A MIXED TYPE IDENTIFICATION PROBLEM RELATED TO A PHASE-FIELD MODEL WITH MEMORY

Size: px
Start display at page:

Download "A MIXED TYPE IDENTIFICATION PROBLEM RELATED TO A PHASE-FIELD MODEL WITH MEMORY"

Transcription

1 Guidetti, D. and Lorenzi, A. Osaka J. Math. 44 (27), A MIXED TYPE IDENTIFICATION PROBLEM RELATED TO A PHASE-FIELD MODEL WITH MEMORY DAVIDE GUIDETTI and ALFREDO LORENZI (Received January 23, 26, revised October 11, 26) Abstract In this paper we consider an integro-differential system consisting of a parabolic and a hyperbolic equation related to phase transition models. The first equation is integro-differential and of hyperbolic type. It describes the evolution of the temperature and also accounts for memory effects through a memory kernel k via the Gurtin-Pipkin heat flux law. The latter equation, governing the evolution of the order parameter, is semilinear, parabolic and of the fourth order (in space). We prove a local in time existence result and a global uniqueness result for the identification problem consisting in recovering the memory kernel k appearing in the first equation. Introduction In order to introduce our mathematical problem, let us consider a smooth bounded container Đ R d, 1 d 3, occupied by the substance undergoing the phase transition. Name and the basic state variables of the process, corresponding to the relative temperature and to the order parameter, respectively. Then, the energy balance equation, describing the evolution of, can be written, under the Gurtin-Pipkin heat flux law (cf. [11]), in the form (.1) D t ( + l) ½(k ) = f, where stands for the standard time convolution product for functions with their supports in R +, k : [, T ] R, k(), is the so-called heat conductivity relaxation kernel, f is a heat source also incorporating an additional term depending on the past history of up to t =, which is assumed to be given, and l is a positive constant accounting for the latent heat. Then, this first equation of hyperbolic type after differentiation, ruling the evolution of, is coupled with the kinetic equation for the phase variable, which is of parabolic type. We will consider the conserved case (.2) D t ½Û =, Û = ½ + () l. 2 Mathematics Subject Classification. 35R3, 45K5, 35M1.

2 58 D. GUIDETTI AND A. LORENZI We stress that system (.1), (.2) of mixed hyperbolic-parabolic type was studied in the paper [4]. We recall that in (.2) is usually assumed to be the sum of a general, possibly multivalued, maximal monotone graph, representing the derivative of the convex part of a double well free energy potential, and of the derivative of a smooth function accounting for its non-convex part. However, in the present paper to solve our problem we will need to consider only sufficiently regular functions. Moreover, relations (.1) and (.2) are assumed to be complemented by homogeneous Neumann boundary conditions both for,, and for the auxiliary unknown Û, generally called chemical potential, and by the Cauchy conditions for and : (.3) (.4) (, x) = (x), (, x) = (x), x ¾ Đ, k = = [ ½ + () l] =, on (, ) Đ, where indicates the normal derivative on Đ. Actually, on account of the boundary condition on Û, it is straightforward to check that the average of is constant in time. The main task of this paper concerns the identification of the kernel k entering equation (.1) under the following additional information involving the temperature : (.5) [(t, )] := Đ ³(x)(t, x) dx = g(t), t ¾ [, ]. We stress that, to derive a fixed point equation for k, we need to differentiate (in time) equations (.1) and (.2) and working on the resulting expression. Further, to perform rigorously such a procedure, however, we need to deal with smooth solutions to problem (.1) (.5). Consequently, we must first obtain preliminary regularity results for the solutions to hyperbolic and parabolic initial and boundary value problems. We point out that these regularity and continuous dependence results for such direct problems, beyond acting as a basis for the subsequent analysis of the inverse problem, might have some independent interest. In the latter part of our paper, we will proceed to solving our identification problem for k. Let us note that identifying memory kernels in systems of partial integro-differential equations (PIDE) related to transition models is a quite new problem, mainly when they are of mixed type. A pioneering paper on this subject is [5] where the author uses analytic semigroup theory to study (locally in time) both the direct and the inverse problems for a non-conserved (parabolic) system of PIDE s (cf. [3, 2, 7]), which couples (.1) provided with an additional term k ½, k, with a second-order parabolic equation for the phase variable. As far as the conserved model considered in this paper is concerned, let us quote the papers [6] and [14], where the authors (using semigroup techniques) study the local

3 PHASE FIELD IDENTIFICATION PROBLEM 581 (in time) identification of a kernel for system (.1), (.2), when the first equation is provided with an additional term k ½. We give now the plan of our paper. In Section 1 we introduce some notation and recall some basic results used throughout the paper. In Section 2 we present some higher-order regularity results concerning the abstract hyperbolic problem (.6) u ¼¼ (t) + Au(t) = f (t), t ¾ [, T ], u() = u, u ¼ () = u 1 where A is a self-adjoint lower bounded operator in the Hilbert space H. In the study of inverse problems, one often needs to work in spaces of quite regular functions. Section 2 is devoted to proving the regularity results given in Proposition 2.9 and Corollary 2.11, which are both used in the final Section 5. In Remark 2.12 we apply our abstract results to the particular case appearing in the inverse problem. In Section 3 we study the higher-order regularity of the solution to an abstract parabolic system in an L p -setting with values in a Hilbert space. Of course, the results obtained are applied in Section 5. In Section 4 we present a careful study of the nonlinear term in our problem. Here also we have to work in spaces of quite regular functions. Finally, Section 5 is devoted to solving locally in time our identification problem (.1) (.5) as well as to proving a global in time uniqueness result. 1. Notation We introduce some notations we shall use in the sequel. We shall indicate with N the set of strictly positive integers, while N will stand for N. R + will denote the set of strictly positive real numbers. If p ¾ [1, +½[, we shall indicate with p ¼ the dual index p(p 1), with p ¼ = ½ if p = 1. The notation W m, p (, T ; X), where X is a Banach space, T ¾ R +, m ¾ N, p ¾ [1, +½[, will denote the space of all measurable functions from (, T ) into X whose distributional derivatives up to the order m belong to L p (, T ; X). Such a space is normed by (1.1) f W m, p (,T ;X) := m 1 k= f (k) () X + f (m) L p (,T ;X).

4 582 D. GUIDETTI AND A. LORENZI We note that this norm is equivalent to the usual one (1.2) f W m, p (,T ;X) := m 1 k= f (k) p L p (,T ;X) 1p, but the constants yielding the equivalence may burst to +½ as T +. Since in some fundamental estimates in this paper occur terms such as f (k) () X and we are going to prove an existence result in the little with respect to time, it will be more appropriate to make use of definition (1.1). Finally, we norm in a similar way C m ([, T ]; X), m ¾ NÒ: (1.3) f C m ([,T ];X) := m 1 k= f (k) () X + f (m) C([,T ];X). It is easily seen that, if f ¾ C 1 ([, T ]; X) and f () =, then (1.4) f C([,T ];X) T f C 1 ([,T ];X) and, if f ¾ C([, T ]; X), (1.5) f L p (,T ;X) T 1p f C([,T ];X). If Đ is an open subset in R n, s, p, q ¾ [1, +½], we shall indicate with B s p,q (Đ) the corresponding Besov space (see, for example, [9], [19]). If X (Đ) is a space of functions of domain Đ, with Đ open and Đ smooth, we shall put, whenever the expression has a meaning, (1.6) (1.7) X B (Đ) := X BB (Đ) := f ¾ X (Đ): f, f ¾ X (Đ): f ½ f. If (X, X 1 ) is a pair of compatible Banach spaces, ¾ (, 1) and p ¾ [1, ½], we shall indicate with (X, X 1 ), p the real interpolation space (see [15]). We shall indicate with C(A, B, ) positive constants depending on A, B,. In a sequence of estimates, we shall often write C 1 (A, B, ), C 2 (A, B, ), etc. 2. Hyperbolic problems We introduce the following basic assumption: (A1) V and H are Hilbert spaces, V H and V is dense in H. We indicate with [, ] the scalar product in H, with (, ) V the scalar product in V, with V and H the norms in V and H respectively, and with V ¼ the space of antilinear bounded functionals in V. We assume that

5 PHASE FIELD IDENTIFICATION PROBLEM 583 (A2) Ú H Ú V, Ú ¾ V in such a way that V is continuously embedded into H. Using Riesz theorem, one can easily verify that the natural norm V ¼ in V ¼ can be obtained from a scalar product, so that V ¼ is by itself a Hilbert space. If h ¾ H, we identify it with the element [h, ] of V ¼. From (A2) it follows immediately that (2.1) h V ¼ h H h ¾ H. If ³ ¾ V ¼ and Ú ¾ V, we set (2.2) [³, Ú] := ³(Ú). (2.2) is consistent with the previous identification of H with a subspace of V ¼. Let now a be a sesquilinear, Hermitian continuous functional of domain V V, such that, for certain and, the estimate (A3) a(ú, Ú) Ú 2 V Ú2 H Ú ¾ V holds (observe that, as a is hermitian, a(ú, Ú) ¾ R). We indicate with A the linear operator from V to V ¼ such that (2.3) [AÚ, Û] = a(ú, Û), Ú, Û ¾ V. Finally, we introduce the following operator A in H: (2.4) D(A) = Ú ¾ V : AÚ ¾ H, AÚ = AÚ, Ú ¾ D(A). It is convenient to consider also the following operator S in H: (2.5) S := A +. Observe that we should obtain S instead of A if we replaced a with a + [, ]. The following properties of S are well known (see, for example, [18], Section 2.2): Theorem 2.1. Assume that the assumptions (A1) (A3) hold. Then, if A and S are the operators defined in (2.4) and (2.5), we have: (I) D(A) is dense in H; (II) S is a positive self-adjoint operator in H; (III) V = D(S 12 ), with equivalent norms. Lemma 2.2. H to V ¼. The operator S 12 can be extended by a linear isomorphism from Proof. If h ¾ H and Ú ¾ V, define (2.6) [T h, Ú] := [h, S 12 Ú].

6 584 D. GUIDETTI AND A. LORENZI Ê If E() ¾ R is the spectral resolution of S, we have S 12 +½ = Ê 12 d E(), so that S 12 is an isomorphism from V to H with inverse S 12 +½ = 12 d E(). So it is clear that the operator T defined in (2.6) is an extension of S 12 to an element of L(H, V ¼ ), as [h, S 12 Ú] h H S 12 Ú H S 12 L(V, H) h H Ú V. Moreover, if ³ ¾ V ¼, the equation T h = ³ has a unique solution h ¾ H: element of H satisfying h is the [h, k] = [³, S 12 k], k ¾ H. REMARK 2.3. In the following, we shall indicate with S 12 the operator T introduced in Lemma 2.2. Proposition 2.4. Set X := V H. Define the following operator G in X: (2.7) D(G) := D(A) V, G(u, Ú) = (Ú, Au), (u, Ú) ¾ D(G). Then G is the infinitesimal generator of a semigroup in X. Proof. With the method developed in the proof of Theorem 2.1, Chapter 5 in [13], if (u, Ú ) ¾ D(A) V, the problem (2.8) u ¼¼ (t) + Au(t) =, t, u() = u, u ¼ () = Ú has a unique solution u in C 2 ([, +½[; H) C 1 ([, +½[; V ) C([, +½[; D(A)). If t, we define (2.9) S(t)(u, Ú ) := (u(t), u ¼ (t)). From Theorem 8.2 in [12] we have also that for every T there exists C(T ) such that, if t T, the estimate (2.1) u(t) V + u ¼ (t) H C(T )(u V + Ú H ) holds. As D(A) V is dense in X, we conclude from (2.1) that for any t ¾ [, +½[ S(t) is extensible to an element of L(X), which we continue to indicate with S(t). From the uniqueness of the solution, it is not difficult to verify that S(t): t is a strongly continuous semigroup of linear bounded operators in X.

7 PHASE FIELD IDENTIFICATION PROBLEM 585 Now we show that its infinitesimal generator is G. We indicate for a moment this generator with G ¼. Then it is clear that G ¼ is an extension of G, as, if (u, Ú ) ¾ D(G), then S( )(u, Ú ) ¾ C 1 ([, +½[; X) and S ¼ ()(u, Ú ) = G(u, Ú ). On the other hand, from ] ½, ] (A) it follows easily [ Ô, +½[ (G). So, if we pick ¾ R sufficiently large in such a way that ¾ (G) (G ¼ ), we obtain easily that ( G) 1 = ( G ¼ ) 1, implying immediately G = G ¼. Lemma 2.5. Let G be the infinitesimal generator of a strongly continuous semigroup S(t): t in the Banach space X and let F ¾ W 1,1 (, T ; X). Then the function u(t) := t S(t s)f(s) ds, t ¾ [, T ], belongs to ¾ C 1 ([, T ]; X)C([, T ]; D(G)) and u ¼ (t) = Gu(t) + F(t) for any t ¾ [, T ]. Proof. It is well known (see, for example, [16], Corollary 2.5) that, if F ¾ C 1 ([, T ]; X), then u ¾ C 1 ([, T ]; X) C([, T ]; D(A)). Moreover, for any t ¾ [, T ], (2.11) (2.12) u ¼ (t) = S(t)F() + t Au(t) = u ¼ (t) F(t). S(t s)f ¼ (s) ds, Fix a sequence (F k ) k¾n in C 1 ([, T ]; X), converging to F in W 1,1 (, T ; X) and indicate with u k the function obtained replacing F with F k. Then, from (2.11) (2.12), we deduce that (u k (t)) k¾n, (u ¼ k (t)) k¾n, (Au k (t)) k¾n converge, uniformly in [, T ], respectively to u(t), S(t)F() + Ê t S(t s)f¼ (s) ds and S(t)F() + Ê t S(t s)f¼ (s) ds F(t). From this the conclusion easily follows. (2.13) Corollary 2.6. Consider the problem u ¼¼ (t) + Au(t) = f (t), t ¾ [, T ], u() = u, u ¼ () = u 1 under the assumption (A1) (A3), where A is the operator defined in (2.4). If f ¾ W 1,1 (, T ; H), u ¾ D(A), u 1 ¾ V, then there exists a unique solution u belonging to C 2 ([, T ]; H) C 1 ([, T ]; V ) C([, T ]; D(A)). Moreover, if f ¾ W 1, p (, T ; V ¼ ), with 1 p +½, then u belongs also to W 3, p (, T ; V ¼ ); if f ¾ C 1 ([, T ]; V ¼ ), then u belongs also to C 3 ([, T ]; V ¼ ). Finally, for any T there exists C(T ) such that, if T T, (2.14) (2.15) u C 2 ([,T ];H) + u C 1 ([,T ];V ) + u C([,T ];D(A)) C(T )(u D(A) + u 1 V + f W 1,1 (,T ;H)), u W 3, p (,T ;V ¼ ) C(T )(u D(A) + u 1 V + f W 1,1 (,T ;H) + f W 1, p (,T ;V ¼ )).

