• Tidak ada hasil yang ditemukan

Daniel Breaz_2. 36KB Jun 04 2011 12:08:35 AM

N/A
N/A
Protected

Academic year: 2017

Membagikan "Daniel Breaz_2. 36KB Jun 04 2011 12:08:35 AM"

Copied!
5
0
0

Teks penuh

(1)

SOME KUMMER-TYPE CONVERGENCE CRITERIA

Daniel Breaz, Nicoleta Breaz

Abstract. Positive terms series are an important part of the mathematical analysis. The study of the convergence of series of this interested many mathematicians, these type of in particular convergent series having many applications in various domains. In this paper we present some convergence criteria based on the Kummer convergence criterion.

Keywords:numerical series, positive terms series, divergence.

Definition 1. A series

=1 n

n

a

is called a positive terms series if

a

n is a positive for every

n

=

1

,

2

,...,

.

Theorem 1.(Kummer). Let

a

1

,

a

2

,...,

a

n

,...

be a sequence of positive numbers. If there is a fixed k ≥0 such that

k a u

u a n

n n

n − + >

+1 1

for every n, then the series

=1 n

n

u

is convergent.

If

k a u

u a n

n n

n − + ≤

+

1 1

for every n and the series

=1 1

n an

is divergent then also the series

=1 n

n

u

is divergent.

Proof. In the first case we have

(

1 1

)

1

1

+ +

+

<

n n

n n

n

a

u

a

u

k

u

hence

(

1 1 2 2

)

2

1

u

a

u

a

k

u

<

(

2 2 3 3

)

3

1

u

a

u

a

k

u

<

(2)

(

1 1

)

1

1

+ +

+

<

n n

n n

n

a

u

a

u

k

u

and

(

1 1 1 1

)

1 1 1 1 1

2 1 1

1

1

...

a

u

k

u

u

u

a

u

a

k

u

u

u

S

n+

=

+

+

+

n+

<

n+ n+

+

<

+

. The sequence

S

n is convergent since it is increasing and upper bounded. In the second case we have

0

1

1

n+ n+

n

n

u

a

u

a

or

n n

n n n

n n

n

u u

a a u

u a

a 1 1 1

1 1

1

+ +

+ +

≤ ⇔

≤ .

Using the comparison criterion and the divergence of series

=1 1

n an ,

we obtain that the series

=1 n

n

u

is divergent as well.

Remark.

1)In the application one calculates

λ

=   

+ + ∞

→ 1

1

lim n

n n n

n u a

u

a .

a)If

λ

>0 the series

=1 n

n

u

is convergent.

b)If

λ

<0 the series

=1 n

n

u

is divergent.

2)If

a

n

=

1

we obtain the d’Alembert criterion. 3)If

a

n

=

n

we obtain the Duhamel criterion.

Corrolary 1. Let us consider the sequence

a

n

=

n

p where p

[ ]

0,1 and

=1 n

n

u

is a

positive terms series. If there is a fixed k >0 such that

( )

n k u

u

n p

n n

p − + >

+

(3)

then the series

=1 n

n

u

is convergent.

If

( )

1 0 1

≤ + −

+

p n

n p

n u

u n

then the series

=1 n

n

u

is divergent.

The proof of this corollary is based on the proof on the previous theorem with the sequence

a

n chosen to be

a

n

=

n

p.

Remark.

1)If

p

=

0

we obtain the d’Alembert criterion.

2)If

p

=

1

we obtain the Duhamel criterion we note that this corollary is a move general form of Raabe-Duhamel criterion where can be obtained from if we set

p

=

1

.

Remark. In application one evaluates

( )

n k u

u

n p

n n p

n =

 

+

+ ∞

→ 1

lim

1

.

a)If k >0 the series

=1 n

n

u

is convergent.

b)If k<0 the series

=1 n

n

u

is divergent.

Corollary 2.Let us consider the sequence

0

,

ln

>

=

n

n

a

n

such that

a

n

>

0

. If

=1 n

n

u

is a positive terms series and there is a k >0 such that

i)

( )

n k

u u n

n

n − + >

   

+

1 ln ln

1

, then the series

=1 n

n

u

is convergent.

ii)

( )

n n u

u

n n

ln 1 ln

1

+ ≤

+

, then the series

=1 n

n

u

is divergent.
(4)

( )

1

ln 1 ln lim lim

1

= + ≤

+ ∞

n

n u

u

n n

n

n ,

we obtain the divergence condition from the raport criterion.

Corollary 3.Let us consider the sequence

>

=

n

N

n

a

n

1

2

0

,

.

If

=1 n

n

u

is a positive terms series and there is a fixed k >0 such that:

i)If

( )

n k u

u n n

n >

+ −

+1 2

2

1 1 1

,

then the series

=1 n

n

u

is convergent.

ii)If

( )

+

+

N

n

n

n

u

u

n n

,

1

2 2

1

,

then the series

=1 n

n

u

is divergent.

Corollary 4.Let us consider the sequence

>

=

n

N

n

a

n

0

,

!

1

.

If

=1 n

n

u

is a positive terms series and there is a fixed k >0 such that:

i)If

( )

k n

u u n n

n >

+ −

+ 1!

1 !

1

1

,

then the series

=1 n

n

u

is convergent.

ii)If

+ ≤ + ∀ ∈

N n n

u u

n n

, 1 1

1

,

then the series

=1 n

n

(5)

These convergence criteria can be used to investigate the convergence of some positive terms series for which classical convergence criteria cannot be applied. Some they are based on the Kummer criterion they can also be called Kummer-type criteria.

REFERENCES

[1].Iacob, C., “Lectures in superior mathematics”, Ed. Tehnica, Bucharest, 1959. (in Romanian)

[2].Marusciac, I., “Mathematical analyses”, UBB-Press Print, Cluj-Napoca, 1985.(in Romanian).

Authors.

Daniel Breaz, Nicoleta Breaz, “1 Decembrie 1918” University of Alba Iulia Str. N. Iorga, No. 11-13, 510009, Romania.

Referensi

Dokumen terkait

In recent years, a lot of attention is made by many mathematicians and scien- tists to develop the methods for exact Solitary wave solutions of partial differential equations, such

Abstract: We use Salem’s method [ 13 , 14 ] to prove an inequality of Kwapie´n and Pełczy´nski concerning a lower bound for partial sums of series of bi-orthogonal vectors in a

In fact, as we will see at the end of the article, our method of proof is exible enough and can be adapted to get a smaller error term (of square root type) for finer domains.. In

Defect lattice mode (DL): The number of re-orientation reversal in- creases with increasing shear rate, and the director filed display a finite number of twist inversion walls

In this paper we find infinitely many Lucas and Lehmer pseudoprimes that are analogous to Cipolla pseudoprimes..

Keywords : fractional Brownian motion, function series expansion, rate of convergence, Gamma-mixed Ornstein–Uhlenbeck process.. AMS subject classification:

interested reader. Note that the upper bounds in Lemma 8 and Corollary 10 neither depend on the structure of G nor on the underlying migration mechanism, as long as there is

We prove that for almost all values of the index α – except for a dense set of Lebesgue measure zero – the asymptotic series which were obtained in [13] are in fact