c Applied Mathematics E-Notes, 7(2007), 9-15 Available free at mirror sites of http://www.math.nthu.edu.tw/∼amen/ ISSN 1607-2510 On A Second-Order Differential Inclusion With Constraints∗ Aurelian Cernea† Received 27 January 2006 Abstract We prove the existence of viable solutions to the Cauchy problem x00 ∈ F (x, x0 ) + f (t, x, x0 ), x(0) = x0 , x0 (0) = y0 , x(t) ∈ K, where K ⊂ Rn is a closed set, F is a set-valued map contained in the Fréchet subdifferential of a φ- convex function of order two and f is a Carathéodory map. 1 Introduction In this note we consider the second order differential inclusions of the form x00 ∈ F (x, x0) + f(t, x, x0), x(0) = x0, x0(0) = y0 , (1) where F (., .) : D ⊂ Rn × Rn → P(Rn ) is a given set-valued map, f(., ., .) : D1 ⊂ R × Rn × Rn → P(Rn ) is a given function and x0, y0 ∈ Rn . Existence of solutions of problem (1.1) that satisfy a constraint of the form x(t) ∈ K, ∀t, well known as viable solutions, has been studied by many authors, mainly in the case when the multifunction is convex valued and f ≡ 0 ([2], [6], [8], [10] etc.). Recently in [1], the situation when the multifunction is not convex valued is considered. More exactly, in [1] it is proved the existence of viable solutions of the problem (1) when F (., .) is an upper semicontinuous, compact valued multifunction contained in the subdifferential of a proper convex function. The result in [1] extends the result in [9] obtained for problems without constraints (i.e., K = Rn ). The aim of this note is to prove existence of viable solutions of the problem (1) in the case when the set-valued map F (., .) is upper semicontinuous compact valued and contained in the Fréchet subdifferential of a φ- convex function of order two. On one hand, since the class of proper convex functions is strictly contained into the class of φ- convex functions of order two, our result generalizes the result in [1]. On the other hand, our result may be considered as an extension of our previous viability result for second-order nonconvex differential inclusions in [5] obtained for a problem without perturbations (i.e., f ≡ 0). ∗ Mathematics Subject Classifications: 34A60 † Faculty of Mathematics and Informatics, University of Bucharest, Academiei 14, Romania 9 010014 Bucharest, 10 Differential Inclusion with Constraints The proof of our result follows the general ideas in [1] and [5]. We note that in the proof we pointed out only the differences that appeared with respect to the other approaches. The paper is organized as follows: in Section 2 we recall some preliminary facts that we need in the sequel and in Section 3 we prove our main result. 2 Preliminaries We denote by P(Rn ) the set of all subsets of Rn and by R+ the set of all positive real numbers. For > 0 we put B(x, ) = {y ∈ Rn; ||y − x|| < } and B(x, ) = {y ∈ Rn ; ||y − x|| ≤ }. With B we denote the unit ball in Rn. By cl(A) we denote the closure of the set A ⊂ Rn, by co(A) we denote the convex hull of A and we put ||A|| = sup{||a||; a ∈ A}. Let Ω ⊂ Rn be an open set and let V : Ω → R ∪ {+∞} be a function with domain D(V ) = {x ∈ Rn; V (x) < +∞}. DEFINITION 2.1. The multifunction ∂F V : Ω → P(Rn), defined as: ∂F V (x) = {α ∈ Rn, lim inf y→x V (y) − V (x) − hα, y − xi ≥ 0} if V (x) < +∞ ||y − x|| and ∂F V (x) = ∅ if V (x) = +∞ is called the Fréchet subdifferential of V . According to [4] the values of ∂F V are closed and convex. DEFINITION 2.2. Let V : Ω → R ∪ {+∞} be a lower semicontinuous function. We say that V is a φ-convex of order 2 if there exists a continuous map φV : (D(V ))2 ×R2 → R+ such that for every x, y ∈ D(∂F V ) and every α ∈ ∂F V (x) we have V (y) ≥ V (x) + hα, x − yi − φV (x, y, V (x), V (y))(1 + ||α||2)||x − y||2. (2) In [4], [7] there are several examples and properties of such maps. For example, according to [4], if M ⊂ R2 is a closed and bounded domain, whose boundary is a C 2 regular Jordan curve, the indicator function of M 0, if x ∈ M V (x) = IM (x) = +∞, otherwise is φ- convex of order 2. In what follows we assume the next assumptions. HYPOTHESIS 2.3. i) Ω = K × 0, where K ⊂ Rn is a closed set and O ⊂ Rn is a nonempty open set. ii) F (., .) : Ω → P(Rn ) is upper semicontinuous (i.e., ∀z ∈ Ω, ∀ > 0 there exists δ > 0 such that ||z − z 0 || < δ implies F (z 0 ) ⊂ F (z) + B) with compact values. iii) f(., ., .) : R ×Ω → Rn is a Carathèodory function, i.e., ∀(x, y) ∈ Ω, t → f(t, x, y) is measurable, for all t ∈ R f(t, .) is continuous and there exists m(.) ∈ L2 (R, R+) such that ||f(t, x, y)|| ≤ m(t) ∀(t, x, y) ∈ R × Ω. A. Cernea 11 iv) For all (t, x, v) ∈ R × Ω, there exists w ∈ F (x, v) such that lim inf h→0 Z 1 h2 d(x + hv + w+ h2 2 t+h f(s, x, v)ds, K) = 0. t v) There exists a proper lower semicontinuous φ- convex function of order two V : Rn → R ∪ {+∞} such that F (x, y) ⊂ ∂F V (y), 3 ∀(x, y) ∈ Ω. The Mmain Result In order to prove our result we need the following lemmas. LEMMA 3.1 ([1]). Assume that Hypotheses 2.3 i)-iv) are satisfied. Consider (x0, y0 ) ∈ Ω, r > 0 such that B(x0 , r) ⊂ O, M := sup{||F (t, x)||; (t, x) ∈ Ω0 := RT [K × B(y0 , r)] ∩ B((x0, y0 ), r)}, T1 > 0 such that 0 1 (m(s) + M + 1) < r3 , T2 = min{ 3(Mr+1) , 3(||y2r } and T ∈ (0, min{T1 , T2}). Then for every > 0 there exists 0 ||+r) η ∈ (0, ) and p ≥ 1 such that for all i = 1, ..., p − 1 there exists (hi, (xi , yi), wi) ∈ [η, ] × Ω0 × Rn with the following properties xi = xi−1 + hi−1yi−1 + h2i−1 wi−1 + 2 Z yi = yi−1 + hi−1wi−1, and hi−2 +hi−1 f(s, xi−1, yi−1 )ds ∈ K, hi−2 wi ∈ F (xi, yi ) + p−1 (xi , yi ) ∈ Ω0, X hi < T ≤ i=0 p X B, T hi . i=0 Pp−1 h2 Pp−1 Moreover, for > 0 sufficiently small we have i=0 2i ≤ i=0 hi < T . For k ≥ 1 and q = 1, ..., p denote by hkq the real number associated to = k1 and (t, x, y) = (hkq−1 , xq , yq ) given by Lemma 3.1. Define t0k = 0, tpk = T , tqp = hk0 + ... + hkq−1 q n and consider the sequence xk (.) : [tq−1 k , tk ] → R , k ≥ 1 defined by xk (0) = x0 , 1 q−1 2 xk (t) = xq−1 + (t − tq−1 k )yq−1 + (t − tk ) wq−1 + 2 Z t tq−1 k (t − s)f(s, xq−1 , yq−1)ds. LEMMA 3.2 ([1]). Assume that Hypotheses 2.3 i)-iv) are satisfied and consider xk (.) the sequence constructed above. Then there exists a subsequence, still denoted by xk (.) and an absolutely continuous function x(.) : [0, T ] → Rn such that i) xk (.) converges uniformly to x(.), ii) x0k (.) converges uniformly to x0(.), iii) x00k (.) converges weakly in L2 ([0, T ], Rn) to x00(.), 12 Differential Inclusion with Constraints iv) The sequence RT 0 P R tqk p q=1 tq−1 k 0 q−1 < x00k (s), f(s, xk (tq−1 ), x (t )) > ds converges to k k k k < x00(s), f(s, x(s), x0 (s)) > ds, v) For every t ∈ (0, T ) there exists q ∈ {1, ..., p} such that 0 q−1 lim d((xk (t), x0k (t), x00k (t) − f(t, xk (tq−1 k ), xk (tk ))), graph(F )) = 0, k→∞ vi) x(.) is a solution of the convexified problem x00 ∈ coF (x, x0) + f(t, x, x0), x(0) = x0, x0(0) = v0. We are now able to prove our result. THEOREM 3.3. Assume that Hypothesis 2.3 is satisfied. Then, for every (x0, y0 ) ∈ Ω there exist T > 0 and x(.) : [0, T ] → Rn a solution of problem (1) that satisfies x(t) ∈ K ∀t ∈ [0, T ]. PROOF. Let (x0 , y0) ∈ Ω and consider r > 0, T > 0 as in Lemma 3.1 and xk (.) : [0, T ] → Rn , x(.) : [0, T ] → Rn as in Lemma 3.2. Let φV the continuous function appearing in Definition 2.2. Since V (.) is continuous on D(V ) (e.g. [7]), by possibly decreasing r one can assume that for all y ∈ B(y0 , r) ∩ D(V ) |V (y) − V (y0 )| ≤ 1. Set S := sup{φv (y1 , y2 , z1, z2); yi ∈ B(y0 , r), zi ∈ [V (y0 ) − 1, V (y0 ) + 1], i = 1, 2}. From the statement vi) in Lemma 3.2 and Hypothesis 2.3 v) it follows that for almost all t ∈ [0, T ], x00(t) − f(t, x(t), x0(t)) ∈ ∂F V (x0 (t)). (3) Since the mapping x(.) is absolutely continuous, from (3) and Theorem 2.2 in [4] we deduce that there exists T3 > 0 such that the mapping t → V (x0 (t)) is absolutely continuous on [0, min{T, T3 }] and (V (x0(t)))0 = hx00(t), x00(t) − f(t, x(t), x0 (t))i a.e. [0, min{T, T3 }]. (4) Without loss of generality we may assume that T = min{T, T3 }. From (4) we have V (x0(T )) − V (y0 ) = Z T ||x00(s)||2 ds − 0 Z T hx00(s), f(s, x(s), x0 (s)i ds. 0 q On the other hand, for q = 1, ..., p and t ∈ [tq−1 k , tk ) q−1 0 q−1 0 q−1 x00k (t) − f(t, xk (tq−1 k ), xk (tk )) ∈ F (xk (tk ), xk (tk )) + 1 B kT and therefore 0 q−1 0 q−1 x00k (t) − f(t, xk (tq−1 k ), xk (tk )) ∈ ∂F V (xk (tk )) + 1 B. kT (5) A. Cernea 13 We deduce the existence of bqk ∈ B such that 0 q−1 x00k (t) − f(t, xk (tq−1 k ), xk (tk )) − bqk ∈ ∂F V (x0k (tq−1 k )). kT Taking into account Definition 2.2 we obtain * V (x0k (tqk )) −V (x0k (tq−1 k )) ≥ x00k (t) − 0 q−1 f(t, xk (tq−1 k ), xk (tk )) bq − k, kT Z tqk tq−1 k + x00k (s)ds 0 q 0 q−1 −φV x0k (tqk ), x0k (tq−1 ), V (x (t )), V (x (t )) k k k k k ! q 00 2 b q−1 q−1 0 k × 1+ xk (t) − f(t, xk (tk ), xk (tk )) − kT 2 × x0k (tqk ) − x0k (tq−1 k ) . q Using the fact that x00k (.) is constant on [tq−1 k , tk ] one may write Z tqk Z tqk bqk 00 00 00 V (x0k (tqk )) − V (x0k (tq−1 x )) ≥ hx (s), x (s)i ds − (s), ds k k k k kT tq−1 tq−1 k k Z tqk D E 0 q−1 − x00k (s), f(s, xk (tq−1 k ), xk (tk )) ds tq−1 k 0 q 0 q−1 −φV x0k (tqk ), x0k (tq−1 ), V (x (t )), V (x (t )) k k k k k ! q 00 2 b q−1 q−1 0 k × 1+ xk (t) − f(t, xk (tk ), xk (tk )) − kT 2 × x0k (tqk ) − x0k (tq−1 k ) . By adding on q the last inequalities we get Z T V (x0k (T )) − V (y0 ) ≥ ||x00k (s)||2 ds + a(k) + b(k) 0 − p Z X q=1 where b(k) = tqk tq−1 k D E 0 q−1 x00k (s), f(s, xk (tq−1 ), x (t )) ds, k k k (6) Z tq p X k 1 a(k) = − hx00k (s), bqk i ds, q−1 kT tk q=1 − p X q=1 0 q 0 q−1 φV x0k (tqk ), x0k (tq−1 k ), V (xk (tk ))), V (xk (tk ) ! q 2 00 2 b 0 q q−1 q−1 0 0 q−1 k × 1+ x (t) − f(t, x (t ), x (t )) − (t ) − x (t ) x . k k k k k k k k k kT 14 Differential Inclusion with Constraints On the other hand, one has Z tk p 1 X q ||bk||. ||x00k (s)||ds kT q=1 tq−1 k Z T Z T 1 1 1 00 ||xk (s)||ds ≤ [M + + m(s)]ds kT 0 kT 0 