8 586 D. GUIDETTI AND A. LORENZI (2.16) Proof. Putting Ú := u ¼, the problem (2.13) is equivalent to the system u ¼ (t) = Ú(t), t T, Ú ¼ (t) = Au(t) + f (t), t T, u() = u, Ú() = u 1. As the function t (, f (t)) belongs to W 1,1 (, T ; V H) and (u, u 1 ) ¾ D(G), G being defined in (2.7), by Lemma 2.5, problem (2.16) has a unique solution U = (u, Ú) belonging to C 1 ([, T ]; V H)C([, T ]; D(A) V ). This implies u ¾ C 1 ([, T ]; V ) C([, T ]; D(A)). As Ú ¾ C 1 ([, T ]; H) and Ú = u ¼, we conclude that u ¾ C 2 ([, T ]; H). Estimate (2.14) can be obtained by the following argument: extend f to an element f ¾ W 1,1 (, T ; H) setting f (t) = f (T ) if T t T. Next, consider the problem (2.13) in [, T ] replacing f with f. If ũ is its solution, evidently u is the restriction of ũ to [, T ]. So we have u C 2 ([,T ];H) + u C 1 ([,T ];V ) + u C([,T ];D(A)) ũ C 2 ([,T ];H) + ũ C 1 ([,T ];V ) + ũ C([,T ];D(A)) C(T )(u D(A) + u 1 V + f W 1,1 (,T ;H)) = C(T )(u D(A) + u 1 V + f W 1,1 (,T ;H)). Assume now that f ¾ W 1,1 (, T ; H) W 1, p (, T ; V ¼ ). As u ¾ C 1 ([, T ]; V ), Au = Au ¾ C 1 ([, T ]; V ¼ ). As u ¼¼ = Au + f, we obtain immediately the conclusion u ¾ W 3, p (, T ; V ¼ ) and the estimate (2.15). The same argument works in case f ¾ C 1 ([, T ]; V ¼ ). If t, u ¾ V, u 1 ¾ H, set (2.17) S(t)(u, u 1 ) = (S 11 (t)u + S 12 (t)u 1, S 21 (t)u + S 22 (t)u 1 ). S 11, S 12, S 21, S 22 are strongly continuous with values in L(V ), L(H, V ), L(V, H), L(H), respectively. Moreover, for any u ¾ D(A), for any u 1 ¾ V, (2.18) S 11 (t)u + S 12 (t)u 1 = u + t [S 21 (s)u + S 22 (s)u 1 ] ds. By continuity, (2.18) can be extended to u ¾ V, u 1 ¾ H. So S ¼ 11 (t)u = S 21 (t)u for any u ¾ V, S ¼ 12 (t)u 1 = S 22 (t)u 1 for any u 1 ¾ H. Finally, S 11 () = I V, S 22 () = I H, S 12 ()u 1 = for any u 1 ¾ H, S 21 ()u = for any u ¾ V.

9 PHASE FIELD IDENTIFICATION PROBLEM 587 The solution u of the problem (2.13) under the assumptions of Corollary 2.6 can be represented in the form (2.19) u(t) = S 11 (t)u + S 12 (t)u 1 + t S 12 (t s) f (s) ds. We consider the function u given by (2.19) under conditions on u, u 1, f which are less restrictive if compared with those of Corollary 2.6. Proposition 2.7. Consider the function u defined in (2.19) assuming that f ¾ L 1 (, T ; H), u ¾ V, u 1 ¾ H. Then u belongs to C 1 ([, T ]; H) C([, T ]; V ). Moreover, if f ¾ L p (, T ;V ¼ ) (1 p +½), then u belongs also to W 2, p (, T ;V ¼ ), if f ¾ C([, T ]; V ¼ ), u ¾ C 2 ([, T ]; V ¼ ). Finally, for any T there exists C(T ) such that, if T T, (2.2) (2.21) u C 1 ([,T ];H) + u C([,T ];V ) C(T )(u V + u 1 H + f L 1 (,T ;H)), u W 2, p (,T ;V ¼ ) C(T )(u V + u 1 H + f L 1 (,T ;H) + f L p (,T ;V ¼ )). Proof. As S 11 and S 12 are strongly continuous with values in L(V ) and L(H, V ) respectively, u ¾ C([, T ]; V ). Moreover, from S ¼ 11 = S 21, S ¼ 12 = S 22 and S 12 () =, we obtain that u ¾ C 1 ([, T ]; H) and u ¼ (t) = S 21 (t)u + S 22 (t)u 1 + t S 22 (t s) f (s) ds t ¾ [, T ]. Consider now a sequence (u,k ) k¾n in D(A), converging to u in V, a sequence (u 1,k ) k¾n in V, converging to u 1 in H and a sequence ( f k ) k¾n in W 1,1 (, T ; H), converging to f in L 1 (, T ; H). Call u k the solution of (2.13) obtained replacing u with u k, u 1 with u 1,k, f with f k. Then u k ¾ C 2 ([, T ]; H)C 1 ([, T ]; V ) and the sequence (u k ) k¾n converges to u in C 1 ([, T ]; H)C([, T ];V ). As u ¼¼ k = Au k + f k, the sequence (u ¼¼ k ) k¾n converges to Au + f in L 1 (, T ; V ¼ ). So u ¼¼ = Au + f, where u ¼¼ is the second derivative of u in the sense of distributions with values in V ¼. As Au ¾ C([, T ]; V ¼ ), if f ¾ L p (, T ; V ¼ ), then u ¾ W 2, p (, T ; V ¼ ). In the same way, it follows that, if f ¾ C([, T ]; V ¼ ), then u ¾ C 2 ([, T ]; V ¼ ). Estimates (2.2) and (2.21) can be obtained with the arguments used to prove estimate (2.14), extending f to ]T, T ] by. DEFINITION 2.8. Let «¾ [ 12, [. We set D(S «) := ³ ¾ V ¼ : Ú ¾ D(S 12+«): ³ = S 12 Ú, where S 12 is defined in H (see Lemma 2.2 and Remark 2.3). Ú is obviously unique

10 588 D. GUIDETTI AND A. LORENZI and coincides with S 12 ³, where S 12 is assumed to be extended to V ¼. If ³ ¾ D(S «), we set (2.22) ³ D(S«) := S 12 ³ D(S 12+«). Corollary 2.6 and Proposition 2.7 admit the following extensions, which can be easily obtained using the fractional powers S «of S: Proposition 2.9. Consider the problem (2.13), under the assumption (A1) (A3), where A is the operator defined in (2.4). Let «¾ [, +½[. Assume that f ¾ W 1,1 (, T ; D(S «)), u ¾ D(S 1+«), u 1 ¾ D(S 12+«). Then the solution u belongs to C 2 ([, T ]; D(S «)) C 1 ([, T ]; D(S 12+«)) C([, T ]; D(S 1+«)). Moreover, if f ¾ W 1, p (, T ; D(S «12 )) (1 p +½), then u belongs also to W 3, p (, T ; D(S «12 )); if f ¾ C 1 ([, T ]; D(S «12 )), then u belongs also to C 3 ([, T ]; D(S «12 )). Finally, for any T there exists C(T ) such that, if T T, (2.23) (2.24) u C 2 ([,T ];D(S «)) + u C 1 ([,T ];D(S 12+«)) + u C([,T ];D(S 1+«)) C(T )(u D(S 1+«) + u 1 D(S 12+«) + f W 1,1 (,T ;D(S ))), «u W 3, p (,T ;D(S «12 )) C(T )(u D(S 1+«) + u 1 D(S 12+«) + f W 1,1 (,T ;D(S «)) + f W 1, p (,T ;D(S «12 ))). Proposition 2.1. Let «¾ [, +½[. Consider the function u defined in (2.19) under the assumptions f ¾ L 1 (, T ; D(S «)), u ¾ D(S «+12 ), u 1 ¾ D(S «). Then u belongs to C 1 ([, T ]; D(S «)) C([, T ]; D(S «+12 )). Moreover, if f ¾ L p (, T ; D(S «12 )), (1 p +½) then u belongs also to W 2, p (, T ; D(S «12 )); if f ¾ C([, T ]; D(S «12 )), then u belongs also to C 2 ([, T ]; D(S «12 )). Finally, for any T there exists C(T ) such that, if T T, (2.25) u C 1 ([,T ];D(S «)) + u C([,T ];D(S «+12 )) C(T )(u D(S «+12 ) + u 1 D(S«) + f L 1 (,T ;D(S «))), and (2.26) u W 2, p (,T ;D(S «12 )) C(T )(u D(S «+12 ) + u 1 D(S«) + f L 1 (,T ;D(S «)) + f L p (,T ;D(S «12 ))). In Section 5 we shall need the following further regularity result: Corollary Consider problem (2.13), with f ¾ W 1, p (, T ;V ¼ )C([, T ];V ) L p (, T ; D(S 32 )), u = u 1 =. Then the function u defined in (2.19) belongs to

11 PHASE FIELD IDENTIFICATION PROBLEM 589 W 3, p (, T ; V ¼ ) C 2 ([, T ]; V ) C 1 ([, T ]; D(S 32 )). Moreover, if T T, there exist C(T ), such that u W 3, p (,T ;V ¼ ) + u C 2 (,T ;V ) + u C 1 (,T ;D(S 32 )) C(T )( f W 1, p (,T ;V ¼ ) + f C([,T ];V ) + f L p (,T ;D(S 32 ))). Proof. By Proposition 2.1, with «= 32, u ¾ C 1 ([, T ]; D(S 32 )), so that Au ¾ C 1 ([, T ]; V ). Again by Proposition 2.1, with «=, u ¾ C 2 ([, T ]; V ¼ ). As u ¼¼ = Au + f ¾ C([, T ]; V ), u ¾ C 2 ([, T ]; V ). As Au + f ¾ W 1, p (, T ; V ¼ ), u ¾ W 3, p (, T ; V ¼ ). Moreover, by Proposition 2.1, (2.27) u C 1 ([,T ];D(S 32 )) C 1 (T ) f L 1 (,T ;D(S 32 )) C 1 (T )T 1p¼ f L p (,T ;D(S 32 )), so that Finally, u C 2 ([,T ];V ) = D 2 t u C([,T ];V ) Au C([,T ];V ) + f C([,T ];V ) T Au C 1 ([,T ];V ) + f C([,T ];V ) CT u C 1 ([,T ];D(S 32 )) + f C([,T ];V ) T 2 1p C 1 (T ) f L p (,T ;D(S 32 )) + f C([,T ];V ). u W 3, p (,T ;V ¼ ) = Dt 2 u W 1, p (,T ;V ¼ ) Au W 1, p (,T ;V ¼ ) + f W 1, p (,T ;V ¼ ) CAu W 1, p (,T ;V ) + f W 1, p (,T ;V ¼ ) CT 1p Au C 1 ([,T ];V ) + f W 1, p (,T ;V ¼ ) C 2 (T )( f L p (,T ;D(S 32 )) + f W 1, p (,T ;V ¼ )). REMARK Operator S is self adjoint and positive. So, for any t ¾ R +, S it is an isometry in H. It follows from Theorem in [19] that, for any «¾ ], 1[, D(S «) coincides with the complex interpolation space [H, D(S)] «, with equivalent norms. We apply the previous result to a specific case useful in the sequel. For this purpose we introduce the following condition: (B1) Đ is an open bounded subset of R n, lying on one side of its boundary Đ, which is a submanifold of R n of class C 4. Under assumption (B1), we set (2.28) V := H 1 (Đ), H := L 2 (Đ).

