<< Предыдущая стр. 43(из 61 стр.)ОГЛАВЛЕНИЕ Следующая >>
x(z + 1) = A(z) x(z),
where the coeп¬ѓcient matrix is exactly as in (3.1), i.e., is holomorphic for
|z| > ПЃ and has at worst a pole at inп¬Ѓnity. The п¬Ѓrst, trivial although
important, observation is that the order of the pole of A(z) does not play
the same crucial role as for systems of ODE: If we set x(z) = О“(z) x(z), then
Лњ
the functional equation of the Gamma function implies that x(z) satisп¬Ѓes
Лњ
в€’1
a diп¬Ђerence equation exactly like before, but with z A(z) as its coeп¬ѓcient
matrix. Thus we learn by repeated application of this argument that we
might restrict ourselves to systems where A(z) is holomorphic at inп¬Ѓnity,
and A(в€ћ) = 0, but for what we shall have to say, this will be not essential.
However, to avoid degeneracies, we shall assume that det A(z) в‰Ў 0, so
that x(z + 1) = A(z) x(z) can be solved for x(z), obtaining an equivalent
diп¬Ђerence equation in the вЂњbackward direction.вЂќ
Under the assumptions made above, the formal theory of such a diп¬Ђerence
equation is well established [80, 81, 92, 133, 134, 223, 269]: Every such
system has a formal fundamental solution of the form

X(z) = F (z) z О›z eQ(z) z L ,
Л† Л†

with:
Л†
вЂў a formal q-meromorphic transformation F (z), for some q в€€ N,
вЂў a diagonal matrix О› of rational numbers with denominators equal to
the same number q,
вЂў a diagonal matrix Q(z) of polynomials in z 1/q without constant term
and of degree strictly less than q, and
вЂў a constant matrix L in Jordan canonical form, commuting with Q(z)
and О›.
Hence, the main diп¬Ђerence between formal solutions of diп¬Ђerence and dif-
ferential equations is the occurrence of the term z О›z . Indeed, if О› vanishes
or is a multiple of the identity matrix, then Braaksma and Faber  have
Л†
shown that F (z) is (k1 , . . . , kp )-summable in all but countably many direc-
tions, with levels 1 = k1 > . . . > kp > 0 that are determined by Q(z)
in exactly the same fashion as for ODE. If О› contains several distinct en-
tries on the diagonal, however, things get considerably more complicated
because of the occurrence of a new level, commonly denoted as level 1+ .
A simple but very instructive example showing this phenomenon may be
13.3 Singular Perturbations 201

found in FaberвЂ™s thesis . Roughly speaking, this example shows that in
presence of level 1+ one can no longer restrict to using Laplace transform
integrating along straight lines, but has to allow other paths of integration.
For a general presentation, see Chapter 3 of the said thesis, containing joint
work of Braaksma, Faber, and Immink.

13.3 Singular Perturbations
Let us consider a system of ODE of the form

ОµПѓ x = g(z, x, Оµ),

where g(z, x, Оµ) is as in Section 13.1, but additionally depends on a parame-
ter Оµ, and Пѓ is a natural number. Analysis of the dependence of solutions on
Оµ is referred to as a singular perturbation problem. Under suitable assump-
tions on the right-hand side, such a system will have a formal solution
в€ћ
x(z, Оµ) = n=0 xn (z) Оµn , with coeп¬ѓcients xn (z) given by diп¬Ђerential re-
Л†
cursion relations. In general, the series is divergent, and classically one has
tried to show existence of solutions of the above system that are asymptotic
to x(z, Оµ) when Оµ в†’ 0 in some sectorial region. Very recently, one has begun
Л†
to investigate Gevrey properties of x, or discuss its (multi-)summability.
Л†
Of the recent articles containing results in this direction, we mention Wal-
let [278, 279], Canalis-Durant , and Canalis-Durant, Ramis, SchВЁfke, a
Since here we meet power series whose coeп¬ѓcients are functions of an-
other variable, the situation is very much analogous to that of formal so-
lutions of partial diп¬Ђerential equations, which we are investigating in the
following section. Here, we shall brieп¬‚y look at a very simple example,
already discussed by Ecalle:
The inhomogeneous equation Оµ x = x в€’ f (z) has the formal solution
в€ћ
x(z, Оµ) = 0 Оµn f (n) (z), for arbitrary f (z) which we assume holomorphic
Л†
near the origin, say, for |z| < ПЃ. According to CauchyвЂ™s formula, the coeп¬ѓ-
cients f (n) (z) grow roughly like n!, so it is natural to study 1-summability
of this series. The formal Borel transformation of order k = 1 of this series
equals the power series expansion of f (z + Оµ) about the point z. Conse-
Л†Л†
quently, holomorphic continuation of B1 x is equivalent to that of f . Hence,
the series x is 1-summable in a direction d if and only if the function f
Л†
admits holomorphic continuation into a sector S(d, Оґ), for some Оґ > 0, and
is of exponential growth at most 1 there. More generally, one can prove a
result on the multisummability of x that is completely analogous to Theo-
Л†
rem 63 (below) for the heat equation. In any case, multisummability of x Л†
is always linked to explicit conditions on the function f (z). This indicates
that for power series of several variables, the notion of multisummability
that has been developed in this book is not general enough. This is due
202 13. Applications in Other Areas, and Computer Algebra

