Académique Documents
Professionnel Documents
Culture Documents
ABSTRACT: Recently, the field of differential equations has been studying in a very abstract method.
Instead of considering the behaviour of one solution of a differential equation, one studies its sheaf-solution
(see[10]). Instead of studying a differential equation, one studies differential inclusion (see[9]). Especially, one
studies fuzzy differential equation ( a differential equation whose variables and derivative are fuzzy sets, see[1-
7]).
In this paper, a set differential equation is generalized to be Set Control Differential Equation (SCDE)
and we present the global controlable (GC) for set control differential equations (SCDE), some examples for
the global controlable (GC) and non – controlable for set control differential equations (SCDE). This paper
is a continuation of some works in this direction (see [10-18]).
1. INTRODUCTION
Before we proceed to investigate problem of Global Controlable for set control
differential equations, let’s note the following facts:
In [5], Prof. V. Lakshmikantham and the other authors considered set differential
equations (SDE) and had some important results on existence, comparison and stability
criteria for SDE
DH X(t) = F(t, X(t)) , (1.1)
1
Corresponding author: Email address: ndphu_dhtn@yahoo.com.vn
1
for the global controlable (GC) and non – controlable for set control differential equations
(SCDE). In section 5, we give conclusion of this paper.
2. PRELIMINARIES
We recall some notations and concepts presented in detail in recent series works of V.
Lakshmikantham et al. See [5].
Let KC (R n ) denote the collection of all nonempty, compact and convex subsets of R n .
Given A, B in KC (R n ) , the Hausdorff distance between A and B defined as
{
D [ A, B] = max supa∈A infb∈B a − b , supb∈B infa∈A a − b , } (2.1)
If α, β ∈ R and A, B ∈ KC (R n ) , then
α(A + B) = αA + αB, α(β A) = (αβ)A, 1A = A (2.3)
and if α, β ∈ R + , then (α + β)A = αA + β A .
It is known that ( KC (R n ) , D) is a complete metric space and if the space KC( R n ) is
equipped with the natural algebraic operations of addition and nonnegative scalar
multiplication, then KC( R n ) becomes a semilinear metric space which can be embedded as
a complete cone into a corresponding Banach space.
Let A, B ∈ Kc( R n ) . The set C ∈ Kc( R n ) satisfying A = B + C is called the Hausdorff
difference (the geometric difference) of the sets A and B and is denoted by the symbol
A − B . Given an interval I in R+ . We say that the mapping F : I → Kc( R n ) has a
Hukuhara derivative DH F( t0 ) at a point t0 ∈ I , if
F( t0 + h ) − F( t0 ) F( t0 ) − F( t0 − h )
lim and lim (2.4)
h →0 +
h h →0 +
h
exist in the topology of KC( R n ) and are equal to DH F( t0 ) .
2
By embeding KC( R n ) as a complete cone in a corresponding Banach space and taking
into account th result on the differentiation of Bochner integral, we find that if
t
F( t) = X 0 + ∫ Φ( s) ds, X 0 ∈ KC( R n ), (2.5)
t0
where Φ : I → KC( R n ) is integrable in the sense of Bochner, then DH F( t) exists and the
equality
DH F( t) = Φ( t), a.e on I holds.
The Hukuhara integral of F is given by
⎡ ⎤
∫I F( s ) ds = ⎢
⎣I
∫ f( s ) ds : f is a continuous selector of F ⎥
⎦
for any compact set I ⊂ R + .
Some properties of the Hukuhara integral are in [4-7].
If F : I → Kc( R n ) is integrable, one has
t2 t1 t2
and
t t
⎣ ⎦ { }
Let us denote D ⎡A, Θn ⎤ = A = sup a : a ∈ A for A ∈ Kc( R n ) (2.9)
, where Θn ∈ K(R n ) is the zero element of R n which is regarded as a one point set.
3. MAIN RESULTS
We consider a set control differential equation (SCDE):
DH X(t) = F(t, X(t), U(t)) (3.1),
X(t0 ) = X 0 ∈ KC (R n )
Where F ∈ C[I × KC (R n ) × KC (R n ), KC (R n )] , t ∈ I ⊂ R + , U(t) ∈ KC (R p ) , whose set
solution is form X(t) = X(t0 , X 0 , t, U(t)) ∈ KC (R n )
Definition 3.1. A state pair of set solutions (X 0 , X1 ) ∈ KC (R n ) will be a controlable if after
time t1 we shall find a set control U(t) ∈ KC (R p ) such that
X(t1 ) = X(t0 , X 0 , t1, U(t1 )) = X1 ∈ KC (R n ) (3.2).
