• Tidak ada hasil yang ditemukan

10_Catalin Mitran. 145KB Jun 04 2011 12:07:22 AM

N/A
N/A
Protected

Academic year: 2017

Membagikan "10_Catalin Mitran. 145KB Jun 04 2011 12:07:22 AM"

Copied!
6
0
0

Teks penuh

(1)

ON SOME ASPECTS REGARDING THE

GENERALIZATION OF THE OPTIMAL FORMULAS OF

SARD TYPE USING THE BOOLEAN-SUM TYPE

OPERATORS

by

C

ă

t

ă

lin Mitran

Abstract. In this paper will be presented some results about the optimal quadrature’s formulas of Sard type and there, starting from considerations, a generalization for the cubature’s formulas of boolean-sum type builts with the help of some operators who satisfy the optimal formulas mentioned above. A better generalization can be obtained using a boolean-sum type formula with a number of n operators and n corresponding remainders.

We shall consider the next quadrature’ formula:

( ) ( )

( )

( )

=

= +

b

a

m

k

m k

kf a R f

A x

d x f

0

λ (1),

where dλ is a positive measure of the integration integral ,

a

k are the nodes of the quadrature’s formula and satisfy the relations:

b a a

a

a0 < 1<...< m≤ ,

k

A are named the coefficients of the formula and Rm

( )

f are named the remainder

Theorem 1

If

f

H

n

[ ]

a

,

b

if we use the Peano’s theorem we shall have the next representation of the remainder:

( )

=

( ) ( )

b

a

n n m

m f K t f t dt

R , ( ) (2),

where

( )

( )

( )

( ) ( )

( )

(

)

    

  

− −

− ⋅ − =     

  

− −

= −

+ =

− + −

+

1

0 1

1

,

! 1 1 !

1

n k m

k k b

a n n

m n

m x t d x A a t

n n

t x R t

K λ

(2)

Remark 1

If the Peano’s kernel has a constant sign on the entire interval of the integration then using a midle formula we shall can write:

[ ]

( )

( )

( )

. b a

, dt t K f

f R

b

a n , m m

m

< <

=

ξ

ξ

(3)

Definition 1

The quadrature’s formula given by (1) will be named optimal of Sard type if:

( )

b

a n m t dt

K , 2 min (4) Theorem 2

If

n

m

the quadrature’s formula given by (1) who is considerated optimal of Sard type has an one solution and this solution will be obtained by the integration of the spline interpolation formula

( ) ( )( )

x

Sf

x

R

( )

f

x

x

[ ]

a

b

f

=

+

m

,

,

,

that means

( ) ( ) ( )( ) ( )

b =

a

b

a

x d x Sf x

d x

f

λ

λ

+

( ) ( )

b

a

m f x d x

R ,

λ

(5).

We shall consider now some questions about the generalization of the optimal quadrature’s formulas of Sard type, more precisely about the cubature formulas of boolean-sum type who results from optimals quadrature’s formulas of Sard type.

Let’s consider the next decomposition of the unit operator:

(

1 2 1 2

)

1 2

2 1 2 1

R

R

P

P

P

P

R

R

P

P

I

+

+

=

=

+

=

(6)

where P1,P2 are interpolation polynoms and R1,R2 are the corresponding remainders:

,I P R

R P

(3)

Remark 2

If we shall integrate the formula given by (6) on a plane domain

[ ] [ ]

a bxc d

D= , , , where P1,P2 are supposed to be defined on

[ ]

a,b , respectively

[ ]

c,d , we will obtain a cubature formula with the remainder given by:

(

)( )

[ ] [ ]

∫∫

=abxcd D

dxdy

y

x

f

R

R

, ,

2

1

,

On R1,R2 we shall suppose that they have the corresponding Peano’s kernels

2 2 1 1,n

,

m,n

m

K

K

given by:

( )

( )

(

)

(

) ( )

( )

(

)

   

  

− −

− ⋅ − =     

  

− −

= −

+ =

− + −

+

1

0 1 1

` 1 1

1

,

1 1 1

1

1

1 1

1 1

1 1!

1 !

1

n k m

k k b

a

n n

m n

m x t d x A a t

n n

t x R t

K λ ,

( )

( )

(

)

(

) ( )

( )

(

)

   

  

− −

− ⋅ − =     

  

− −

= −

+ =

− + −

+

1

0 1

2 2

1

,

2 2 2

2 2 2

2

2 2

2 2

2 1!

1 !

1

n k m

k k b

a

n n

m n

m x t d x A a t

n n

t x R t

K λ

Definition 2

About a cubature’s formula generated by the integration of a boolean-sum type formula we shall say that is optimal of Sard type if it verifies the relation:

( )

( )

∫∫

D

n m n

m s K t dsdt

K , 2 , 2 min

2 2 1

1

Theorem 3

If P1,P2are the operators corresponding to an optimal quadrature’s formula and R1,R2 the corresponding remainders then the cubature’s formula obtained by the integration of a (6) type formula is optimal of Sard type in the sens of the definition 2. Remark 3

In order to demonstrate theorem 3 first is easily to see that

( )

( )

∫∫

=

D

n m n

m s K t dsdt

K , 2 , 2

2 2 1

1

( )

( )

b

a

d

c n m n

m s ds K t dt

K , 2 , 2

2 2 1

(4)

If we take care that the quadrature’s formulas generated from the 2

1,P

P operators and the corresponding remainders R1,R2 are optimals of Sard type, that means

( )

( )

→ →

d

c n m b

a n m

dt t K

ds s K

min min

2 ,

2 ,

2 2

1 1

will be evident taking care of (8) that

( )

( )

∫∫

D

n m n

m s K t dsdt

K , 2 , 2 min

2 2 1

1

Remark 4

Such types of optimals quadrature’s formulas can be generalizated starting from a number of n operators

P

1

,

P

2

,...,

P

s and the corresponding remainders

s

R

R

R

1

,

2

,...,

using an n-dimensional boolean-sum type formula:

s s

R

R

R

P

P

P

I

=

1

2

...

