Girls In Friends Colony Women Seeking Men
 
GenCyber Cyber Security Day Presentation
GenCyber Cyber Security Day PresentationGenCyber Cyber Security Day Presentation
GenCyber Cyber Security Day Presentation
 
Powerful Google developer tools for immediate impact! (2023-24 C)
Powerful Google developer tools for immediate impact! (2023-24 C)Powerful Google developer tools for immediate impact! (2023-24 C)
Powerful Google developer tools for immediate impact! (2023-24 C)
 
Boost Fertility New Invention Ups Success Rates.pdf
Boost Fertility New Invention Ups Success Rates.pdfBoost Fertility New Invention Ups Success Rates.pdf
Boost Fertility New Invention Ups Success Rates.pdf
 
08448380779 Call Girls In Diplomatic Enclave Women Seeking Men
08448380779 Call Girls In Diplomatic Enclave Women Seeking Men08448380779 Call Girls In Diplomatic Enclave Women Seeking Men
08448380779 Call Girls In Diplomatic Enclave Women Seeking Men
 
A Domino Admins Adventures (Engage 2024)
A Domino Admins Adventures (Engage 2024)A Domino Admins Adventures (Engage 2024)
A Domino Admins Adventures (Engage 2024)
 
2024: Domino Containers - The Next Step. News from the Domino Container commu...
2024: Domino Containers - The Next Step. News from the Domino Container commu...2024: Domino Containers - The Next Step. News from the Domino Container commu...
2024: Domino Containers - The Next Step. News from the Domino Container commu...
 
08448380779 Call Girls In Civil Lines Women Seeking Men
08448380779 Call Girls In Civil Lines Women Seeking Men08448380779 Call Girls In Civil Lines Women Seeking Men
08448380779 Call Girls In Civil Lines Women Seeking Men
 