Definition 3.2. The set control differential equation (SCDE) (3.1) is said to be
3
(GC): Global controlable if every state pair of set solutions (X 0 , X1 ) ∈ KC (R n ) will be a
controlable (3.3).
(GR): Global achieve if every X1 ∈ KC (R n ) we have state pair of set solutions
(Θn , X1 ) ∈ KC (R n ) will be a controlable (3.4).
(GNC): Global achieve to Θn ∈ KC (R n ) if every X1 ∈ KC (R n ) we have state pair of set
solutions (X1, Θn ) ∈ KC (R n ) will be a controlable (3.5).
Suppose that, the set Global achieve GR from X 0 ∈ KC (R n ) after time t:
{
R t (X 0 ) = X(t) ∈ KC (R n ) | ∃U(t) ∈ KC (R p ) : X(t0 , X 0 , t, U(t)) = X(t)} (3.6).
(ii) R(X 0 ) = ∪ R t (X 0 ) (3.7).
t>0
t
If we put L t =
t0
∫ W (t, s)W (t , s)ds
0 0 1
(3.12),
t1
then we have L t =
1
t0
∫ W (t, s)W (t , s)ds
0 0 1
(3.13).
4
Theorem 3.2.
The stationary linear set control differential equation (SLSCDE) (3.10) is (GC) iff
exist B−1 & L−t1 (3.14).
1
Proof.
t
a/ Sufficient condition: By L t = ∫ W (t, s)W (t , s)ds , for every
t0
0 0 1
t1 > t0 we find the control
U(t) ∈ KC (R p ) in type:
U(t) = −B−1W0 (t, s)L−t1[W0 (t, t0 )X 0 − X1 ] (3.15).
1
{
= W0 (t1, t0 )X 0 + ∫ W0 (t, s)B −B−1W0 (t1, s)L−t1[W0 (t1, t0 )X 0 − X1 ] ds
t0
1
}
t1
given operator B has not B that means B is decline then SLSCDE (3.10) isn’t controlabe. If
−1
t t
X(t) = X(t0 , X 0 , t, U(t)) = W0 (t, t0 )X 0 + ∫ W0 (t, s)BU(s)ds + ∫ R(s, X(s),U(s))ds (3.18)
t0 t0
5
t
If we put L t = ∫ W (t, s)W (t , s)ds
t0
0 0 1
(3.19)
t1
then L t =
1 ∫ W (t, s)W (t , s)ds
t0
0 0 1
(3.20).
t t
X(t) = X(t0 , X 0 , t, U(t)) = W1(t, t0 )X 0 + ∫ W1(t, s)B(s)U(s)ds + ∫ R(s, X(s), U(s))ds
t0 t0
(3.21).
R(t, X, U)
Let lim =0 (3.22).
X →0 X
Theorem 3.3.
The stationary linear set control differential equation with disturb (USLSCDED) (3.17)
with (3.22) is (GC) iff exist B−1 & L−t1 . (3.23).
1
t
If we put Λ t = ∫ W (t, s)W (t , s)ds
t0
1 1 1
(3.26),
t1
then Λ t =
1 ∫ W (t, s)W (t , s)ds
t0
1 1 1
(3.27).
Theorem 3.4.
The Unstationary Linear Set Control Differential Equation (SLSCDE) (3.24) is (GC)
iff exist B−1(s)& Λ −t 1 (3.28).
1
6
Proof .
t
a/ Sufficient condition: By Λ t = ∫ W (t, s)W (t , s)ds , for every
t0
1 1 1
t1 > t0 we find the control
U(t) ∈ KC (R p ) in type:
U(t) = −B−1(s)W1(t, s)Λ −t 1[W1(t, t0 )X 0 − X1 ] (3.29).
1
{
= W1(t1, t0 )X 0 + ∫ W1(t, s)B(s) −B−1(s)W1(t1, s)Λ −t 1[W1(t1, t0 )X 0 − X1 ] ds
t0
1
}
t1
= W1(t1, t0 )X 0 − Λ t Λ t−1[W1(t1, t0 )X 0 − X1 ]
1 1
given operator B(t) has not B (t) that means B(t) is decline then USLSCDE (3.24) isn’t
−1
(3.32).
7
R(t, X, U)
Let lim =0 (3.33).
X →0 X
Theorem 3.5.
The Unstationary Linear Set Control Differential Equation with Disturb (USLSCDED)
(3.31) with (3.33) is (GC) iff exist B−1(s)& Λ −t 1 (3.34).