T q |a(k)| ≤ ≤ and |b(k)| ≤ p X S(1 + M 2)|| Z tq−1 k q=1 ≤ ≤ tqk x00k (s)ds||2 Z p Z p X 1 tk 1 T 00 ||x00k (s)||2 ds ≤ S(1 + M 2) ||xk (s)||2ds p−1 k k t 0 k q=1 Z T 1 1 [M + + m(s)]2 ds. S(1 + M 2) k T 0 S(1 + M 2) We infer that lim a(k) = lim b(k) = 0. k→∞ k→∞ Hence using also statement iv) in Lemma 3.2 and the continuity of the function V (.) by passing to the limit as k → ∞ in (6) we obtain V (x0 (T )) − V (y0 ) ≥ lim sup Z k→∞ T ||x00k (s)||2ds − Z 0 T 00 x (s), f(s, x( s), x0(s)) ds. (7) 0 Using (4) we infer that lim sup Z k→∞ T ||x0k(t)||2dt ≤ Z 0 T ||x00(t)||2dt 0 and, since x00k (.) converges weakly in L2 ([0, T ], Rn) to x00(.), by the lower semicontinuity of the norm in L2 ([0, T ], Rn) (e.g. Prop. III.30 in [3]) we obtain that lim k→∞ Z T ||x00k (t)||2dt = 0 Z T ||x00(t)||2dt 0 i.e., x00k (.) converges strongly in L2 ([0, T ], Rn). Hence, there exists a subsequence (still denoted) x00k (.) that converges pointwise to x00(.). From the statement v) in Lemma 3.2 it follows that d((x(t), x0(t), x00(t) − f(t, x(t), x0(t))), graph(F )) = 0 a.e. [0, T ]. and since by Hypothesis 2.4 graph(F ) is closed we obtain x00(t) ∈ F (x(t), x0(t)) + f(t, x(t), x0 (t)) a.e. [0, T ]. A. Cernea 15 In order to prove the viability constraint satisfied by x(.) fix t ∈ [0, T ]. There exists a sequence (tqk )k such that t = limk→∞ tqk . But limk→∞ ||x(t) − xk (tqk )|| = 0 and xk (tqk ) ∈ K. So the fact that K is closed gives x(t) ∈ K and the proof is complete. REMARK 3.4. If V (.) : Rn → R is a proper lower semicontinuous convex function then (e.g. [7]) ∂F V (x) = ∂V (x), where ∂V (.) is the subdifferential in the sense of convex analysis of V (.), and Theorem 3.3 yields the result in [1]. At the same time if in Theorem 3.3 f ≡ 0 then Theorem 3.3 yields the result in [5]. References [1] S. Amine, R. Morchadi and S. Sajid, Carathéodory perturbation of a second-order differential inclusions with constraints, Electronic J. Diff. Eq., 2005(2005), No. 114, 1-11. [2] A. Auslender and J. Mechler, Second order viability problems for differential inclusions, J. Math. Anal. Appl., 181(1984), 205-218. [3] H. Brezis, Analyse fonctionelle, theorie et applications, Masson, Paris, 1983. [4] T. Cardinali, G. Colombo, F. Papalini and M. Tosques, On a class of evolution equations without convexity, Nonlinear Anal., 28(1996), 217-234. [5] A. Cernea, On the existence of viable solutions for a class of second-order differential inclusions, Discuss. Math. Diff. Incl. Control Optim., 22(2002), 67-78. [6] B. Cornet, B. Haddad, Théoreme de viabilité pour inclusions differentielles du second order, Israel J. Math., 57(1987), 225-238. [7] M. Degiovanni, A. Marino and M. Tosques, Evolution equations with lack of convexity, Nonlinear Anal., 9(1995), 1401-1443. [8] T. X. D. Ha and M. Marques, Nonconvex second order differential inclusions with memory, Set-valued Anal., 5(1995), 71-86. [9] V. Lupulescu, Existence of solutions for nonconvex second-order differential inclusions, Appl. Math. E-Notes, 3(2003), 115-123. [10] L. Marco and J. A. Murillo, Viability theorems for higher-order differential inclusions, Set-valued Anal., 6(1998), 21-37. [11] M. Tosques, Quasi-autonomus parabolic evolution equations associated with a class of non linear operators, Ricerche Mat., 38(1989), 63-92.