12 59 D. GUIDETTI AND A. LORENZI If Ú and Û are elements of V, we set (2.29) a(ú, Û) := Observe that, for any Ú ¾ V, Đ ÖÚ(x) ÖÛ(x) dx. a(ú, Ú) = Ú 2 V Ú 2 H, so that (A3) is satisfied. It is well known that, under the assumption (B1), (2.3) D(A) = D(S) = Ú ¾ H 2 (Đ): Ú = in Đ and (2.31) Au = ½u u ¾ D(A). Owing to Remark 2.12 and Theorem 4.1 in [17], we have, for any «¾ ], 1[, with «= 34, (2.32) D(S «) = H 2«(Đ) if «3 4, H 2«B (Đ) if 3 4 «1, where H 2«B (Đ) denotes the space defined by (1.6) with X = H 2«(Đ). Owing to the regularity of Đ and known results of regularity for solutions of elliptic boundary value problems (see, for example, [9] in case p = q = 2), we have also, for any «¾ ], 1[, with «= 34, (2.33) D(S 1+«) = H 2(1+«) B (Đ) if «3 4, H 2(1+«) BB (Đ) if 3 4 «1, where, if «34, H 2(1+«) BB (Đ) denotes the space defined by (1.7) with X = H 2(1+«) (Đ). 3. Parabolic problems We start with the following result, due to De Simon (see [8]): Theorem 3.1. Let B be the infinitesimal generator of an analytic semigroup

13 PHASE FIELD IDENTIFICATION PROBLEM 591 (e t B ) t in the Hilbert space X. Let T ¾ R +, p ¾ ]1, +½[, f ¾ L p (, T ; X). Set, for t ¾ [, T ], Ú(t) := t Then Ú ¾ W 1, p (, T ; X) L p (, T ; D(B)). e (t s)b f (s) ds. Corollary 3.2. Let B be the infinitesimal generator of an analytic semigroup (e t B ) t in the Hilbert space X. Let T ¾ R +, p ¾ ]1, +½[, f ¾ L 1 (, T ; X), Ú ¾ X. Then the Cauchy problem (3.1) Ú ¼ (t) = BÚ(t) + f (t), t ¾ [, T ], Ú() = Ú has a solution Ú in W 1, p (, T ; X) L p (, T ; D(B)) if and only if f ¾ L p (, T ; X) and Ú ¾ (X, D(B)) 1p ¼, p. In this case the solution with such regularity is unique and can be represented in the form (3.2) Ú(t) := e t B Ú + t e (t s)b f (s) ds. Finally, for any T ¾ R +, there exists C(T ), such that, if T T, L p (, T ; X) and Ú ¾ (X, D(B)) 1p ¼, p, f ¾ (3.3) Ú W 1, p (,T ;X) + Ú L p (,T ;D(B)) C(T )[Ú (X, D(B))1p ¼, p + f L p (,T ;X)]. Proof. The condition f ¾ L p (, T ; X) is clearly necessary to get a solution with the required regularity. Moreover, Ú() Ú ¾ W 1, p (, T ; X) L p (, T ; D(B)) = (X, D(B)) 1p ¼, p (see [15], Proposition 1.2.1). On the other hand, assume that f ¾ L p (, T ; X) and Ú ¾ (X, D(B)) 1p ¼, p. Let Ú be the function defined in (3.2). Then, by Theorem 3.1, t Ê t e(t s)b f (s) ds ¾ W 1, p (, T ; X) L p (, T ; D(B)). Moreover, by Proposition in [15], t e t B Ú belongs to L p (, T ; D(B)). As its derivative is Be t B Ú, we conclude that Ú ¾ W 1, p (, T ; X) L p (, T ; D(B)). If f ¾ C «([, T ]; X), for some «¾ R +, and Ú ¾ D(B), it is well known that Ú ¾ C 1 ([, T ]; X) C([, T ]; D(B)). Moreover, Ú ¼ (t) = BÚ(t) + f (t) for all t ¾ [, T ]. Approximating f by a sequence ( f k ) k¾n with values in C 1 ([, T ]; X), converging to f in L p (, T ; X), and Ú with a sequence (Ú k ) k¾n with values in D(B), converging to Ú in (X, D(B)) 1p ¼, p (which exists, owing to Proposition in [15]), we conclude that Ú ¼ (t) BÚ(t) = f (t) almost everywhere in ], T [ and in the sense of vector valued distributions.

14 592 D. GUIDETTI AND A. LORENZI To show the uniqueness, let Ú ¾ W 1, p (, T ; X) L p (, T ; D(B)) be such that Ú() = and Ú ¼ (t) = BÚ(t) almost everywhere in ], T [. Set, for t ¾ [, T ], V (t) := t Ú(s) ds. Then V ¾ C 1 ([, T ]; X) C([, T ]; D(B)). Moreover, for all t ¾ [, T ], V ¼ (t) = Ú(t) = t BÚ(s) ds = BV (t). Owing to well known properties of semigroups, we conclude that V, and so Ú = V ¼. To prove estimate (3.3), one can argue as in the proof of the estimate (2.14), extending f to the element f of L p (, T ; X) such that f (t) = if t ¾ ]T, T ]. Lemma 3.3. Let B be the infinitesimal generator of an analytic semigroup (e t B ) t in the Hilbert space X. Let T ¾ R +, p ¾]1, +½[, f ¾ L 1 (, T ; X), Ú ¾ X. Then: (I) the Cauchy problem (3.1) has a solution Ú in W 2, p (, T ; X) W 1, p (, T ; D(B)) if and only if f ¾ W 1, p (, T ; X), Ú ¾ D(B) and Ú 1 := BÚ + f () ¾ (X, D(B)) 1p ¼, p; (II) the Cauchy problem (3.1) has a solution Ú in W 3, p (, T ; X) W 2, p (, T ; D(B)) if and only if f ¾ W 2, p (, T ; X), Ú, Ú 1 ¾ D(B) and Ú 2 := BÚ 1 + f ¼ () ¾ (X, D(B)) 1p ¼, p. Moreover, for any T ¾ R +, there exists C(T ), such that, in case (I) with T T, f ¾ W 1, p (, T ; X), Ú ¾ D(B), Ú 1 ¾ (X, D(B)) 1p ¼, p, (3.4) Ú W 2, p (,T ;X) + Ú W 1, p (,T ;D(B)) C(T )[Ú D(B) + Ú 1 (X, D(B))1p ¼, p + f W 1, p (,T ;X)], in case (II), with T T, f ¾ W 2, p (, T ; X), Ú, Ú 1 ¾ D(B), Ú 2 ¾ (X, D(B)) 1p ¼, p, (3.5) Ú W 3, p (,T ;X) + Ú W 2, p (,T ;D(B)) C(T )[Ú D(B) + Ú 1 D(B) + Ú 2 (X, D(B))1p ¼, p + f W 2, p (,T ;X)]. Proof. As W 1, p (,T ;D(B)) C([,T ];D(B)), if Ú ¾ W 2, p (,T ;X)W 1, p (,T ;D(B)) and solves (3.1), then Ú ¾ D(B) and f ¾ W 1, p (, T ; X). Moreover, Û := Ú ¼ belongs to W 1, p (, T ; X) L p (, T ; D(B)) and solves the problem (3.6) Û ¼ (t) = BÛ(t) + f ¼ (t), t ¾ [, T ], Û() = Ú 1. So, necessarily Ú 1 ¾ (X, D(B)) 1p ¼, p.

15 PHASE FIELD IDENTIFICATION PROBLEM 593 On the other hand, if Û ¾ W 1, p (, T ; X) L p (, T ; D(B)) and solves (3.6), setting Ú(t) := Ú + 1 Û, we get Ú ¾ W 2, p (, T ; X) W 1, p (, T ; D(B)). Moreover, for any t ¾ ], T [, Ú ¼ (t) = Û(t) = Ú 1 + t (BÛ(s) + f ¼ (s)) ds = Ú 1 + BÚ(t) BÚ + f (t) f () = BÚ(t) + f (t). Thus (I) is proved. (II) can be shown with similar arguments. The final estimates can be proved applying (3.3) to the derivatives of Ú. We consider now operator B := A 2, with A defined in (2.3) (2.31), under assumption (B1), in the space X = H = L 2 (Đ). So, from well known results concerning the regularity of solutions of elliptic problems, its domain is HBB 4 (Đ). It is well known that B is the infinitesimal generator of an analytic semigroup in H. Next, we set V := H 1 (Đ) and consider the following operator B in V ¼ of domain HB 3(Đ): (3.7) (BÚ, Û) = Đ Ö(½Ú)(x) ÖÛ(x) dx, Ú ¾ H 3 B (Đ), Û ¾ V. Owing to the isomorphism S 12, even B is the infinitesimal generator of an analytic semigroup in V ¼. The following result holds: (3.8) Proposition 3.4. Consider the Cauchy problem Ú ¼ (t) = BÚ(t) + f (t), t ¾ [, T ], Ú() = Ú. Let p ¾ ]1, +½[. Then the following conditions are necessary and sufficient in order that there exist a unique solution Ú ¾ W 3, p (, T ; V ¼ ) W 2, p (, T ; HB 3(Đ)) W 1, p (, T ; HBB 4 (Đ)): (I) f ¾ W 2, p (, T ; V ¼ ) W 1, p (, T ; L 2 (Đ)); (II) Ú ¾ HBB 4 (Đ)); (III) Ú 1 := BÚ + f () ¾ (HB 3(Đ), H BB 4 (Đ)) 1p ¼, p; (IV) Ú 2 := BÚ 1 + f ¼ () ¾ (V ¼, HB 3(Đ)) 1p ¼, p. Moreover, for any T ¾ R + there exists C(T ), such that, if T T and the conditions (I) (IV) are satisfied, (3.9) Ú W 3, p (,T ;V ¼ ) + Ú W 2, p (,T ;H 3 (Đ)) + Ú W 1, p (,T ;H 4 (Đ)) C(T ) f W 2, p (,T ;V ¼ ) + f W 1, p (,T ;L 2 (Đ)) + Ú H 4 (Đ) + Ú 1 (H 3 B (Đ), H 4 BB (Đ)) 1p ¼, p + Ú 2 (V ¼, H 3 B (Đ)) 1p ¼, p.

16 594 D. GUIDETTI AND A. LORENZI Proof. Clearly the only possible solution is (3.1) Ú(t) = e t B Ú + t e (t s)b f (s) ds = e tb Ú + t e (t s)b f (s) ds. Owing to Lemma 3.3, necessary and sufficient conditions for (3.8) to have a solution with the desired regularity are: (a) f ¾ W 2, p (, T ; V ¼ ) W 1, p (, T ; L 2 (Đ)); (b) Ú ¾ HBB 4 (Đ)); (c) Ú 1 ¾ (L 2 (Đ), HBB 4 (Đ)) 1p ¼, p HB 3(Đ); (d) Ú 2 ¾ (V ¼, HB 3(Đ)) 1p ¼, p. So (I) (II) and (IV) are necessary. As Ú ¼ ¾ W 1, p (, T ; HB 3(Đ)) L p (, T ; HBB 4 (Đ)) and Ú ¼ () = Ú 1, even (III) is necessary, again by [15], Proposition On the other hand, (I) (IV) imply (a) (d). (3.9) follows from (3.4) and (3.5). Proposition 3.5. Let p 2. Then W 3, p (, T ; V ¼ ) W 2, p (, T ; H 3 (Đ)) C 2 ([, T ]; H 1 (Đ)). In particular, if assumptions (I) (IV) of Proposition 3.4 are fulfilled, the solution Ú of problem (3.8) belongs to C 2 ([, T ]; H 1 (Đ)). Moreover, for any T ¾ R +, there exists C(T ) ¾ R +, such that, if T ¾ ], T ], (3.11) Ú C 2 ([,T ];H 1 (Đ)) C(T ) f W 2, p (,T ;V ¼ ) + Ú H 3 (Đ) + Ú 1 H 3 (Đ) + Ú 2 (V ¼, H 3 B (Đ)) 1 1p, p. Proof. Let Ú ¾ W 3, p (, T ; V ¼ ) W 2, p (, T ; H 3 (Đ)). Then we have that Dt 2 Ú ¾ W 1, p (, T ; V ¼ ) L p (, T ; H 3 (Đ)) W 1,2 (, T ; V ¼ ) L 2 (, T ; H 3 (Đ)). So, by Theorem 3.1 in Chapter 1 of [12], Dt 2 Ú is continuous with values in the complex interpolation space [V ¼, H 3 (Đ)] 12, coinciding with H 1 (Đ). As Ú() and D t Ú() are elements of H 3 (Đ), Ú ¾ C 2 ([, T ]; H 1 (Đ)). Let now Ú be the solution of (3.8), with T ¾ ], T ]. If the conditions (I) (IV) are satisfied and p 2, from the first part of the statement, we have Ú ¾ C 2 ([, T ]; H 1 (Đ)). Then, employing again the argument used to show estimate (2.14), we obtain estimate (3.11). REMARK 3.6. (3.12) (H 3 B (Đ), H 4 BB (Đ)) 1p ¼, p = It is well known (see, for example, [1], Theorem 3.5), that B 4 1p 2, p, B (Đ) if 1 p 2, B 4 1p 2, p,bb (Đ) if 2 p +½. A more involved characterization is known in case p = 2 (see [1]).

17 PHASE FIELD IDENTIFICATION PROBLEM 595 Proposition 3.7. (3.13) (V ¼, H 3 B (Đ)) 1p ¼, p = Proof. Under the assumption (B1), we have that S 12 B 4(1p¼ ) 2, p (Đ) if 1 p 8 5, B 3 4p 2, p (Đ) if Owing to Theorem 3.5 in [1], we have (3.14) (L 2 (Đ), H 4 BB (Đ)) 1p ¼, p = 4 3 p 8 3, B 3 4p 2, p, B (Đ) if 8 3 p +½. B 4p¼ 2, p (Đ) if 4 p ¼ 3 2, B 4p¼ 2, p, B (Đ) if p ¼ 5 2, B 4p¼ 2, p,bb (Đ) if 4 p ¼ 5 2. As HBB 4 (Đ) = D(S2 ) and HB 3(Đ) = D(S32 ) (see (2.33)), S 12 is an isomorphism between HBB 4 (Đ) and H B 3 (Đ). So, by Lemma 2.2, if 1 p 85. Moreover, as (V ¼, H 3 B (Đ)) 1p ¼, p = S 12 ((L 2 (Đ), H 4 BB (Đ)) 1p ¼, p) = S 12 B 4p¼ 2, p (Đ) (L 2 (Đ), H 4 BB (Đ)) 14,1 H 1 (Đ) (L 2 (Đ), H 4 BB (Đ)) 14,½ (by Theorem 3.5 and Proposition 1.6 in [1]), using again the isomorphism S 12, we can say that (V ¼, H 3 B (Đ)) 14,1 L 2 (Đ) (V ¼, H 3 B (Đ)) 14,½. So, by the reiteration theorem, if p 43 (so that 1p ¼ 14), we have (V ¼, H 3 B (Đ)) 1p ¼, p = (L 2 (Đ), H 3 B (Đ)) 1 43p, p. So we can get the result applying again Theorem 3.5 in [1]. 4. An auxiliary nonlinear operator In this section we study the nonlinear operator V S( +1 V )V, where S: R R is appropriately smooth, V is a suitably regular function of domain [, T ] Đ and (4.1) (1 V )(t, x) := t V (s, x) ds, t ¾ [, T ].

