Does the section and the bottom manifold equivalent in the bundle? Here the equivalent means:in the bundle ,B is bottom manifold(also called base space),S is section,we can also treat S as a bottom and B as a section in the same bundle.
Is it true in the vector bundle? and also in the fiber bundle?   
 A: First, a minor point: As Mariano notes in a comment, in English one says base manifold, not bottom manifold.  In any case, I will let $E$ denote the total space of the fibre bundle and $B$ denote the base.
Let $\pi: E \to B$ be a fibre bundle, and let $\sigma:B \to E$ be a section, with image $S = \sigma(B) \subset E$.
Your question is, as far as I understand, the following:
can one find a projection $\pi': E \to S$ which makes $E$ a bundle over $S$,
so that furthermore $B$ is now realized as a section.
In some sense the answer is "Yes", but for trivial reasons:
One can always find such a projection $\pi'$, because $\pi$ induces a diffeomorphism $S \to B$, so that abstractly $B$ and $S$ are the same space. 
More concretely, define $\pi': E \to S$ via $\pi' = \sigma\pi.$  This realizes
$E$ as a fibre bundle over $S$.  However, nothing very exciting has happened; we have just replaced the base $B$ by the diffeomorphic manifold $S$.
We may now define the section $\sigma': S \to E$ via $\sigma'(s) = s.$
The image of $\sigma'$ is just $S$ again, thought of as a submanifold of $E$,
but if you want we can identify this with $B$ via the projection $\pi$.
Again, this is not very exciting.
My sense is that you are thinking of something a little deeper than this, but I'm not sure what.  One problem in general is that the base $B$ and a section
$S$ have a different nature: $S$ maps into $E$, while $E$ maps onto $B$ (so $B$ is not really "in" the fibre bundle at all).
Perhaps the vector bundle case is the one you want to focus on: then
$B$ can be made into a section via the zero section.  But I still don't see
what you are going to do with this.  (One thing you can do is that you can intersect $B$ and $S$, and this is how one produces characteristic classes geometrically; is that of any interest to you at all?)
A: I think what you meant is some duality between $\Gamma(E)$ and $B$. I have never seen bottom manifold, etc defined in literature. I think there is no such relationship. Maybe you can ask some Chinese for help in English so others can understand you. 
The best way to treat sections is to think them as an equivalence between the category of vector bundles over $X$ and the categrory of finitely generated projective modules over $\mathbb{C}(X)$ (complex value functions on $X$). 