1
Lemma 3.1: Suppose that, the Lyapunov-like function V(t, X(t)(t)) (see {…}) satisfies the
following:
i/ V ∈ C(R + × KC (R n ) × KC (R p ), KC (R n )) ,
ii/ V(t, X1, U1 ) − V(t, X2 , U2 ) ≤ L(D[X1, X2 ] + D[U1, U2 ]) ,
iii/ D+ V(t, X, U) ≤ 0 ,
iv/ b( X ) ≤ V(t, X, U) ≤ a(t, X )
for t ∈ R + , X1, X2 , X ∈ KC (R n ), U1, U2 , U ∈ KC (R p ) ,
where b(.), a(t,.) ∈ K = {σ ∈ C[R + , R + ] : σ(ϖ) is increasing in ϖ and σ(ϖ) → ∞ as
ϖ → ∞ } then for every α > 0, t0 ∈ R + , ∃β = β(t, α) > 0 such that X 0 < α implies
X(t) < β , ∀t > t0 . (3.36)
Proof.
Let α > 0, t0 ∈ R + be given.
We can choose β = β(t, α) > 0 such that a(t0 , α) < b(β) (3.37),
8
then (3.37) is hold. If this is not true, there would exists
t
X(t) = X(t0 , X 0 , t, U(t)) = X 0 + ∫ F(s, X(s), U(s))ds - set solution of (NLSCDE) (3.35) and
t0
such that ∃t1 > t0 such that X(t1 ) = β . Because the Lyapunov – like function V(t, X(t)(t))
is decreas then V(t, X(t), U(t)) ≤ V(t1, X(t1 ), U(t1 )) , ∀t > t1 or X(t) ≤ β ,as results,
condition iv/ of lemma 3.1 and (3.37) yield
( )
b(β) ≤ b X(t1 ) ≤ V(t1, X(t1 ), U(t1 )) ≤ V(t0 , X(t0 ), U(t0 )) ≤ a(t0 , X 0 ) < a(t0 , α) < b(β)
This contradiction proves (3.36). (, )
Theorem 3.6.
Suppose that for nonlinear set control differential equation (NLSCDE) (3.35) satisfy the
followings:
i/ The Lyapunov-like function V(t, X(t)(t)) satisfies the conditions of lemma 3.1,
ii/ F ∈ C[I × KC (R n ) × KC (R n ), KC (R n )] with: F(t, X(t), U(t) ≤ L( X + U ) , L ∈ R +
iii/ For every pair of set states (X 0 , X1 ) in KC (R n ) , ∀t1 > t0 ∃U(t) ∈ KC (R p ) such that
t1
X1 exp[ −L(t1 − t0 ) − X 0
∫
t0
U(s) ds =
L
(3.38),
Proof.
Before we choose the controlable set:
t
Because the set solution of (3.35) is X(t) = X(t0 , X 0 , t, U(t)) = X 0 + ∫ F(s, X(s), U(s))ds -
t0
t1
then X(t1 ) = X(t0 , X 0 , t1, U(t1 )) = X 0 + ∫ F(s, X(s), U(s))ds implies that
t0
t1 t1
X(t1 ) ≤ X 0 + ∫
t0
F(s, X(s), U(s)) ds ≤ X 0 + ∫ L( X(s) + U(s) )ds
t0
t1 t1
t1
X1 exp[ −L(t1 − t0 ) − X 0
Replaceing ∫
t0
U(s) ds =
L
from (3.38) into (3.39) we received
X(t1 ) ≤ X1 .
We must prove that X(t1 ) = X1 . If this is not true, that if X(t1 ) < X1 . But in this
case we have by lemma 3.1 , X(t1 ) = β implies:
9
( )
b(β) ≤ b X(t1 ) ≤ V(t1, X(t1 ), U(t1 )) ≤ V(t1, X1, U(t1 )) ≤ a(t1, X1 ) < a(t1, β) < b(β)
Proof.
Before we choose the controlable set:
t
Because the set solution of (3.35) is X(t) = X(t0 , X 0 , t, U(t)) = X 0 + ∫ F(s, X(s), U(s))ds -
t0
t1
then X(t1 ) = X(t0 , X 0 , t1, U(t1 )) = X 0 + ∫ F(s, X(s), U(s))ds implies that
t0
t1 t1
X(t1 ) ≤ X 0 + ∫
t0
F(s, X(s), U(s)) ds ≤ X 0 + ∫ L( X(s) + U(s) )ds
t0
t1 t1
t1
X1 exp[ −L(t1 − t0 ) − X 0
Replaceing ∫ λ(s)ds =
t0 L X1
from (3.40) into (3.41) we received
X(t1 ) ≤ X1 .