18 596 D. GUIDETTI AND A. LORENZI We introduce the following conditions: (C1) (B1) holds; (C2) n 7. Lemma 4.1. Under the conditions (C1) (C2), H 4 (Đ) is continuously embedded in C(Đ) and is a space of pointwise multipliers for H s (Đ), for any s ¾ [, 4]. Proof. The first statement follows from Sobolev embedding theorem. The embedding of H 4 (Đ) into C(Đ) implies also that H 4 (Đ) is a space of pointwise multipliers for L 2 (Đ). By Theorem 5.23 in [1], H 4 (Đ) is also a Banach algebra. So the second statement follows by complex interpolation (see [19], Theorem 2 in Section 4.3.1). Lemma 4.2. Assume that (C1) (C2) hold. Let j ¾ Z, j 4, and S ¾ C j (R). Then, for any u ¾ H 4 (Đ), S Æ u is a pointwise multiplier for H j (Đ). Moreover, if u H 4 (Đ) R and Ú ¾ H j (Đ), (4.2) S(u)Ú H j (Đ) C(R)Ú H j (Đ). Finally, if S ¾ C j+1 (R), u, u 1 belong to H 4 (Đ), maxu H 4 (Đ), u 1 H 4 (Đ) R and Ú belongs to H j (Đ) (4.3) S(u )Ú S(u 1 )Ú H j (Đ) C(R)u u 1 H 4 (Đ)Ú H j (Đ). Proof. First of all, if u ¾ H 4 (Đ) and Ú ¾ H j (Đ), then S(u)Ú ¾ H j (Đ). In fact, it is easily seen that, if ¾ N n and j, (S(u)Ú) is a linear combination of terms of the form S (k) (u) 1 u k u Ú, with k j, k + j (here we use the convention that there are no derivatives of u if k = ). Now, S (k) (u) ¾ L ½ (Đ), while 1 u k u Ú ¾ L 2 (Đ). To verify this fact, observe that, if 4 n2, u ¾ L ½ (Đ), if j n2, Ú ¾ L ½ (Đ). So it suffices to show that the two following products are in L 2 (Đ): (I) products of È the form 1 u k u with 1 k 4, i 4 n2 for each k i ¾ 1,, k, i=1 i j; (II) products of the form 1 u k u Ú with 1 k 4, i 4 n2 for each i ¾ 1,, k, j n2, È k i=1 i + j. We consider the case (I). By the Sobolev embedding theorem, we have, if i 4 n2, i u ¾ L 2n(n 2(4 i )) (Đ), while, if i = 4 n2, i u ¾ Ì 1p+½ L p (Đ). So, by Hölder s inequality, we have to show that È k i=1 (n 2(4 i))(2n) 12

19 PHASE FIELD IDENTIFICATION PROBLEM 597 and È k i=1 (n 2(4 i))(2n) 12, when i = 4 n2 for some i. Recalling that n 8, we get k i=1 n 2(4 i) 2n = k(n 8) + 2 È k i=1 i 2n È n k i=1 i 1 2n 2. We observe, moreover, that we have equality only in case k = 1 and 1 = 4. This can occur only when j = 4. So (I) is completely treated. We consider (II). Arguing as in the first case, we have to show that È k i=1 (n 2(4 i ))(2n) + (n 2( j ))(2n) 12 and È k i=1 (n 2(4 i))(2n) + (n 2( j ))(2n) 12, when i = 4 n2 for some i, or = j n2. We have: k i=1 n 2(4 i) 2n + n 2( j ) 2n = k(n 8) + 2 È k i=1 i + + n 2 j 2n n 8 + n 2n 1 2. Estimate (4.2) follows from the previous considerations and Sobolev embedding theorem, implying, for k j, S (k) Æ u L ½ (Đ) sup S (k) () C(R). u L ½ (Đ) Finally, estimate (4.3) is a consequence of the fact that, if S ¾ C j+1 (R), k j and maxu H 4 (Đ), u 1 H 4 (Đ) R, then maxu L ½ (Đ), u 1 L ½ (Đ) C 1 (R), so that S (k) Æ u S (k) Æ u 1 L ½ (Đ) C 2 (R)u u 1 L ½ (Đ). In the future we shall need the following simple lemma, the proof of which can be obtained by using Hölder and Young s inequalities. Lemma 4.3. Let X be a Banach space, ¾ R +, p ¾ ]1, +½[, z ¾ W 1, p (, ; X), with z() =. Then (4.4) z L ½ (,;X) 1p¼ z W 1, p (,;X), and (4.5) z L p (,;X) z W 1, p (,;X).

20 598 D. GUIDETTI AND A. LORENZI Lemma 4.4. Assume that (C1) (C2) are satisfied, S ¾ C 4 (R), ¾ H 4 (Đ). Let R ¾ R +, T, V 1 and V 2 be elements of W 1, p (, ; H 4 (Đ)) W 2, p (, ; L 2 (Đ)), V 1 () = V 2 (), max j¾1,2 V j W 1, p (,;H 4 (Đ)) R, ¾ W 4, p (Đ). Then S( + 1 V 1 )V 1 S( + 1 V 2 )V 2 W 1, p (,;H 2 (Đ)) C(R, T ) 1(2p) (V 1 V 2 W 1, p (,;H 4 (Đ)) + V 1 V 2 W 2, p (,;L 2 (Đ))). Proof. We begin with some useful estimates. First of all, (4.6) V j L ½ (,;H 4 (Đ)) C(R, T ). In fact, employing (4.4) V j L ½ (,;H 4 (Đ)) V j () H 4 (Đ) + 1 D t V j L ½ (,;H 4 (Đ)) V j () H 4 (Đ) + 1p¼ 1 D t V j W 1, p (,;H 4 (Đ)) = V j () H 4 (Đ) + 1p¼ D t V j L p (,;H 4 (Đ)) (1 T 1p¼ )R. Next, (4.7) + 1 V j L ½ (,;H 4 (Đ)) C(R, T ). In fact, + 1 V j L ½ (,;H 4 (Đ)) H 4 (Đ) + 1 V j L ½ (,;H 4 (Đ)) H 4 (Đ) + V j L ½ (,;H 4 (Đ)) R + T V j L ½ (,;H 4 (Đ)). We show that (4.8) 1 (V 1 V 2 ) L ½ (,;H 4 (Đ)) 1+1p¼ V 1 V 2 W 1, p (,;H 4 (Đ)). In fact, using (4.4), we have We show that 1 (V 1 V 2 ) L ½ (,;H 4 (Đ)) V 1 V 2 L ½ (,;H 4 (Đ)) 1+1p¼ V 1 V 2 W 1, p (,;H 4 (Đ)). (4.9) D t (V 1 V 2 ) L p (,;H 2 (Đ)) C(R, T ) 1(2p) (V 1 V 2 W 2, p (,;L 2 (Đ)) + V 1 V 2 W 1, p (,;H 4 (Đ))).

21 PHASE FIELD IDENTIFICATION PROBLEM 599 In fact, as H 2 (Đ) ¾ J 12 (L 2 (Đ), H 4 (Đ)), by Hölder s inequality, C D t (V 1 V 2 )(t) p H 2 (Đ) dt 1p D t (V 1 V 2 )(t) p2 L 2 (Đ) D t(v 1 V 2 )(t) p2 H 4 (Đ) dt 1p CD t (V 1 V 2 ) 12 L p (,;L 2 (Đ)) D t(v 1 V 2 ) 12 L p (,;H 4 (Đ)). Moreover, by Minkowski s inequality, so that D t (V 1 V 2 ) L p (,;L 2 (Đ)) 1p D t (V 1 V 2 )() L 2 (Đ) + 1 D 2 t (V 1 V 2 ) L p (,;L 2 (Đ)) 1p D t (V 1 V 2 )() L 2 (Đ) + D 2 t (V 1 V 2 ) L p (,;L 2 (Đ)) max1, T 1p¼ 1p V 1 V 2 W 2, p (,;L 2 (Đ)), D t (V 1 V 2 ) L p (,;H 2 (Đ)) C(T ) 1(2p) V 1 V 2 12 W 2, p (,;L 2 (Đ)) D t(v 1 V 2 ) 12 L p (,;H 4 (Đ)) C(T ) 1(2p) (V 1 V 2 W 2, p (,;L 2 (Đ)) + V 1 V 2 W 1, p (,;H 4 (Đ))). Now we prove the lemma. As V 1 () = V 2 (), we have S( + 1 V 1 )V 1 S( + 1 V 2 )V 2 W 1, p (,;H 2 (Đ)) = S ¼ ( + 1 V 1 )V S( + 1 V 1 )D t V 1 S ¼ ( + 1 V 2 )V 2 2 S( + 1 V 2 )D t V 2 L p (,;H 2 (Đ)). For almost every t ¾ [, ], employing Lemmata 4.1, 4.2, 4.3, (4.7), (4.8), we get S ¼ ( + 1 V 1 (t))v 1 (t) 2 + S( + 1 V 1 (t))d t V 1 (t) S ¼ ( + 1 V 2 (t))v 2 (t) 2 S( + 1 V 2 (t))d t V 2 (t) H 2 (Đ) V 1 (t)[s ¼ ( + 1 V 1 (t)) S ¼ ( + 1 V 2 (t))]v 1 (t) H 2 (Đ) + (V 1 (t) + V 2 (t))s ¼ ( + 1 V 2 (t))(v 1 (t) V 2 (t)) H 2 (Đ) + [S( + 1 V 1 (t)) S( + 1 V 2 (t))]d t V 1 (t) H 2 (Đ) + S( + 1 V 2 (t))[d t V 1 (t) D t V 2 (t)] H 2 (Đ) C(R, T )[1 (V 1 V 2 )(t) H 4 (Đ)(1 + D t V 1 (t) H 2 (Đ)) + V 1 (t) V 2 (t) H 2 (Đ) + D t (V 1 V 2 )(t) H 2 (Đ)]

22 6 D. GUIDETTI AND A. LORENZI C(R, T )[ 1+1p¼ V 1 V 2 W 1, p (,;H 4 (Đ))(1 + D t V 1 (t) H 2 (Đ)) We conclude, applying (4.5) and (4.9), that + V 1 (t) V 2 (t) H 2 (Đ) + D t (V 1 V 2 )(t) H 2 (Đ)]. S( + 1 V 1 )V 1 S( + 1 V 2 )V 2 W 1, p (,;H 2 (Đ)) C(R, T )[ 1+1p¼ V 1 V 2 W 1, p (,;H 4 (Đ))( 1p + D t V 1 L p (,;H 2 (Đ))) + V 1 V 2 L p (,;H 2 (Đ)) + D t (V 1 V 2 ) L p (,;H 2 (Đ))] C 1 (R, T ) 1(2p) (V 1 V 2 W 2, p (,;L 2 (Đ)) + V 1 V 2 W 1, p (,;H 4 (Đ))). In the following, we shall use without comment the elementary fact that, if T and «, C «, with C ¾ R +, depending only on «and T. Lemma 4.5. Assume that (C1) (C2) are satisfied and S ¾ C 4 (R). Let R ¾ R +, T, V 1 and V 2 be elements of W 1, p (, ; H 4 (Đ)) W 2, p (, ; H 3 B (Đ)) W 3, p (,;V ¼ ), such that V 1 () = V 2 (), D t V 1 () = D t V 2 (), max j¾1,2 (V j W 2, p (,;H 3 (Đ))+ V j W 1, p (,;H 4 (Đ))) R. Then S( + 1 V 1 )V 1 S( + 1 V 2 )V 2 W 2, p (,;H 1 (Đ)) C(R) (1p¼ )1(2p) (V 1 V 2 W 1, p (,;H 4 (Đ)) + V 1 V 2 W 2, p (,;H 3 (Đ)) + V 1 V 2 W 3, p (,;V ¼ )). (4.1) Proof. We start with a couple of useful estimates. First of all, D t (V 1 V 2 ) L p (,;H 1 (Đ)) V 1 V 2 W 2, p (,;H 3 (Đ)). In fact, as D t (V 1 V 2 )() =, D t (V 1 V 2 ) L p (,;H 1 (Đ)) = 1 D 2 t (V 1 V 2 ) L p (,;H 1 (Đ)) D 2 t (V 1 V 2 ) L p (,;H 1 (Đ)) V 1 V 2 W 2, p (,;H 3 (Đ)). Next, (4.11) D 2 t (V 1 V 2 ) L p (,;H 1 (Đ)) C 1(2p) (V 1 V 2 W 3, p (,;V ¼ ) + V 1 V 2 W 2, p (,;H 3 (Đ))). This estimate can be obtained with the same method used to prove (4.9), using the fact that H 1 (Đ) ¾ J 12 (V ¼, H 3 B (Đ)).

23 PHASE FIELD IDENTIFICATION PROBLEM 61 We pass to prove the stated result. As V 1 () = V 2 () and D t V 1 () = D t V 2 (), we have S( + 1 V 1 )V 1 S( + 1 V 2 )V 2 W 2, p (,;H 1 (Đ)) = D 2 t [S( + 1 V 1 )V 1 S( + 1 V 2 )V 2 ] L p (,;H 1 (Đ)) = S ¼¼ ( + 1 V 1 )V1 3 S ¼¼ ( + 1 V 2 )V [S ¼ ( + 1 V 1 )V 1 D t V 1 S ¼ ( + 1 V 2 )V 2 D t V 2 ] + S( + 1 V 1 )D 2 t V 1 S( + 1 V 2 )D 2 t V 2 L p (,;H 1 (Đ)). For almost every t in [, ] we have, employing Lemma 4.1, Lemma 4.2, (4.6), (4.8), (4.9), (4.4), S ¼¼ ( + 1 V 1 (t))v 1 (t) 3 S ¼¼ ( + 1 V 2 (t))v 2 (t) 3 + 3[S ¼ ( + 1 V 1 (t))v 1 (t)d t V 1 (t) S ¼ ( + 1 V 2 (t))v 2 (t)d t V 2 (t)] + S( + 1 V 1 (t))d 2 t V 1(t) S( + 1 V 2 (t))d 2 t V 2(t) H 1 (Đ) V 1 (t) 2 [S ¼¼ ( + 1 V 1 (t)) S ¼¼ ( + 1 V 2 (t))]v 1 (t) H 1 (Đ) + (V 1 (t) 2 + V 1 (t)v 2 (t) + V 2 (t) 2 )S ¼¼ ( + 1 V 2 (t))(v 1 (t) V 2 (t)) H 1 (Đ) + 3V 1 (t)[s ¼ ( + 1 V 1 (t)) S ¼ ( + 1 V 2 (t))]d t V 1 (t) H 1 (Đ) + 3(V 1 (t) V 2 (t))s ¼ ( + 1 V 2 (t))d t V 1 (t) H 1 (Đ) + 3V 2 (t)s ¼ ( + 1 V 2 (t))d t (V 1 V 2 )(t) H 1 (Đ) + [S( + 1 V 1 (t)) S( + 1 V 2 (t)])d 2 t V 1(t) H 1 (Đ) + S( + 1 V 2 (t))d 2 t (V 1 V 2 )(t) H 1 (Đ) C(R, T )[1 (V 1 V 2 )(t) H 4 (Đ)(1 + D t V 1 (t) H 1 (Đ) + D 2 t V 1(t) H 1 (Đ)) + V 1 (t) V 2 (t) H 1 (Đ) + V 1 (t) V 2 (t) H 4 (Đ)D t V 1 (t) H 1 (Đ) + D t (V 1 V 2 )(t) H 1 (Đ) + D 2 t (V 1 V 2 )(t) H 1 (Đ)] C(R, T )V 1 V 2 W 1, p (,;H 4 (Đ))[ 1+1p¼ (1 + D t V 1 (t) H 1 (Đ) + D 2 t V 1(t) H 1 (Đ)) So, using (4.1) (4.11), we get + 1p¼ (1 + D t V 1 (t) H 1 (Đ))] + D t (V 1 V 2 )(t) H 1 (Đ) + D 2 t (V 1 V 2 )(t) H 1 (Đ). S( + 1 V 1 )V 1 S( + 1 V 2 )V 2 W 2, p (,;H 1 (Đ)) C 1 (R, T ) 1p¼ V 1 V 2 W 1, p (,;H 4 (Đ)) + D t (V 1 V 2 )(t) L p (,;H 1 (Đ)) + D 2 t (V 1 V 2 )(t) L p (,;H 1 (Đ))