to the fact that here we can only treat one variable at a time, while the
remaining ones are treated as parameters.

13.4 Partial Diп¬Ђerential Equations
There is a classical theory discussing convergence of power series solutions
for Cauchy problems for certain classes of partial diп¬Ђerential equations
(PDE). Recently, eп¬Ђorts have been made to show that formal power series
arising as solutions of such problems have a certain Gevrey order; for such
results, see [108, 190вЂ“194, 213] and the literature cited there. There are
also results, e.g., by Ouchy , showing that such formal solutions are
asymptotic representations of proper solutions of the underlying equation.
Not much is known so far about summability of formal solutions of such
problems: Lutz, Miyake, and SchВЁfke  obtained a п¬Ѓrst result for the
a
complex heat equation, which has been generalized in . Very recent work
by Balser and Miyake  treats an even more general case, showing that
the results obtained are not so much dependent on having a formal solution
of a partial diп¬Ђerential equation, but carry over to series whose coeп¬ѓcients
are given by certain diп¬Ђerential recursions.
Here, we brieп¬‚y investigate the following situation: Given a function П•(z),
analytic in some G containing the origin, try to п¬Ѓnd another function u(t, z)
of two complex variables, in some subset of C 2 near the origin, such that

в€‚2
в€‚
u = 2 u, u(0, z) = П•(z). (13.2)
в€‚t в€‚z
в€ћ
Obviously, (13.2) has the formal solution u(t, z) = n=0 tn П•(2n) (z)/n! In
Л†
an arbitrary compact subset K вЉ‚ G, one can show, using CauchyвЂ™s formula
for derivatives, that П•(2n) K = supzв€€K |П•(2n) (z)| в‰¤ cn О“(1 + 2n), for suп¬ѓ-
ciently large c. Hence u can be regarded as a formal power series in t with
Л†
coeп¬ѓcients in the Banach space E K of functions that are bounded on K
and analytic in its interior. The above estimate implies u в€€ E K [[t]]1 . Thus,
Л†
it is natural to ask for 1-summability of u. If П• happens to be entire, how-
Л†
ever, the above estimate upon its derivatives may be improved, and then u Л†
may be summable of another type. Indeed, one can show that for arbitrary
k1 > . . . > kq > 1/2, under certain explicit conditions, best expressed in
the coeп¬ѓcients of the power series expansion П•(z) = П•n z n , one even has
(k1 , . . . , kq )-summability of x. In detail, the following holds :
Л†
в€ћ
П•n z n for |z| < ПЃ, and deп¬Ѓne u as above,
Theorem 63 Let П•(z) = Л†
0
resp. for 0 в‰¤ j в‰¤ 2:
в€ћ в€ћ
(2n + j)! n Л†В±
Л† П•n О“(1 + n/2) (В±t)n .
П€j (t) = П•2n+j t , П€ (t) =
n!
0 0
13.4 Partial Diп¬Ђerential Equations 203

Given any type of multisummability k = (k1 , . . . , kОЅ ) with kОЅ в‰Ґ 1, and
any multidirection d = (d1 , . . . , dОЅ ), admissible with respect to k, let 2k =
(2k1 , . . . , 2kОЅ ), d/2 = (d1 /2, . . . , dОЅ /2). Then the following statements are
equivalent:
(a) For every K вЉ‚ D(0, ПЃ), u в€€ E K {t}k,d .
Л†
Л†
(b) For 0 в‰¤ j в‰¤ 1, П€j в€€ C {t}k,d .