Optimal control problem for processes

  • 1. International Journal on Cybernetics & Informatics (IJCI) Vol. 3, No. 4, August 2014 OPTIMAL CONTROL PROBLEM FOR PROCESSES REPRESENTED BY STOCHASTIC SEQUENTIAL MACHINE Yakup H. HACI and Muhammet CANDAN Department of Mathematics, Canakkale Onsekiz Mart University, Canakkale, Turkey ABSTRACT In this paper, optimal control problem for processes represented by stochastic sequential machine is analyzed. Principle of optimality is proven for the considered problem. Then by using method of dynamical programming, solution of optimal control problem is found. Keywords Optimal control problem, stochastic sequential machine, dynamical programming. 1. INTRODUCTION Stochastic sequential machine (SSM) is the one of the most developing field of discrete system theory [2], [3]. It plays an important role in many areas such as construction of finite dynamical system, imitation modelling problem, coding of discrete systems and identification problems. Thus, it points out that it requires a comprehensive research. 2. SSM SSM is generalization of multi-parametric finite sequential machine[3] but it contains probability variable. General form of this system is defined by[6]: K =< X , S,Y, s , p( ), F (.),G > 0 n w v =1,2,...k (1) where X = [ GF(2) ] r , S = [ GF(2) ]m and Y = [ GF(2) ]q are input, state and output index (alphabet) respectively; s 0 is initial state vector, p(w) is determitistic discrete probability distribution ( W = { w , w ,..., w } is finite set, p ( w ) = { p ( w ) : w ÎW, p ( w ) = 1}) , 1 2 p i i i w ÎW i characteristic Boolean vector functions [7] denoted by ( × ) = { ( × ),..., ( × )} F F Fn n n which are also 1 m known as transfer functions are nonlinear functions defined on the set [ ] [ ]k m r Z × GF(2) × GF(2) and G are an output characteristic functions defined on GF(2) where GF(2) is Galois field[8] and the symbol (.) denotes (c, s(c), x(c)) for simplicity. In addition to the definition, SSM is represented by: s(c e ) F (c, s(c), x(c),w(c)) n n + = v = 1,2,...k (2) DOI: 10.5121/ijci.2014.3403 21
  • 2. International Journal on Cybernetics Informatics (IJCI) Vol. 3, No. 4, August 2014 )y(c) = G(s(c ) (3) where s(c), x(c), y(c) are m, r and q dimensional state, input and output vectors at the point c respectively, w(c) is a random variable [6], c {c c Z c c c c c c c Z} i 22 = Î k , £ £ L ,..., 0 £ £ L k , Î k k k 1 1 0 1 1 is point in k Z , determining position i L , ( i = 1,2,...k positive integer) is the duration of the stage n n e = . i of this process. Z is set of integers and (0,...,0,1,0,...,0) In SSM, each random variable w has a special case. For instance, w is an input variable in the identification problem of the SSM. However, w is a set of all possible states in the synthesis of optimal sequential machine Moreover, the state of the system depends on the random variable w which affects not only parameters of the SSM but also the input variable. Finally, equation (2) is transformed to: s(c e ) F (c, s(c), x(c) w(c)) n n + = Å v = 1,2,..., k (4) where symbol Åmeans that x Å w is always in input alphabet X . The discrete optimal processes given by SSM are characterized by functional: ( ) { ( ( ))} __ L J x M f s c w = (5) where {.} w M is a mathematical expected value of w . 3. OPTIMAL CONTROL PROBLEM AND PRINCIPLE OF OPTIMALITY We can state optimal control problem [2] for processes represented by SSM as below: In order for given SSM to start from the known inital state 0 s to go any desired state ( ) * L s c , to which we expect to access in L steps, a control x(c)Î X [4] must exist such that the functional in (1) has a minimal value: d s(c + e ) = Fˆ (c, s(c), x(c),w(c)),cÎG n n v =1,2,...k (6) 0 0 s(c ) = s , x(c)Î X , d cÎG (7) ˆ ( , ˆ ( , ( ), ( )), ( ), ( )) ˆ ( ˆ ( , ( ), ( )), ( ) ( ) n μ μ μ μ μ n n n n F c + e F c s c x c x c + e w c + e = F c + e F c s c x c x c + e w c + e (8) ( ) { ( ( )} min __ = ® L J x M f s c w (9) where ˆ (.) n F (n = 1,..., k) denotes the pseudo Boolean expression of the Boolean vector ^ F n = k and k L = L + L +...+ L 1 2 is the time duration of this process. function[7] (.)( 1,..., ) It is well-known that method of dynamic programming [5] is used for solution of optimal control problem. If we make use of this method to solve the considered problem then, (6)-(9) can be formulate as an optimal problem: ( ) ˆ ( , ( ), ( ),w( )), (s ) n n d s c + e = F c s c x c c cÎG (10)
  • 3. International Journal on Cybernetics Informatics (IJCI) Vol. 3, No. 4, August 2014 23 s(s ) =À (11) ( ) , (s ) d x c Î X cÎG (12) ˆ ( , ˆ ( , ( ), ( ), ( )), ( ), ( )) ˆ ( , ˆ ( , ( ), ( ), ( )), ( ), ( )) n μ μ μ μ μ n n n n F c+e F c s c x c w c x c+e w c+e =F c+e F c s c x c w c x c+e w c+e (13) ( ) { ( ( ))} min __ = ® L J x M f s c w (14) where À is an arbitrary element in S . As it can be seen from (10)-(14), if we substitute 0 s = c or 0 À= s into problem (10)-(14) we obtain first problem stated above. If the conditions for existence of unique solution are satisfied then for the given initial condition s(s ) =Àand given ( )( (s )), d x c cÎG we find a unique s(c) . That is, the functional (14) is the function of the parameters À and ( )( (s )) d x c cÎG : __ __ J ( x ) = J ( À , x ( G ( s )) (15) d where ( (s )) d x G denotes the range of the control x(c) on the points (s ) d cÎG : ( (s )) { ( ), (s )} d d x G = x c cÎG (16) from the unique solution condition of the system (6), we find that the stochastic process can be investigated in the set (s ) d G and also in the set { } d k k k G s = c c £ c s c £ c s 0 ( ) ; 0 ,..., 1 1 1 1 (17) Definition. We say that the control ( ), ( (s )) d x c cÎG which minimizes the functional (5) in the problem (10)-(14) is optimal control with respect to the initial pair (s ,À) on the region (s ) d G Suppose that ( ) 0 x c is an optimal control with respect to the initial pair ( , ) 0 0 c s on the region d G and ( ) 0 s c is admissible optimal trajectory. Then ( ) 0 x c is an optimal control with respect to the initial pair ( , ( )) 0 s s s on the region (s ) d G for every d G . Proof. Suppose the contrary. Then there exists ( ) , (s ) d x c Î X cÎG such that we have __ __ ( , ( ( )) ( , ( ( )) 0 s s d d J À x G J À x G (18) We choose a new control process d ~x(c), cÎG as follows: Î s ( ), ( ) 0 x c c G ~( ) 1 x c (19) Î = s d ( ), ( ) x c c G d As it can be seen, (19) is an admissible control such that