24 62 D. GUIDETTI AND A. LORENZI C 1 (R, T ) 1p¼ V 1 V 2 W 1, p (,;H 4 (Đ)) + V 1 V 2 W 2, p (,;H 3 (Đ)) which implies the conclusion. 5. The inverse problem + 1(2p) (V 1 V 2 W 3, p (,;V ¼ ) + V 1 V 2 W 2, p (,;H 3 (Đ))), In this section we will be concerned with the identification of the kernel k entering in the boundary value problem (.1) (.4). We start by introducing some notations, which will be useful in the following. As in the previous sections, we indicate with V ¼ the dual space of H 1 (Đ). Let s ¾ N, p ¾ [2, +½[. We set: (5.1) X s+3, p () := W s+3, p (, ; V ¼ ) C s+2 ([, ]; H 1 (Đ)) C s+1 ([, ]; H 2 B (Đ)) C s ([, ]; H 3 B (Đ)), (5.2) Y s+3, p () := W s+3, p (, ; V ¼ ) W s+2, p (, ; HB 3 (Đ)) W s+1, p (, ; HBB 4 (Đ)), (5.3) Z s+3, p () := X s+3, p () Y s+3, p () W s+1, p (, ) endowed with their natural norms X s+3, p (), Y s+3, p (), Z s+3, p () defined by (5.4) Û X s+3, p () := Û W s+3, p (,;V ¼ ) + 3 j=1 Û C s+3 j ([,];H j (Đ)), (5.5) (5.6) Ú Y s+3, p () := Ú W s+3, p (,;V ¼ ) + Ú W s+2, p (,;H 3 (Đ)) + Ú W s+1, p (,;H 4 (Đ)) + Ú C s+2 ([,];H 1 (Đ)), (u, Ú, h) Z s+3, p () := u X s+3, p () + Ú Y s+3, p () + h W s+1, p (,). The norm (5.5) is motivated by the fact that, owing to Proposition 3.5, in the case p 2, Y s+3, p () C s+2 ([, ]; H 1 (Đ)). Explicitly we will search for a solution (,, k) ¾ Z 4, p () to the following problem, where ¾ (, T ]: (5.7) D t ( + l)(t, x) (k ½ x )(t, x) = f (t, x), (t, x) ¾ [, ] Đ, D t (t, x) ½ x [ ½ x + () l](t, x) =, (t, x) ¾ [, ] Đ, (, x) = (x), (, x) = (x), x ¾ Đ, (t, x) = (t, x) = ½ x (t, x) =, (t, x) ¾ [, ] Đ, Đ ³(x)(t, x) dx = g(t), t ¾ [, ]. As far as the data are concerned, we shall make the following assumptions:

25 PHASE FIELD IDENTIFICATION PROBLEM 63 (D1) Đ is an open bounded subset of R n, n 7, lying on one side of its boundary Đ, which is a submanifold of R n of class C 4 ; (D2) ¾ C 5 (R), ³ ¾ H 1 (Đ); (D3) f ¾ W 3, p (, T ; H 1 (Đ)), g ¾ W 4, p (, T ), T ¾ R +, p ¾ [2, +½[; (D4) ¾ H 3 B (Đ), ¾ H 4 BB (Đ); (D5) Ú := ½ 2 x + ½ x [ ( ) l ] ¾ H 4 BB (Đ); (D6) u := f (, ) lú ¾ H 3 B (Đ); (D7) Ú 1 := ½ 2 x Ú + ½ x [ ¼ ( )Ú lu ] ¾ (H 3 B (Đ), H 4 BB (Đ)) 1 1p, p; (D8) Ê Đ ³(x)½ x (x) dx = ; (D9) k := g ¼¼ () + Ê Đ ³(x)[lÚ 1(x) D t f (, x)] dx Ê Đ ³(x)½ x (x) dx 1 ¾ R + ; (D1) u 1 := k ½ x + D t f (, ) lú 1 ¾ H 2 B (Đ); (D11) Ê Đ ³(x) (x) dx = g(), Ê Đ ³(x)u (x) dx = g ¼ (), Ê Đ ³(x)u 1(x) dx = g ¼¼ (); (D12) Ú 2 := BÚ 1 + ½ x [ ¼¼ ( )Ú 2 + ¼ ( )Ú 1 ] l½ x u 1 ¾ (V ¼, H 3 B (Đ)) 1p ¼, p, p ¼ = p(p 1). REMARK 5.1. Observe that the more complex boundary conditions (.4) can be replaced with the simpler ones in (5.7) under the assumption k() (cf. (D9) and (5.19). Indeed, since (,, k) ¾ Z 4, p (), the following relations hold: (5.8) k() + k¼ =, on (, ) Đ, µ (5.9) () = ¼ () =, on (, ) Đ. =, on (, ) Đ, Consequently, from (5.7), (5.8), (5.9) we easily deduce the simpler boundary conditions in (5.7). Assume now that a solution (,, k) ¾ Z 4, p () does exist. Then, differentiating with respect to t the first equation in (5.7), we obtain, for (t, x) ¾ [, ] Đ, (5.1) D 2 t (t, x) + l D2 t (t, x) = k()½ x (t, x) + (k ¼ ½ x )(t, x) + D t f (t, x). Differentiating again (5.1) and the second equation in (5.7) and setting (5.11) u := D t, Ú := D t, h := D t k, we get (5.12) Dt 2 u(t, x) + l D2 t Ú(t, x) (5.13) = k()½ x u(t, x) + h(t)½ x (x) + (h ½ x u)(t, x) + D 2 t f (t, x), (t, x) ¾ [, ] Đ, D t Ú(t, x) ½ x [ ½ x Ú + ¼ ()Ú lu](t, x) =, (t, x) ¾ [, ] Đ.

26 64 D. GUIDETTI AND A. LORENZI Moreover, (5.14) Ú(, ) = D t (, ) = Ú, u(, ) = f (, ) l D t (, ) = u, D 2 t (, ) = D t Ú(, ) = Ú 1. If Û ¾ V ¼, we set (5.15) [Û] := [Û, ³], where ³ denotes the function conjugate to ³. Therefore the last condition in (5.7) can be written in the form (5.16) [(t, )] = g(t), t ¾ [, ]. Applying ³ to (5.1), we obtain (5.17) g ¼¼ (t) + l[dt 2 (t, )] = k()[½ x (t, )] + [(h ½ x )(t, )] + [D t f (t, )], t ¾ [, ], implying (5.18) g ¼¼ () + l[ú 1 ] = k()[½ x ] + [D t f (, )]. It follows from (D8) and (D9) that (5.19) k() = g¼¼ () + l[ú 1 ] [D t f (, )] [½ x ] Next, = k. (5.2) D t u(, ) = D 2 t (, ) = k()½ x + D t f (, ) l D 2 t (, ) = k ½ x + D t f (, ) lú 1 = u 1. Finally, (5.21) u Ú (t, x) = (t, x) = ½ x Ú (t, x) =, (t, x) ¾ [, ] Đ. From these considerations, if (,, k) ¾ Z 4, p () solves (5.7), then triplet (u, Ú, h) ¾ Z 3, p (). Moreover, if A is the operator defined in (2.3) (2.31), B = A 2 and S := ¼,

27 PHASE FIELD IDENTIFICATION PROBLEM 65 then (u, Ú, h) ¾ Z 3, p () solves the problem Dt (5.22) 2 u(t, ) + l D2 t Ú(t, ) = k Au(t, ) h(t)a (h Au)(t, ) + Dt 2 f (t, ), D t Ú(t, ) BÚ(t, ) + A[S(( + 1 Ú)(t, ))Ú(t, )] l Au(t, ) =, u(, ) = u, D t u(, ) = u 1, Ú(, ) = Ú, g (3) (t) + l[dt 2 Ú(t, )] = k [Au(t, )] h(t)[a ] [(h Au)(t, )] + [D 2 t f (t, )], t ¾ [, ], t ¾ [, ], t ¾ [, ]. Lemma 5.2. Let assumptions (D1) (D12) be satisfied. Let (u, Ú, h) ¾ Z 3, p () solve system (5.22) and set := + 1 u, := + 1 Ú, k := k + 1 h. Then (,, k) ¾ Z 4, p () and solves problem (5.7). Proof. Since u = and Ú = ½Ú = on Đ, we deduce that also and satisfy such boundary conditions according to (D4). This implies that (,, k) ¾ Z 4, p () according to definitions (5.1) (5.3). Moreover, it is easy to check that the initial conditions (, ) =, (, ) = in (5.7) are satisfied. Since the second equation in (5.22) can be rewritten in the form D t D t (t, x) ½ x [ ½ x + () l](t, x), by virtue of (D5), the second equation in (5.7) is satisfied. Next, we have (5.23) D 2 t (k ½ x ) = D t (k ½ x + h ½ x ) = k ½ x u + h½ x + h ½ x u. Hence, the first equation in (5.22) can be rewritten in the form Owing to (D6), (D7) and (D1), we get D 2 t D t ( + l) (k ½ x )(t, x) D 2 t f. (5.24) D t D t ( + l) (k ½ x )(, ) = D t u(, ) + l D t Ú(, ) k ½ x = u 1 + lú 1 k ½ x = D t f (, ), i.e. even the first equation in (5.7) is satisfied. Finally, applying to the first equation in (5.22) and comparing the result with the final equation, we get, for all t ¾ [, T ], [D 3 t (t, )] = [D2 t u(t, )] = g(3) (t). Thus the final equation in (5.7) follows from (D11).

28 66 D. GUIDETTI AND A. LORENZI REMARK 5.2. in (5.22) we get If (u, Ú, h) is a solution of the system, from the last equation (5.25) h() = h := [ lú 2 k Au + Dt 2 f (, )] g (3) (). [A ] Lemma 5.3. Let X be a Banach space and let ¾ R +, p ¾ [1, +½[, h ¾ W 1, p (, ), z ¾ C([, ]; X). Then h z ¾ W 1, p (, ; X) and (5.26) Proof. h z W 1, p (,;X) ( 1p )h W 1, p (,)z C([,];X). As (h z)() = and h ¼ z L p (,;X) h ¼ L p (,)z L 1 (,;X), we get h z W 1, p (,;X) = (h z) ¼ L p (,;X) = h()z + h ¼ z L p (,;X) (5.27) h()z L p (,;X) + h ¼ L p (,)z L 1 (,;X) ( 1p h() + h ¼ L p (,))z C([,];X) ( 1p )h W 1, p (,)z C([,];X). Lemma 5.4. Let assumptions (D1) (D12) be satisfied and let U be the solution to the Cauchy problem (5.28) D 2 t U(t, ) + k AU(t, ) =, t ¾ [, T ], U(, ) = u, D t U(, ) = u 1. Then U ¾ Ì 3 j= C3 j ([, T ]; H j (Đ)). Proof. It follows from Proposition 2.9, with «= 12, p = +½. Lemma 5.5. Let assumption (D1) be satisfied. Let Ú ¾ HBB 4 (Đ), Ú 1 ¾ (HB 3(Đ), HBB 4 (Đ)) 1p ¼, p, Ú 2 ¾ (V ¼, HB 3(Đ)) 1p ¼, p. Then there exists V ¾ Y 3, p (T ) (cf. (5.2)) such that V () = Ú, D t V () = Ú 1, Dt 2V () = Ú 2. Proof. As Ú 2 BÚ 1 ¾ (V ¼, L 2 (Đ)) 1 1p, p, by virtue of Proposition in [15] there exists g ¾ W 1, p (, T ; V ¼ ) L p (, T ; L 2 (Đ)), such that g() = Ú 2 BÚ 1. Then we set f := Ú 1 BÚ + 1 g