(c) П€ В± в€€ C {t}2k,d/2 .
Л†

Proof: Assume (a). For K1 вЉ‚ D(0, ПЃ), let K2 вЉ‚ D(0, ПЃ) be so that K1
is contained in the interior of K2 . Then CauchyвЂ™s formula for the deriva-
tive implies that D = d/dz is a bounded linear operator from E K2 into
E K1 . From Theorem 52 (p. 171) we conclude that u в€€ E K2 {t}k,d im-
Л†
в€ћ (2n+1)
(z)/n! в€€ E K1 {t}k,d . Taking z = 0 and us-
plies uz (t, z) =
Л† 0П•
ing Exercise 4 on p. 172, this implies (b). From Theorem 51 (p. 166)
we obtain that (b) is equivalent to П€j (t2 ) being in C {t}2k,d/2 . Apply-
Л†
ing Theorem 50 (p. 164) and Exercise 4 on p. 109, one can then show
в€ћ в€ћ
equivalence with 0 О“(1 + n) П•2n t2n and 0 О“(3/2 + n) П•2n+1 t2n be-
ing in C {t}2k,d/2 . This in turn is equivalent to (c). This leaves to con-
clude (a) from either (b) or (c): To do so, п¬Ѓrst observe that, as a con-
sequence of Theorem 50 (p. 164), it suп¬ѓces to consider the case ОЅ = 1.
в€ћ Л† Л† Л†
In this situation, verify that u(t, z) = m=0 z m П€m (t), with П€0 (t), П€1 (t)
Л†
Л†(Вµ)
Л†
as in (b), and П€m (t) = П€j (t)/m! , for m = 2Вµ + j, 0 в‰¤ j в‰¤ 1. So
(Вµ)
Л† Л†
П€m в€€ C {t}k,d , and П€m (t) = (Sk,d П€m )(t) = П€j (t)/m! in some sectorial
region G = G(d, О±), О± > ПЂ/k, independent of m. From Proposition 9 (p. 70)
ВЇ
we obtain for every closed subsector S вЉ‚ G existence of constants C, c > 0
ВЇ
with |П€m (t)| в‰¤ C c О“(1 + m s)/m!, t в€€ S, m в‰Ґ 0, for s = (1 + 1/k)/2.
m
в€ћ
Owing to k в‰Ґ 1, the series u(t, z) = m=0 z m П€m (t) converges uniformly
(in two variables) for |z| в‰¤ ПЃ, ПЃ > 0 suп¬ѓciently small and t в€€ S 1 . So
ВЇ
u(t, z) в€€ H(G, E K ), for every K as in (a). Diп¬Ђerentiating with respect to
t, one can show
в€ћ (Вµ+n) (Вµ+n)
2Вµ П€0 (t) 2Вµ+1 П€1 (t)
в†’ П•(2n) (z),
n
в€‚t u(t, z) = z +z
(2Вµ)! (2Вµ + 1)!
Вµ=0

as t в†’ 0. Hence we п¬Ѓnd that u(t, z) has u(t, z) as its asymptotic expansion.
Л†
To show that this expansion is of Gevrey order k, we use Proposition 9
(p. 70) again to obtain
в€ћ
|в€‚t u(t, z)| в‰¤ C (c |z|)m О“(1 + s(m + n))/m!
n

m=0

1 In fact, if k > 1, i.e., П•(z) entire, convergence takes place for every z.
204 13. Applications in Other Areas, and Computer Algebra

в€ћ
xsn exp[в€’x + c|z| xs ] dx.
=C
0

The integral can be bounded by О“(1+sn) times some constant to the power
2
n, completing the proof.
It is worthwhile observing that for t = 0 the series u(t, z) in fact converges
for every z, and hence represents an entire function in z that may be seen
to be of exponential size at most 2. This coincides with the classical result
Л†
on the convergence of u(t, z) (see below). The formal series П€j (z) resp.
Л†
П€ В± (z) are explicitly related to П•(z), and using acceleration and Laplace
Л†
operators, one can explicitly reformulate (a), (b) in terms of transforms
of the function П•(z). However, it is in general diп¬ѓcult to check (a) or (b)
directly through investigation of П•(z), except for the special case of ОЅ = 1,
k = k1 = 1. In this situation, the above theorem essentially coincides with
the result obtained by Lutz, Miyake, and SchВЁfke : Conditions (a) or
a
(b) then are equivalent to П•(z) admitting holomorphic continuation into
small sectors bisected by rays arg z = d and arg z = d + ПЂ, and being of
exponential size not more than 2 there.
As follows from the proof of the above theorem, in case of kОЅ > 1 we have
summability of u(t, z) for every z. For kОЅ = 1, however, summability takes
Л†
place only in a disc whose radius in general is smaller than the radius of
convergence of П•(z). Since k-summability in all directions is equivalent to
Л†
convergence, we obtain as a corollary of the above theorem that convergence
of u(t, z) is equivalent to the initial condition being an entire function of
Л†
exponential size at most two. This, however, is a well-known classical result.

 << Предыдущая стр. 43(из 61 стр.)ОГЛАВЛЕНИЕ Следующая >>