  • 4. International Journal on Cybernetics Informatics (IJCI) Vol. 3, No. 4, August 2014 24 ( , ~( )) ( , ~( ( ) ( )). 1 0 __ 0 __ s s d d d J s x G = J s x G ÈG (20) According to the condition, ( ) =À 0 s s . Thus we have __ ( , ~( ( ) È ( )) = ( ( ), ~( ( )) = 0 J s 0 x G s G s J s s x G s = (À, ( ( ))) (À, 0 ( ( ))) = d d d __ 1 __ __ s s d d J x G J x G __ ( ( ), 0 ( ( ))) ( 0 , 0 ( )) __ d d = J s s x G s = J s x G (21) and by virtue of (2.10) and (2.11) we can obtain __ ( 0 , ~( )) ( 0 , 0 ( )) __ d d J s x G J s x G (22) Hence, (22) contradicts the hypothesis that the control d x (c),cÎG 0 is optimal. This completes the proof of the theorem. Let a function (for every fixed s and À) be corresponding to the optimal value of pseudo Boolean functional in the problem (10)-(14): ( , ) min { ( ( ))} L B s M f s c w À = . (23) where minimization on the set of admissible control ( ), (s ) d x c cÎG . Now, we determine method of dynamical programming(It is known as Bellman equation) [5] for B(s ,À) . Suppose that d x (c),cÎG 0 is the admissible control corresponding to (10)-(14) with initial condition and ( ), ( ) 0 s d s c cÎG is also the optimal trajectory. Let the point s G ( )(v 1,2,...,k) d x Î s = n and any element y(c)Î X be specified. If x(s ) = y(c) , then the state of the system in the point x s n is determined by (x s ) ˆ (s , , ,w(s )) n n s = F À y (24) We consider the following problem: x ( ) ˆ ( , ( ), ( ),w( )), (x s ) n n d n s c = F c s c x c c cÎG (25) s(x s ) = Fˆ (s ,À, y) n n (26) ( ) ( ), (x s ) d n x c Î X c cÎG (27) ( ) { ( ( ))} min __ = ® L J x M f s c w (28) If ˆ( ), (x s ) d n y c cÎG and ˆ( ), (x s ) d n s c cÎG are optimal control and optimal trajectory respectively, then
  • 5. International Journal on Cybernetics Informatics (IJCI) Vol. 3, No. 4, August 2014 {f ( ( ))} (x s , ˆ (s , , ˆ,w(s ))) w n n M s c B F y L = À (29) x c (30) 25 can be found. For (10)-(14), let ~x (c) be an admissible control below. = Î = , y c ˆ( ), ( ) ~( ) x s s d n y c c G Also, ~s (c) can be obtained by ˆ ( , , , ( )), s À w s = s F y c s c (31) Î = ˆ( ) , ( ) ~( ) x s n n d s c c G __ It is evident that the value of ( ) { ( ( ))} L J x M f s c w = to control ~x (c) is determined by { (~( ))}= { ( ˆ( ))}= L L M f s c M f s c w w (x s , ˆ (s , , ˆ,w(s ))) n n = B F À y . (32) Since ~( ), (s ) d x c cÎG is not largely optimal control, we can state { (~( ))} { ( ( ))} ( , ) 0 f ³ f = s À w w M s c M s c B L L (33) Thus, we have (s , ) (x s , ˆ (s , , ˆ,w(s ))) n n B À £ B F À y (34) On the other hand, if ( ) ( ), 0 y c = x s then by the principle of optimality [2], ˆ( )( ( )) ( )( ( )) 0 x s x s d n d n y c cÎG = x c c ÎG (35) Therefore, ( , ) ( , ˆ ( , , ( ), ( ))) 0 s x s s s w s n n B À = B F À x (36) By (34) and (35), Bellman equation[5] can be determined by ( , ) min ( , ˆ ( , , ˆ, ( ))), s x s s w s n n B B F y S y X ÀÎ À = À Î s (37) ( )
  • 6. International Journal on Cybernetics Informatics (IJCI) Vol. 3, No. 4, August 2014 26 4. CONCLUSIONS It is shown that Bellman equation for optimal processes with stochastic sequential machine is obtained and the principle of optimality is proven. REFERENCES [1] Anderson, J. A., (2004) Discrete Mathematics witk Combinatorics, Prentice Hall, New Jersey, p.45 “This is my paper”, ABC Transactions on ECE, Vol. 10, No. 5, pp120-122. [2] Boltyanskii, V. G., (1978) Optimal Control of Discrete Systems, John Willey, New York, p.363. [3] Gaishun, I. V., (1983) Completely Solvable Multidimensional Differential Equations, Nauka and Tekhnika, Minsk, p.231. [4] Hacı, Y., Ozen, K., (2009) “Terminal Control Problem for Processes Represented by Nonlinear Multi Binary Dynamic System”, Control and Cybernetics, Vol. 38, No. 3, pp625-633. [5] Bellman, R., (1957) Dynamic Programming, Princeton University Press, Princeton, p.12 [6] Yermolyev, Y. M., (1976) Stochastic Programming Methods, Nauka (in Russian), p.240. [7] Yablonsky, S. V., (1989) Introduction to Discrete Mathematics, Mir Publishers, Moscow, p.9. [8] Fraileigh, J. B., (1998) A First Course in Abstract Algebra, 6th ed. Addison--Wesley Publisher, p.419-423 Authors Yakup H. HACI received Ph.D at physics and mathematics in Azerbaijan NationalAcademy of Sciences at the Institute of Cybernetics. He is currently working as professor and head of department of Mathematics, in Canakkale Onsekiz Mart University. His primary research interests are in the areas of the theory of multiparametric binary sequential machines,linear and nonlinear optimization and discrete mathematics. He is an author(co-author) of over 38 papers. Muhammet CANDAN received his M.Sc degree from Canakkale Onsekiz Mart University, Canakkale, Turkey, in 2012. He is currently Ph.D. candidate and working as a research assistant in Canakkale Onsekiz Mart University. His research interests cover applied mathematics, finite dynamical systems, graph theory.