29 PHASE FIELD IDENTIFICATION PROBLEM 67 and consider the solution V to problem (3.8). It is easily seen that all conditions (I) (IV) in Proposition 3.4 are fulfilled. So V ¾ Y 3, p (T ) and it is easily seen that D t V () = Ú 1, Dt 2V () = Ú 2. Now we are able to state and prove the first of the main results of our paper. Theorem 5.6. Let assumptions (D1) (D12) be fulfilled. Then there exists ¾ ], T ] such that the problem (5.7) has a solution (,, k) ¾ Z 4, p (). Proof. We define (5.29) Z 3, p () := (u, Ú, h) ¾ X 3, p () Y 3, p () W 1, p (, ): u() = u, D t u() = u 1, Ú() = Ú, D t Ú() = Ú 1, D 2 t Ú() = Ú 2, h() = h. Observe first that, owing to Lemmata 5.4 and 5.5, Z 3, p () is a nonempty closed subset of Z 3, p (). Then, for any fixed (U, V, H) ¾ Z 3, p () we consider the problem (5.3) Dt 2 u(t, ) + k Au(t, ) = Dt 2 f (t, ) H(t)A (H AU)(t, ) l Dt 2 V (t, ), t ¾ [, ], D t Ú(t, ) BÚ(t, ) = l AU(t, ) A[S(( + 1 V )(t, ))V (t, )], u(, ) = u, D t u(, ) = u 1, Ú(, ) = Ú, h(t)[a ] = g (3) (t) + [Dt 2 f (t, )] k [AU(t, )] [(H AU)(t, )] l[dt 2 t ¾ [, ]. V (t, )], It is easily seen that, for any fixed (U, V, H) ¾ Z 3, p (), problem (5.3) has a unique solution (u, Ú, h) in Z 3, p (). In fact, owing to (D3), (D4) and Lemma 5.3, D 2 t f (t, ) H( )A H AU ¾ W 1, p (, ; H 1 (Đ)). So, by virtue of Proposition 2.9, with «= 12, the problem (5.31) D 2 t z(t, ) + k Az(t, ) = D 2 t f (t, ) H(t)A z(, ) = u, D t z(, ) = u 1 (H AU)(t, ), t ¾ [, ], has a unique solution z ¾ C 2 ([, ]; H 1 (Đ)) C 1 ([, ]; H 2 B (Đ)) C([, ]; H 3 B (Đ)) W 3, p (, ; L 2 (Đ)). Moreover, by Proposition 3.5 and Corollary 2.11, the problem (5.32) D 2 t (t, ) + k A(t, ) = l Dt 2 V (t, ), (, ) =, D t (, ) = t ¾ [, ], has a unique solution belonging to W 3, p (,;V ¼ )C 2 ([,];H 1 (Đ))C 1 ([,];H 3 B (Đ)). Moreover, l AU ¾ C 2 ([,];V ¼ )C 1 ([,];L 2 (Đ)), so that l AU A[S(( +1 V ))V (t, )] ¾

30 68 D. GUIDETTI AND A. LORENZI W 2, p (, ; V ¼ ) W 1, p (, ; L 2 (Đ)) due to Lemmata 4.4 and 4.5. So, owing to assumptions (D5), (D7) and (D12), Proposition 3.4 allows to conclude that the second equation in (5.3) has a unique solution Ú ¾ Y 3, p (). Finally, as ³ ¾ H 1 (Đ) and AU ¾ C 1 ([, ]; V ¼ ), g (3) + [Dt 2 f ] k [AU] [H AU] l[dt 2V ] ¾ W 1, p (, ), owing to (D3) and Lemma 5.3. So problem (5.3) has a unique solution in Z 3, p (). It is not difficult to check that it belongs, in fact, to Z 3, p (). We denote this solution by S(U, V, H) = (S 1 (U, V, H), S 2 (U, V ), S 3 (U, V, H)), and stress that S 2 is independent of H. We observe that S is a (nonlinear) operator from Z 3, p () into itself. Clearly, to solve our identification problem (5.22) we have to look for a fixed point of S. Let R ¾ R + and (U j, V j, H j ) ¾ Z 3, p (), j ¾ 1, 2 and T, be such that (5.33) max j¾1,2 (U j, V j, H j ) Z 3, p () R. Taking Proposition 2.9, Corollary 2.11 and Lemma 5.3 into account, we get S 1 (U 1, V 1, H 1 ) S 1 (U 2, V 2, H 2 ) X 3, p () C((H 1 H 2 )A W 1, p (,;H 1 (Đ)) + H 1 A(U 1 U 2 ) W 1, p (,;H 1 (Đ)) (5.34) + (H 1 H 2 ) AU 2 W 1, p (,;H 1 (Đ)) + Dt 2 (V 1 V 2 ) L p (,;H 3 (Đ)) + Dt 2 (V 1 V 2 ) W 1, p (,;V ¼ ) + Dt 2 (V 1 V 2 ) C([,];H 1 (Đ))) C 1 (R) 1p (U 1 U 2 X 3, p () + H 1 H 2 W 1, p (,)) + C 1 (V 1 V 2 Y 3, p () + H 1 H 2 W 1, p (,)). Next, from Propositions 3.4 and 3.5, Lemmata 4.4 and 4.5 we obtain S 2 (U 1, V 1 ) S 2 (U 2, V 2 ) Y 3, p () C(U 1 U 2 W 2, p (,;H 1 (Đ)) + U 1 U 2 W 1, p (,;H 2 (Đ)) (5.35) + S( + 1 V 1 )V 1 S( + 1 V 2 )V 2 W 2, p (,;H 1 (Đ)) + S( + 1 V 1 )V 1 S( + 1 V 2 )V 2 W 1, p (,;H 2 (Đ))) C 2 (D 2 t (U 1 U 2 ) L p (,;H 1 (Đ)) + D t (U 1 U 2 ) L p (,;H 2 (Đ))) + C 2 (R) 1(2p) V 1 V 2 Y 3, p () C 2 1p U 1 U 2 X 3, p () + C 2 (R) 1(2p) V 1 V 2 Y 3, p ().

BANACH AND HILBERT SPACE REVIEW

BANACH AND HILBERT SPACE REVIEW BANACH AND HILBET SPACE EVIEW CHISTOPHE HEIL These notes will briefly review some basic concepts related to the theory of Banach and Hilbert spaces. We are not trying to give a complete development, but

More information

Let H and J be as in the above lemma. The result of the lemma shows that the integral

Let H and J be as in the above lemma. The result of the lemma shows that the integral Let and be as in the above lemma. The result of the lemma shows that the integral ( f(x, y)dy) dx is well defined; we denote it by f(x, y)dydx. By symmetry, also the integral ( f(x, y)dx) dy is well defined;

More information

Metric Spaces. Chapter 7. 7.1. Metrics

Metric Spaces. Chapter 7. 7.1. Metrics Chapter 7 Metric Spaces A metric space is a set X that has a notion of the distance d(x, y) between every pair of points x, y X. The purpose of this chapter is to introduce metric spaces and give some

More information

Separation Properties for Locally Convex Cones

Separation Properties for Locally Convex Cones Journal of Convex Analysis Volume 9 (2002), No. 1, 301 307 Separation Properties for Locally Convex Cones Walter Roth Department of Mathematics, Universiti Brunei Darussalam, Gadong BE1410, Brunei Darussalam

More information

Linear Algebra. A vector space (over R) is an ordered quadruple. such that V is a set; 0 V ; and the following eight axioms hold:

Linear Algebra. A vector space (over R) is an ordered quadruple. such that V is a set; 0 V ; and the following eight axioms hold: Linear Algebra A vector space (over R) is an ordered quadruple (V, 0, α, µ) such that V is a set; 0 V ; and the following eight axioms hold: α : V V V and µ : R V V ; (i) α(α(u, v), w) = α(u, α(v, w)),

More information

Inner Product Spaces

Inner Product Spaces Math 571 Inner Product Spaces 1. Preliminaries An inner product space is a vector space V along with a function, called an inner product which associates each pair of vectors u, v with a scalar u, v, and

More information

Properties of BMO functions whose reciprocals are also BMO

Properties of BMO functions whose reciprocals are also BMO Properties of BMO functions whose reciprocals are also BMO R. L. Johnson and C. J. Neugebauer The main result says that a non-negative BMO-function w, whose reciprocal is also in BMO, belongs to p> A p,and

More information

Mathematics Course 111: Algebra I Part IV: Vector Spaces

Mathematics Course 111: Algebra I Part IV: Vector Spaces Mathematics Course 111: Algebra I Part IV: Vector Spaces D. R. Wilkins Academic Year 1996-7 9 Vector Spaces A vector space over some field K is an algebraic structure consisting of a set V on which are

More information

Class Meeting # 1: Introduction to PDEs

Class Meeting # 1: Introduction to PDEs MATH 18.152 COURSE NOTES - CLASS MEETING # 1 18.152 Introduction to PDEs, Fall 2011 Professor: Jared Speck Class Meeting # 1: Introduction to PDEs 1. What is a PDE? We will be studying functions u = u(x

More information

MATH 425, PRACTICE FINAL EXAM SOLUTIONS.

MATH 425, PRACTICE FINAL EXAM SOLUTIONS. MATH 45, PRACTICE FINAL EXAM SOLUTIONS. Exercise. a Is the operator L defined on smooth functions of x, y by L u := u xx + cosu linear? b Does the answer change if we replace the operator L by the operator

More information

{f 1 (U), U F} is an open cover of A. Since A is compact there is a finite subcover of A, {f 1 (U 1 ),...,f 1 (U n )}, {U 1,...

{f 1 (U), U F} is an open cover of A. Since A is compact there is a finite subcover of A, {f 1 (U 1 ),...,f 1 (U n )}, {U 1,... 44 CHAPTER 4. CONTINUOUS FUNCTIONS In Calculus we often use arithmetic operations to generate new continuous functions from old ones. In a general metric space we don t have arithmetic, but much of it

More information

16.3 Fredholm Operators

16.3 Fredholm Operators Lectures 16 and 17 16.3 Fredholm Operators A nice way to think about compact operators is to show that set of compact operators is the closure of the set of finite rank operator in operator norm. In this

More information

Continued Fractions and the Euclidean Algorithm

Continued Fractions and the Euclidean Algorithm Continued Fractions and the Euclidean Algorithm Lecture notes prepared for MATH 326, Spring 997 Department of Mathematics and Statistics University at Albany William F Hammond Table of Contents Introduction

More information

Fixed Point Theorems

Fixed Point Theorems Fixed Point Theorems Definition: Let X be a set and let T : X X be a function that maps X into itself. (Such a function is often called an operator, a transformation, or a transform on X, and the notation

More information

Duality of linear conic problems

Duality of linear conic problems Duality of linear conic problems Alexander Shapiro and Arkadi Nemirovski Abstract It is well known that the optimal values of a linear programming problem and its dual are equal to each other if at least

More information

Linear Maps. Isaiah Lankham, Bruno Nachtergaele, Anne Schilling (February 5, 2007)

Linear Maps. Isaiah Lankham, Bruno Nachtergaele, Anne Schilling (February 5, 2007) MAT067 University of California, Davis Winter 2007 Linear Maps Isaiah Lankham, Bruno Nachtergaele, Anne Schilling (February 5, 2007) As we have discussed in the lecture on What is Linear Algebra? one of

More information

1 if 1 x 0 1 if 0 x 1

1 if 1 x 0 1 if 0 x 1 Chapter 3 Continuity In this chapter we begin by defining the fundamental notion of continuity for real valued functions of a single real variable. When trying to decide whether a given function is or

More information

A PRIORI ESTIMATES FOR SEMISTABLE SOLUTIONS OF SEMILINEAR ELLIPTIC EQUATIONS. In memory of Rou-Huai Wang

A PRIORI ESTIMATES FOR SEMISTABLE SOLUTIONS OF SEMILINEAR ELLIPTIC EQUATIONS. In memory of Rou-Huai Wang A PRIORI ESTIMATES FOR SEMISTABLE SOLUTIONS OF SEMILINEAR ELLIPTIC EQUATIONS XAVIER CABRÉ, MANEL SANCHÓN, AND JOEL SPRUCK In memory of Rou-Huai Wang 1. Introduction In this note we consider semistable

More information

n k=1 k=0 1/k! = e. Example 6.4. The series 1/k 2 converges in R. Indeed, if s n = n then k=1 1/k, then s 2n s n = 1 n + 1 +...

n k=1 k=0 1/k! = e. Example 6.4. The series 1/k 2 converges in R. Indeed, if s n = n then k=1 1/k, then s 2n s n = 1 n + 1 +... 6 Series We call a normed space (X, ) a Banach space provided that every Cauchy sequence (x n ) in X converges. For example, R with the norm = is an example of Banach space. Now let (x n ) be a sequence

More information

MATH10212 Linear Algebra. Systems of Linear Equations. Definition. An n-dimensional vector is a row or a column of n numbers (or letters): a 1.

MATH10212 Linear Algebra. Systems of Linear Equations. Definition. An n-dimensional vector is a row or a column of n numbers (or letters): a 1. MATH10212 Linear Algebra Textbook: D. Poole, Linear Algebra: A Modern Introduction. Thompson, 2006. ISBN 0-534-40596-7. Systems of Linear Equations Definition. An n-dimensional vector is a row or a column

More information

Functional analysis and its applications

Functional analysis and its applications Department of Mathematics, London School of Economics Functional analysis and its applications Amol Sasane ii Introduction Functional analysis plays an important role in the applied sciences as well as

More information

FUNCTIONAL ANALYSIS LECTURE NOTES: QUOTIENT SPACES

FUNCTIONAL ANALYSIS LECTURE NOTES: QUOTIENT SPACES FUNCTIONAL ANALYSIS LECTURE NOTES: QUOTIENT SPACES CHRISTOPHER HEIL 1. Cosets and the Quotient Space Any vector space is an abelian group under the operation of vector addition. So, if you are have studied

More information

4.6 Null Space, Column Space, Row Space

4.6 Null Space, Column Space, Row Space NULL SPACE, COLUMN SPACE, ROW SPACE Null Space, Column Space, Row Space In applications of linear algebra, subspaces of R n typically arise in one of two situations: ) as the set of solutions of a linear

More information

BOUNDED, ASYMPTOTICALLY STABLE, AND L 1 SOLUTIONS OF CAPUTO FRACTIONAL DIFFERENTIAL EQUATIONS. Muhammad N. Islam

BOUNDED, ASYMPTOTICALLY STABLE, AND L 1 SOLUTIONS OF CAPUTO FRACTIONAL DIFFERENTIAL EQUATIONS. Muhammad N. Islam Opuscula Math. 35, no. 2 (215), 181 19 http://dx.doi.org/1.7494/opmath.215.35.2.181 Opuscula Mathematica BOUNDED, ASYMPTOTICALLY STABLE, AND L 1 SOLUTIONS OF CAPUTO FRACTIONAL DIFFERENTIAL EQUATIONS Muhammad

More information

Vector Spaces II: Finite Dimensional Linear Algebra 1

Vector Spaces II: Finite Dimensional Linear Algebra 1 John Nachbar September 2, 2014 Vector Spaces II: Finite Dimensional Linear Algebra 1 1 Definitions and Basic Theorems. For basic properties and notation for R N, see the notes Vector Spaces I. Definition

More information

Systems with Persistent Memory: the Observation Inequality Problems and Solutions

Systems with Persistent Memory: the Observation Inequality Problems and Solutions Chapter 6 Systems with Persistent Memory: the Observation Inequality Problems and Solutions Facts that are recalled in the problems wt) = ut) + 1 c A 1 s ] R c t s)) hws) + Ks r)wr)dr ds. 6.1) w = w +

More information

x if x 0, x if x < 0.

x if x 0, x if x < 0. Chapter 3 Sequences In this chapter, we discuss sequences. We say what it means for a sequence to converge, and define the limit of a convergent sequence. We begin with some preliminary results about the

More information

The Dirichlet Unit Theorem

The Dirichlet Unit Theorem Chapter 6 The Dirichlet Unit Theorem As usual, we will be working in the ring B of algebraic integers of a number field L. Two factorizations of an element of B are regarded as essentially the same if

More information

HOMEWORK 5 SOLUTIONS. n!f n (1) lim. ln x n! + xn x. 1 = G n 1 (x). (2) k + 1 n. (n 1)!