+

1 2

...

. (9) Remark 5

A such type of boolean-sum type formula can be obtained supposing that we have verifieds the conditions

j i n j

i P P P

Pi j = j i,∀, =1,..., , ≠ and using a recursive formula

(

n

)

n

n P P P P

P P

P12⊕...⊕ = 12⊕...⊕ 1 ⊕ . (10) For example, using a number of 3 operators

P

1

,

P

2

,

P

3 we shall have:

(

1 2

)

3 1 2 3 1 2 1 3 2 3 1 2 3

3 2

1

P

P

P

P

P

P

P

P

P

P

P

P

P

P

P

P

P

(5)

Remark 6

It is easily to verify that the formula given by (10) has the property of associativity.

Definition 3

About a cubature’s formula generated by the integration of a boolean-sum type formula given by (9) we shall say that is optimal of Sard type if it verifies the relation:

( )

( )

( )

∫∫ ∫

D s s n m n m n

m t K t K t dt dt dt

K

s

s ... min

...

... , 1 2 , 2 2 , 2 1 2

2 2 1

1

with

D

=

[ ] [

a

1

,

b

1

x

a

2

,

b

2

] [

x

...

x

a

s

,

b

s

]

.

Supposing that the s quadrature’s formulas builts with the operators and the remainders mentionated above are optimal of Sard type, that means:

( )

( )

... min min 2 2 2 2 1 1 1 1 2 , 2 ,

→ → b a n m b a n m dt t K dt t K (11)

( )

s

s s s b a n m t dt

K , 2 min

where

( )

( )

(

)

(

) ( )

( )

(

)

       − − − ⋅ − =         − − = − + = − + − +

1 0 1 1 ` 1 1 1 , 1 1 1 1 1 1 1 1 1

1 1!

1 ! 1 n k m k k b a n n m n

m x t d x A a t

n n t x R t K λ

( )

( )

(

)

(

) ( )

( )

(

)

       − − − ⋅ − =         − − = − + = − + − +

1 0 1 2 2 1 , 2 2 2 2 2 2 2 2 2 2 2

2 1!

1 ! 1 n k m k k b a n n m n

m x t d x A a t

n n t x R t K λ …

( )

( )

(

)

(

) ( )

( )

(

)

       − − − ⋅ − =         − − = − + = − + − +

1 0 1 1 , ! 1 1 ! 1 s s s s s s s s s s s s n k m k k b a n s s n m n

m x t d x A a t

n n t x R t K λ

(6)

( )

( )

( )

( )

= 1 →

1 1 1 1

1 ... ... ... min

2 , 1

2 1 , 1

2 , 2 1 ,

b

a

b

a

s s n m n

m D

s s

n m n

m

s

s s s s

s t dt dt K t dt K t dt

K t K

where

D

=

[ ] [

a

1

,

b

1

x

a

2

,

b

2

] [

x

...

x

a

s

,

b

s

]

REFERENCES

1. Coman Gh.,Analiză numerică, Ed. Libris, Cluj-Napoca, 1995

2. Coman Gh.,D.D.Stancu, Blaga Petru- Analiză numericăşi teoria aproximării, vol. I+II, Presa Universitară Clujeană, Cluj-Napoca, 2002

3. ***-Approximation and optimization: Proceedings of ICAOR-International Conference on Approximation and optimization, Transilvania Press, Cluj-Napoca, 1996

Author:

Referensi

Dokumen terkait

Keywords and phrases : quadrature rule, numerical integration, Gauss- Lobatto quadrature formulae, Gauss-Radau quadrature formulae, Chakalov- Popoviciu quadrature formulae,

In this paper we will make a discussion on the conditions when a strongly collisionfree hash function is also one-way hash function and also some considerations about the security

Dans cet article, nous donnons une expression en s´erie de type Atkinson et explorons quelques une des propri´et´es de la moyenne comme fonction en

Applying this result to the matrix A/α , some properties on diagonal dominance and nonsingularity for the matrix A/α d will be presented to establish some results about the

Compared with standard applications of the second moment method in related problems, the analysis here requires the control of second order terms in the large deviation

If D is C 2 then it is known that the stationary distribution for this continuous time process is the product of the uniform measure in D and the standard Gaussian distribution; see [

There are basically three important families of polynomials associated with the root system of type A, namely, the Jack type polynomials, the Laguerre polynomials, and the MP

A conve- nient starting hypothesis which would explain the ubiquity of Lagrangian theories is to consider all successful Lagrangian theories as effective theories arising from