Abstract
We establish two optimal inequalities among power mean , arithmetic mean , logarithmic mean , and geometric mean .
1. Introduction
For , the power mean of order of two positive numbers and is defined by
In the recent past, the power mean has been the subject of intensive research. In particular, many remarkable inequalities for the mean can be found in literature [1–11]. It is well-known that is continuous and strictly increasing with respect to for fixed and .
If we denote by
, and the identric mean, logarithmic mean, arithmetic mean, geometric mean and the harmonic mean, of two positive real numbers and , respectively, then
with equality if and only if in each inequality.
In [12], Alzer and Janous established the following sharp double inequality (see also [13, page 350]):
for all real numbers .
In [14], Mao proved
for all real numbers , and the constant in the left side inequality cannot be improved.
In [15–17], the authors presented the bounds for and in terms of and as follows:
for all with .
Alzer [18] proved
In [5], Alzer and Qiu established
for , and with .
The main purpose of this paper is to present the optimal bounds for and for all in terms of the power mean . Moreover, two optimal inequalities among , , and are proved.
2. Lemmas
In order to establish our main results we need two inequalities, which we present in this section.
Lemma 2.1. If , then for .
Proof. Let , , , , , , , , and , then simple computation yields
From (2.10) we clearly see that is strictly increasing in . Therefore, Lemma 2.1 follows from (2.2)–(2.9) and (2.11) together with the monotonicity of .
Lemma 2.2. If , then for .
Proof. Let , , , , , and , then simple computation leads to
From (2.18) we clearly see that is strictly increasing in . Therefore, Lemma 2.2 follows from (2.13)–(2.17) and (2.19) together with the monotonicity of .
3. Main Results
Theorem 3.1. If , then with equality if and only if , and the parameter cannot be improved.
Proof. If , then we clearly see that
If , then without loss of generality we assume that and let ; hence elementary calculation yields
Let
then
where
From Lemma 2.1 and (3.6) we know that
for .
Therefore, we get
for that follows from (3.3)–(3.5) and (3.8).
Next, we prove that the parameter cannot be improved.
For any , let , then (1.1) leads to
where
Making use of the Taylor expansion we get
Equations (3.10) and (3.12) imply that for any and , there exists , such that
for .
Remark 3.2. For any we have
for all , with equality if and only if , and the parameter in the lower bound cannot be improved.
In fact, if , then we clearly see that . If , then follows from .
Next, we prove that the parameter in the lower bound cannot be improved.
For any , we have
Equation (3.15) implies that for any , there exists , such that
for .
Theorem 3.3. If , then for all , with equality if and only if , and the parameter cannot be improved.
Proof. If , then we clearly see that
If , then without loss of generality, we assume that , let ; hence simple computation leads to
Let
then
where
From Lemma 2.2 and (3.22) we know that
for .
Therefore, we get
for that follows from (3.19)–(3.21) and (3.24).
Next, we prove that the constants cannot be improved.
For any , let , then (1.1) leads to
where
Making use of the Taylor expansion we get
Equations (3.26) and (3.28) imply that for any and , there exists , such that
for .
Remark 3.4. For any , we have Therefore, (3.30) implies that inequality holds for all and , with equality if and only if and the parameter in the lower bound cannot be improved.
Acknowledgments
This work was supported by the National Natural Science Foundation of China (Grant no. 60850005) and the Natural Science Foundation of Zhejiang Province (Grant no. D7080080 and no. Y607128).