HOMEWORK 5 SOLUTIONS. n!f n (1) lim. ln x n! + xn x. 1 = G n 1 (x). (2) k + 1 n. (n 1)! Math 7 Fall 205 HOMEWORK 5 SOLUTIONS Problem. 2008 B2 Let F 0 x = ln x. For n 0 and x > 0, let F n+ x = 0 F ntdt. Evaluate n!f n lim n ln n. By directly computing F n x for small n s, we obtain the following

More information

Math 4310 Handout - Quotient Vector Spaces

Math 4310 Handout - Quotient Vector Spaces Math 4310 Handout - Quotient Vector Spaces Dan Collins The textbook defines a subspace of a vector space in Chapter 4, but it avoids ever discussing the notion of a quotient space. This is understandable

More information

Section 6.1 - Inner Products and Norms

Section 6.1 - Inner Products and Norms Section 6.1 - Inner Products and Norms Definition. Let V be a vector space over F {R, C}. An inner product on V is a function that assigns, to every ordered pair of vectors x and y in V, a scalar in F,

More information

Chapter 5. Banach Spaces

Chapter 5. Banach Spaces 9 Chapter 5 Banach Spaces Many linear equations may be formulated in terms of a suitable linear operator acting on a Banach space. In this chapter, we study Banach spaces and linear operators acting on

More information

0 <β 1 let u(x) u(y) kuk u := sup u(x) and [u] β := sup

0 <β 1 let u(x) u(y) kuk u := sup u(x) and [u] β := sup 456 BRUCE K. DRIVER 24. Hölder Spaces Notation 24.1. Let Ω be an open subset of R d,bc(ω) and BC( Ω) be the bounded continuous functions on Ω and Ω respectively. By identifying f BC( Ω) with f Ω BC(Ω),

More information

Further Study on Strong Lagrangian Duality Property for Invex Programs via Penalty Functions 1

Further Study on Strong Lagrangian Duality Property for Invex Programs via Penalty Functions 1 Further Study on Strong Lagrangian Duality Property for Invex Programs via Penalty Functions 1 J. Zhang Institute of Applied Mathematics, Chongqing University of Posts and Telecommunications, Chongqing

More information

1. Let P be the space of all polynomials (of one real variable and with real coefficients) with the norm

1. Let P be the space of all polynomials (of one real variable and with real coefficients) with the norm Uppsala Universitet Matematiska Institutionen Andreas Strömbergsson Prov i matematik Funktionalanalys Kurs: F3B, F4Sy, NVP 005-06-15 Skrivtid: 9 14 Tillåtna hjälpmedel: Manuella skrivdon, Kreyszigs bok

More information

Matrix Representations of Linear Transformations and Changes of Coordinates

Matrix Representations of Linear Transformations and Changes of Coordinates Matrix Representations of Linear Transformations and Changes of Coordinates 01 Subspaces and Bases 011 Definitions A subspace V of R n is a subset of R n that contains the zero element and is closed under

More information

F. ABTAHI and M. ZARRIN. (Communicated by J. Goldstein)

F. ABTAHI and M. ZARRIN. (Communicated by J. Goldstein) Journal of Algerian Mathematical Society Vol. 1, pp. 1 6 1 CONCERNING THE l p -CONJECTURE FOR DISCRETE SEMIGROUPS F. ABTAHI and M. ZARRIN (Communicated by J. Goldstein) Abstract. For 2 < p

More information

THE BANACH CONTRACTION PRINCIPLE. Contents

THE BANACH CONTRACTION PRINCIPLE. Contents THE BANACH CONTRACTION PRINCIPLE ALEX PONIECKI Abstract. This paper will study contractions of metric spaces. To do this, we will mainly use tools from topology. We will give some examples of contractions,

More information

A REMARK ON ALMOST MOORE DIGRAPHS OF DEGREE THREE. 1. Introduction and Preliminaries

A REMARK ON ALMOST MOORE DIGRAPHS OF DEGREE THREE. 1. Introduction and Preliminaries Acta Math. Univ. Comenianae Vol. LXVI, 2(1997), pp. 285 291 285 A REMARK ON ALMOST MOORE DIGRAPHS OF DEGREE THREE E. T. BASKORO, M. MILLER and J. ŠIRÁŇ Abstract. It is well known that Moore digraphs do

More information

Adaptive Online Gradient Descent

Adaptive Online Gradient Descent Adaptive Online Gradient Descent Peter L Bartlett Division of Computer Science Department of Statistics UC Berkeley Berkeley, CA 94709 bartlett@csberkeleyedu Elad Hazan IBM Almaden Research Center 650

More information

EXISTENCE AND NON-EXISTENCE RESULTS FOR A NONLINEAR HEAT EQUATION

EXISTENCE AND NON-EXISTENCE RESULTS FOR A NONLINEAR HEAT EQUATION Sixth Mississippi State Conference on Differential Equations and Computational Simulations, Electronic Journal of Differential Equations, Conference 5 (7), pp. 5 65. ISSN: 7-669. UL: http://ejde.math.txstate.edu

More information

Theory of Sobolev Multipliers

Theory of Sobolev Multipliers Vladimir G. Maz'ya Tatyana O. Shaposhnikova Theory of Sobolev Multipliers With Applications to Differential and Integral Operators ^ Springer Introduction Part I Description and Properties of Multipliers

More information

GROUPS SUBGROUPS. Definition 1: An operation on a set G is a function : G G G.

GROUPS SUBGROUPS. Definition 1: An operation on a set G is a function : G G G. Definition 1: GROUPS An operation on a set G is a function : G G G. Definition 2: A group is a set G which is equipped with an operation and a special element e G, called the identity, such that (i) the

More information

Extremal equilibria for reaction diffusion equations in bounded domains and applications.

Extremal equilibria for reaction diffusion equations in bounded domains and applications. Extremal equilibria for reaction diffusion equations in bounded domains and applications. Aníbal Rodríguez-Bernal Alejandro Vidal-López Departamento de Matemática Aplicada Universidad Complutense de Madrid,

More information

No: 10 04. Bilkent University. Monotonic Extension. Farhad Husseinov. Discussion Papers. Department of Economics

No: 10 04. Bilkent University. Monotonic Extension. Farhad Husseinov. Discussion Papers. Department of Economics No: 10 04 Bilkent University Monotonic Extension Farhad Husseinov Discussion Papers Department of Economics The Discussion Papers of the Department of Economics are intended to make the initial results

More information

ON LIMIT LAWS FOR CENTRAL ORDER STATISTICS UNDER POWER NORMALIZATION. E. I. Pancheva, A. Gacovska-Barandovska

ON LIMIT LAWS FOR CENTRAL ORDER STATISTICS UNDER POWER NORMALIZATION. E. I. Pancheva, A. Gacovska-Barandovska Pliska Stud. Math. Bulgar. 22 (2015), STUDIA MATHEMATICA BULGARICA ON LIMIT LAWS FOR CENTRAL ORDER STATISTICS UNDER POWER NORMALIZATION E. I. Pancheva, A. Gacovska-Barandovska Smirnov (1949) derived four

More information

I. GROUPS: BASIC DEFINITIONS AND EXAMPLES

I. GROUPS: BASIC DEFINITIONS AND EXAMPLES I GROUPS: BASIC DEFINITIONS AND EXAMPLES Definition 1: An operation on a set G is a function : G G G Definition 2: A group is a set G which is equipped with an operation and a special element e G, called

More information

Numerical Verification of Optimality Conditions in Optimal Control Problems

Numerical Verification of Optimality Conditions in Optimal Control Problems Numerical Verification of Optimality Conditions in Optimal Control Problems Dissertation zur Erlangung des naturwissenschaftlichen Doktorgrades der Julius-Maximilians-Universität Würzburg vorgelegt von

More information

Undergraduate Notes in Mathematics. Arkansas Tech University Department of Mathematics

Undergraduate Notes in Mathematics. Arkansas Tech University Department of Mathematics Undergraduate Notes in Mathematics Arkansas Tech University Department of Mathematics An Introductory Single Variable Real Analysis: A Learning Approach through Problem Solving Marcel B. Finan c All Rights

More information

Introduction to the Finite Element Method

Introduction to the Finite Element Method Introduction to the Finite Element Method 09.06.2009 Outline Motivation Partial Differential Equations (PDEs) Finite Difference Method (FDM) Finite Element Method (FEM) References Motivation Figure: cross

More information

TOPIC 4: DERIVATIVES

TOPIC 4: DERIVATIVES TOPIC 4: DERIVATIVES 1. The derivative of a function. Differentiation rules 1.1. The slope of a curve. The slope of a curve at a point P is a measure of the steepness of the curve. If Q is a point on the

More information

Convex analysis and profit/cost/support functions

Convex analysis and profit/cost/support functions CALIFORNIA INSTITUTE OF TECHNOLOGY Division of the Humanities and Social Sciences Convex analysis and profit/cost/support functions KC Border October 2004 Revised January 2009 Let A be a subset of R m

More information

T ( a i x i ) = a i T (x i ).

T ( a i x i ) = a i T (x i ). Chapter 2 Defn 1. (p. 65) Let V and W be vector spaces (over F ). We call a function T : V W a linear transformation form V to W if, for all x, y V and c F, we have (a) T (x + y) = T (x) + T (y) and (b)

More information

Mathematical Methods of Engineering Analysis

Mathematical Methods of Engineering Analysis Mathematical Methods of Engineering Analysis Erhan Çinlar Robert J. Vanderbei February 2, 2000 Contents Sets and Functions 1 1 Sets................................... 1 Subsets.............................

More information

Subsets of Euclidean domains possessing a unique division algorithm

Subsets of Euclidean domains possessing a unique division algorithm Subsets of Euclidean domains possessing a unique division algorithm Andrew D. Lewis 2009/03/16 Abstract Subsets of a Euclidean domain are characterised with the following objectives: (1) ensuring uniqueness

More information

IRREDUCIBLE OPERATOR SEMIGROUPS SUCH THAT AB AND BA ARE PROPORTIONAL. 1. Introduction

IRREDUCIBLE OPERATOR SEMIGROUPS SUCH THAT AB AND BA ARE PROPORTIONAL. 1. Introduction IRREDUCIBLE OPERATOR SEMIGROUPS SUCH THAT AB AND BA ARE PROPORTIONAL R. DRNOVŠEK, T. KOŠIR Dedicated to Prof. Heydar Radjavi on the occasion of his seventieth birthday. Abstract. Let S be an irreducible

More information

ALMOST COMMON PRIORS 1. INTRODUCTION

ALMOST COMMON PRIORS 1. INTRODUCTION ALMOST COMMON PRIORS ZIV HELLMAN ABSTRACT. What happens when priors are not common? We introduce a measure for how far a type space is from having a common prior, which we term prior distance. If a type

More information

CONTINUED FRACTIONS AND PELL S EQUATION. Contents 1. Continued Fractions 1 2. Solution to Pell s Equation 9 References 12

CONTINUED FRACTIONS AND PELL S EQUATION. Contents 1. Continued Fractions 1 2. Solution to Pell s Equation 9 References 12 CONTINUED FRACTIONS AND PELL S EQUATION SEUNG HYUN YANG Abstract. In this REU paper, I will use some important characteristics of continued fractions to give the complete set of solutions to Pell s equation.

More information

Research Article Stability Analysis for Higher-Order Adjacent Derivative in Parametrized Vector Optimization

Research Article Stability Analysis for Higher-Order Adjacent Derivative in Parametrized Vector Optimization Hindawi Publishing Corporation Journal of Inequalities and Applications Volume 2010, Article ID 510838, 15 pages doi:10.1155/2010/510838 Research Article Stability Analysis for Higher-Order Adjacent Derivative

More information

PYTHAGOREAN TRIPLES KEITH CONRAD

PYTHAGOREAN TRIPLES KEITH CONRAD PYTHAGOREAN TRIPLES KEITH CONRAD 1. Introduction A Pythagorean triple is a triple of positive integers (a, b, c) where a + b = c. Examples include (3, 4, 5), (5, 1, 13), and (8, 15, 17). Below is an ancient

More information

Linear Codes. In the V[n,q] setting, the terms word and vector are interchangeable.