We must prove that X(t1 ) = X1 . If this is not true, that if X(t1 ) < X1 . But in this
case we have by lemma 3.1 , X(t1 ) = β implies:
( )
b(β) ≤ b X(t1 ) ≤ V(t1, X(t1 ), U(t1 )) ≤ V(t1, X1, U(t1 )) ≤ a(t1, X1 ) < a(t1, β) < b(β)
Corollary 3.1.
10
i/ If approcimation of F(t, X(t), U(t)) ≈ AX + BU + R(t, X, U) then problem of GC of
NLSCDE (3.35) will be a problem of GC of USLSCDED (3.17).
ii/ If approcimation of F(t, X(t), U(t)) ≈ A(t)X + B(t)U + R(t, X, U) then problem of GC
of NLSCDE (3.35) will be a problem of GC of USLSCDED (3.31).
4. APPLICATIONS
We present some examples for two cases: Global controlable and non controlable
A. Global controlable case : The Oil sp on the see. For every pair of set states (X 0 , X1 ) in
KC (R n ) , ∀t1 > t0 ∃U(t) ∈ KC (R p ) such that U(t) = λ(t)X1 where
t1
X1 exp[ −L(t1 − t0 ) − X 0
∫
t0
λ(s)ds =
L X1
(3.40),
5. CONCLUSION
In this paper we give a concept of Problem of Global Control for Set Control Differential
Equation (SCDE) and study its many kinds of feeback controls (3.15), (3.29) and (3.38).
Some more results on new prived derivative and integral in Haussdorff metric space and
there applications shall be presented in next works.
Acknowledgements
I wish to thank the refree for his (her) carefull reading and valuable remarks which
improve the presentation of the paper.
11
References
[1] Kaleva, O., Fuzzy differential equations,
J. Fuzzy Sets and Systems, 24(1987), pp.301-317.
[2] Kaleva, O., Fuzzy differential equations,
J. Fuzzy Sets and Systems, 35(1990), pp. 389-396.
[3] Kaleva, O., Fuzzy differential equations,
J. Fuzzy Sets and Systems, 24(1987), pp. 301-317.
[4] Lakshmikantham V., Set differential equations versus fuzzy differential equations,
J. Applied Mathematics and Computation 164 (2005) ,pp. 277-294.
[5] Lakshmikantham V., Gnana Bhaskar T., Vasundhara Devi J., Theory of set differential
equations in metric spaces,
Cambridge Scientific Publisher, UK, 2006.
[6] Lakshmikantham V., Mohapatra R., Theory of fuzzy differential equations and inclusions,
Taylor & Francis, London, 2003.
[7] LakshmikanthamV., Leela S., Fuzzy differential systems and the new concept of stability,
J. Nonlinear Dynamics and Systems Theory, 1(2) (2001),pp. 111-119.
[8] Phu N. D., Genaral views in theory of systems,
VNU – Publishing House, HCM City, 2003.
[9] Phu N.D., Huong N.T., Multivalued Differential Equations,
VNU – Publishing House, HCM City, 2005.
[10] Phu N. D., Tung T.T., Sheaf optimal control problems in fuzzy type,
J. Science and Technology Development 8 (12) (2005), pp 5-11.
[11] Phu N. D., Tung T.T., The comparison of sheaf- solutions in fuzzy control problems,
J. Science and Technology Development 9 (2) (2006), pp 5-10.
[12] Phu N. D., Tung T.T., Some properties of sheaf-solutions of sheaf fuzzy control
problems.
E. Journal of Differential Equations Vol (2006), N. 108, pp 1-8.
[13] Phu N. D., Tung T.T., Some results on sheaf solutions of sheaf set control problems.
J. Nonlinear Analysis,Vol 9 (2007), pp 1309 – 1315.
[14] Phu N. D.,Tung T.T., Existence of solutions of fuzzy control differential equations.
J. Science and Technology Development 10 (5) (2007), pp 5-12.
[15] Phu N. D., Tung T.T., Existence of solutions of set control differential equations, J.
Science and Technology Development 10 (6) (2007), pp 5-14.
[16] Phu N. D.,Tung T.T., Some new results on the fuzzy control differential equations .
J. Science and Technology Development 11 (1) (2008), pp 5-20.
[17] Phu N. D., Quang L.T, Tung T.T., Stability criteria for some set control
differential equations,
J. Nonlinear Analysis, 69(2008), pp 3715-3721.
[18] Phu N. D., Quang L.T, Tung T.T., Criteria for boundedness of solutios for set
control differential equations,
J. Evolution Equations ( on Line).
12
13