Linear Codes. In the V[n,q] setting, the terms word and vector are interchangeable. Linear Codes Linear Codes In the V[n,q] setting, an important class of codes are the linear codes, these codes are the ones whose code words form a sub-vector space of V[n,q]. If the subspace of V[n,q]

More information

1 = (a 0 + b 0 α) 2 + + (a m 1 + b m 1 α) 2. for certain elements a 0,..., a m 1, b 0,..., b m 1 of F. Multiplying out, we obtain

1 = (a 0 + b 0 α) 2 + + (a m 1 + b m 1 α) 2. for certain elements a 0,..., a m 1, b 0,..., b m 1 of F. Multiplying out, we obtain Notes on real-closed fields These notes develop the algebraic background needed to understand the model theory of real-closed fields. To understand these notes, a standard graduate course in algebra is

More information

PERTURBATION, INTERPOLATION, AND MAXIMAL REGULARITY

PERTURBATION, INTERPOLATION, AND MAXIMAL REGULARITY PERTURBATION, INTERPOLATION, AND MAXIMAL REGULARITY BERNHARD HAAK, MARKUS HAASE, AND PEER C. KUNSTMANN Abstract. We prove perturbation theorems for sectoriality and R sectoriality in Banach spaces, which

More information

Finite dimensional topological vector spaces

Finite dimensional topological vector spaces Chapter 3 Finite dimensional topological vector spaces 3.1 Finite dimensional Hausdorff t.v.s. Let X be a vector space over the field K of real or complex numbers. We know from linear algebra that the

More information

SOLUTIONS TO EXERCISES FOR. MATHEMATICS 205A Part 3. Spaces with special properties

SOLUTIONS TO EXERCISES FOR. MATHEMATICS 205A Part 3. Spaces with special properties SOLUTIONS TO EXERCISES FOR MATHEMATICS 205A Part 3 Fall 2008 III. Spaces with special properties III.1 : Compact spaces I Problems from Munkres, 26, pp. 170 172 3. Show that a finite union of compact subspaces

More information

MASSACHUSETTS INSTITUTE OF TECHNOLOGY 6.436J/15.085J Fall 2008 Lecture 5 9/17/2008 RANDOM VARIABLES

MASSACHUSETTS INSTITUTE OF TECHNOLOGY 6.436J/15.085J Fall 2008 Lecture 5 9/17/2008 RANDOM VARIABLES MASSACHUSETTS INSTITUTE OF TECHNOLOGY 6.436J/15.085J Fall 2008 Lecture 5 9/17/2008 RANDOM VARIABLES Contents 1. Random variables and measurable functions 2. Cumulative distribution functions 3. Discrete

More information

A Transmission Problem for Euler-Bernoulli beam with Kelvin-Voigt. Damping

A Transmission Problem for Euler-Bernoulli beam with Kelvin-Voigt. Damping Applied Mathematics & Information Sciences 5(1) (211), 17-28 An International Journal c 211 NSP A Transmission Problem for Euler-Bernoulli beam with Kelvin-Voigt Damping C. A Raposo 1, W. D. Bastos 2 and

More information

3. Mathematical Induction

3. Mathematical Induction 3. MATHEMATICAL INDUCTION 83 3. Mathematical Induction 3.1. First Principle of Mathematical Induction. Let P (n) be a predicate with domain of discourse (over) the natural numbers N = {0, 1,,...}. If (1)

More information

UNIVERSITETET I OSLO

UNIVERSITETET I OSLO NIVERSITETET I OSLO Det matematisk-naturvitenskapelige fakultet Examination in: Trial exam Partial differential equations and Sobolev spaces I. Day of examination: November 18. 2009. Examination hours:

More information

OPTIMAL CONTROL OF A COMMERCIAL LOAN REPAYMENT PLAN. E.V. Grigorieva. E.N. Khailov

OPTIMAL CONTROL OF A COMMERCIAL LOAN REPAYMENT PLAN. E.V. Grigorieva. E.N. Khailov DISCRETE AND CONTINUOUS Website: http://aimsciences.org DYNAMICAL SYSTEMS Supplement Volume 2005 pp. 345 354 OPTIMAL CONTROL OF A COMMERCIAL LOAN REPAYMENT PLAN E.V. Grigorieva Department of Mathematics

More information

ON COMPLETELY CONTINUOUS INTEGRATION OPERATORS OF A VECTOR MEASURE. 1. Introduction

ON COMPLETELY CONTINUOUS INTEGRATION OPERATORS OF A VECTOR MEASURE. 1. Introduction ON COMPLETELY CONTINUOUS INTEGRATION OPERATORS OF A VECTOR MEASURE J.M. CALABUIG, J. RODRÍGUEZ, AND E.A. SÁNCHEZ-PÉREZ Abstract. Let m be a vector measure taking values in a Banach space X. We prove that

More information

Metric Spaces. Chapter 1

Metric Spaces. Chapter 1 Chapter 1 Metric Spaces Many of the arguments you have seen in several variable calculus are almost identical to the corresponding arguments in one variable calculus, especially arguments concerning convergence

More information

PARTIAL DIFFERENTIAL EQUATIONS

PARTIAL DIFFERENTIAL EQUATIONS CHAPTER 1 PARTIAL DIFFERENTIAL EQUATIONS Many natural processes can be sufficiently well described on the macroscopic level, without taking into account the individual behavior of molecules, atoms, electrons,

More information

Notes on metric spaces

Notes on metric spaces Notes on metric spaces 1 Introduction The purpose of these notes is to quickly review some of the basic concepts from Real Analysis, Metric Spaces and some related results that will be used in this course.

More information

Pacific Journal of Mathematics

Pacific Journal of Mathematics Pacific Journal of Mathematics GLOBAL EXISTENCE AND DECREASING PROPERTY OF BOUNDARY VALUES OF SOLUTIONS TO PARABOLIC EQUATIONS WITH NONLOCAL BOUNDARY CONDITIONS Sangwon Seo Volume 193 No. 1 March 2000

More information

FACTORING POLYNOMIALS IN THE RING OF FORMAL POWER SERIES OVER Z

FACTORING POLYNOMIALS IN THE RING OF FORMAL POWER SERIES OVER Z FACTORING POLYNOMIALS IN THE RING OF FORMAL POWER SERIES OVER Z DANIEL BIRMAJER, JUAN B GIL, AND MICHAEL WEINER Abstract We consider polynomials with integer coefficients and discuss their factorization

More information

The determinant of a skew-symmetric matrix is a square. This can be seen in small cases by direct calculation: 0 a. 12 a. a 13 a 24 a 14 a 23 a 14

The determinant of a skew-symmetric matrix is a square. This can be seen in small cases by direct calculation: 0 a. 12 a. a 13 a 24 a 14 a 23 a 14 4 Symplectic groups In this and the next two sections, we begin the study of the groups preserving reflexive sesquilinear forms or quadratic forms. We begin with the symplectic groups, associated with

More information

The Heat Equation. Lectures INF2320 p. 1/88

The Heat Equation. Lectures INF2320 p. 1/88 The Heat Equation Lectures INF232 p. 1/88 Lectures INF232 p. 2/88 The Heat Equation We study the heat equation: u t = u xx for x (,1), t >, (1) u(,t) = u(1,t) = for t >, (2) u(x,) = f(x) for x (,1), (3)

More information

LINEAR ALGEBRA W W L CHEN

LINEAR ALGEBRA W W L CHEN LINEAR ALGEBRA W W L CHEN c W W L Chen, 1997, 2008 This chapter is available free to all individuals, on understanding that it is not to be used for financial gain, and may be downloaded and/or photocopied,

More information

THIN GROUPS OF FRACTIONS

THIN GROUPS OF FRACTIONS Contemporary Mathematics THIN GROUPS OF FRACTIONS Patrick DEHORNOY Abstract. A number of properties of spherical Artin Tits groups extend to Garside groups, defined as the groups of fractions of monoids

More information

SOME RESULTS ON HYPERCYCLICITY OF TUPLE OF OPERATORS. Abdelaziz Tajmouati Mohammed El berrag. 1. Introduction

SOME RESULTS ON HYPERCYCLICITY OF TUPLE OF OPERATORS. Abdelaziz Tajmouati Mohammed El berrag. 1. Introduction italian journal of pure and applied mathematics n. 35 2015 (487 492) 487 SOME RESULTS ON HYPERCYCLICITY OF TUPLE OF OPERATORS Abdelaziz Tajmouati Mohammed El berrag Sidi Mohamed Ben Abdellah University

More information

Monotone maps of R n are quasiconformal

Monotone maps of R n are quasiconformal Monotone maps of R n are quasiconformal K. Astala, T. Iwaniec and G. Martin For Neil Trudinger Abstract We give a new and completely elementary proof showing that a δ monotone mapping of R n, n is K quasiconformal

More information

Inner Product Spaces and Orthogonality

Inner Product Spaces and Orthogonality Inner Product Spaces and Orthogonality week 3-4 Fall 2006 Dot product of R n The inner product or dot product of R n is a function, defined by u, v a b + a 2 b 2 + + a n b n for u a, a 2,, a n T, v b,

More information

TRIPLE POSITIVE SOLUTIONS FOR BOUNDARY VALUE PROBLEM OF A NONLINEAR FRACTIONAL DIFFERENTIAL EQUATION. Communicated by Mohammad Asadzadeh

TRIPLE POSITIVE SOLUTIONS FOR BOUNDARY VALUE PROBLEM OF A NONLINEAR FRACTIONAL DIFFERENTIAL EQUATION. Communicated by Mohammad Asadzadeh Bulletin of the Iranian Mathematical Society Vol. 33 No. 2 (27), pp -. TRIPLE POSITIVE SOLUTIONS FOR BOUNDARY VALUE PROBLEM OF A NONLINEAR FRACTIONAL DIFFERENTIAL EQUATION R. DEHGHANI AND K. GHANBARI*

More information

These axioms must hold for all vectors ū, v, and w in V and all scalars c and d.

These axioms must hold for all vectors ū, v, and w in V and all scalars c and d. DEFINITION: A vector space is a nonempty set V of objects, called vectors, on which are defined two operations, called addition and multiplication by scalars (real numbers), subject to the following axioms

More information

Lecture Notes on Measure Theory and Functional Analysis

Lecture Notes on Measure Theory and Functional Analysis Lecture Notes on Measure Theory and Functional Analysis P. Cannarsa & T. D Aprile Dipartimento di Matematica Università di Roma Tor Vergata cannarsa@mat.uniroma2.it daprile@mat.uniroma2.it aa 2006/07 Contents

More information

ON A GLOBALIZATION PROPERTY

ON A GLOBALIZATION PROPERTY APPLICATIONES MATHEMATICAE 22,1 (1993), pp. 69 73 S. ROLEWICZ (Warszawa) ON A GLOBALIZATION PROPERTY Abstract. Let (X, τ) be a topological space. Let Φ be a class of realvalued functions defined on X.

More information

Settling a Question about Pythagorean Triples

Settling a Question about Pythagorean Triples Settling a Question about Pythagorean Triples TOM VERHOEFF Department of Mathematics and Computing Science Eindhoven University of Technology P.O. Box 513, 5600 MB Eindhoven, The Netherlands E-Mail address:

More information

Scalar Valued Functions of Several Variables; the Gradient Vector

Scalar Valued Functions of Several Variables; the Gradient Vector Scalar Valued Functions of Several Variables; the Gradient Vector Scalar Valued Functions vector valued function of n variables: Let us consider a scalar (i.e., numerical, rather than y = φ(x = φ(x 1,

More information

NOTES ON LINEAR TRANSFORMATIONS

NOTES ON LINEAR TRANSFORMATIONS NOTES ON LINEAR TRANSFORMATIONS Definition 1. Let V and W be vector spaces. A function T : V W is a linear transformation from V to W if the following two properties hold. i T v + v = T v + T v for all

More information

Essential spectrum of the Cariñena operator

Essential spectrum of the Cariñena operator Theoretical Mathematics & Applications, vol.2, no.3, 2012, 39-45 ISSN: 1792-9687 (print), 1792-9709 (online) Scienpress Ltd, 2012 Essential spectrum of the Cariñena operator Y. Mensah 1 and M.N. Hounkonnou

More information

Mathematics for Econometrics, Fourth Edition

Mathematics for Econometrics, Fourth Edition Mathematics for Econometrics, Fourth Edition Phoebus J. Dhrymes 1 July 2012 1 c Phoebus J. Dhrymes, 2012. Preliminary material; not to be cited or disseminated without the author s permission. 2 Contents

More information

Dynamical Systems Method for solving nonlinear operator equations

Dynamical Systems Method for solving nonlinear operator equations Dynamical Systems Method for solving nonlinear operator equations A.G. Ramm Mathematics Department, Kansas State University, Manhattan, KS 66506-2602, USA E:mail: ramm@math.ksu.edu Abstract Consider an

More information

LEARNING OBJECTIVES FOR THIS CHAPTER

LEARNING OBJECTIVES FOR THIS CHAPTER CHAPTER 2 American mathematician Paul Halmos (1916 2006), who in 1942 published the first modern linear algebra book. The title of Halmos s book was the same as the title of this chapter. Finite-Dimensional

More information

Multiple Solutions to a Boundary Value Problem for an n-th Order Nonlinear Difference Equation

Multiple Solutions to a Boundary Value Problem for an n-th Order Nonlinear Difference Equation Differential Equations and Computational Simulations III J. Graef, R. Shivaji, B. Soni, & J. Zhu (Editors) Electronic Journal of Differential Equations, Conference 01, 1997, pp 129 136. ISSN: 1072-6691.

More information

THE KADISON-SINGER PROBLEM IN MATHEMATICS AND ENGINEERING: A DETAILED ACCOUNT

THE KADISON-SINGER PROBLEM IN MATHEMATICS AND ENGINEERING: A DETAILED ACCOUNT THE ADISON-SINGER PROBLEM IN MATHEMATICS AND ENGINEERING: A DETAILED ACCOUNT PETER G. CASAZZA, MATTHEW FICUS, JANET C. TREMAIN, ERIC WEBER Abstract. We will show that the famous, intractible 1959 adison-singer

More information

Moreover, under the risk neutral measure, it must be the case that (5) r t = µ t.

Moreover, under the risk neutral measure, it must be the case that (5) r t = µ t. LECTURE 7: BLACK SCHOLES THEORY 1. Introduction: The Black Scholes Model In 1973 Fisher Black and Myron Scholes ushered in the modern era of derivative securities with a seminal paper 1 on the pricing

More information

Ideal Class Group and Units

Ideal Class Group and Units Chapter 4 Ideal Class Group and Units We are now interested in understanding two aspects of ring of integers of number fields: how principal they are (that is, what is the proportion of principal ideals

More information

ON SEQUENTIAL CONTINUITY OF COMPOSITION MAPPING. 0. Introduction

ON SEQUENTIAL CONTINUITY OF COMPOSITION MAPPING. 0. Introduction ON SEQUENTIAL CONTINUITY OF COMPOSITION MAPPING Abstract. In [1] there was proved a theorem concerning the continuity of the composition mapping, and there was announced a theorem on sequential continuity

More information

1. Let X and Y be normed spaces and let T B(X, Y ).

1. Let X and Y be normed spaces and let T B(X, Y ). Uppsala Universitet Matematiska Institutionen Andreas Strömbergsson Prov i matematik Funktionalanalys Kurs: NVP, Frist. 2005-03-14 Skrivtid: 9 11.30 Tillåtna hjälpmedel: Manuella skrivdon, Kreyszigs bok

More information