some theorems of random operator equations随机算子方程的一些定理.pdf

IJMMS 30:9 (2002) 511–514 PII. S0161171202111252 © Hindawi Publishing Corp. SOME THEOREMS OF RANDOM OPERATOR EQUATIONS ZHU CHUANXI and XU ZONGBEN Received 17 January 2002 We investigate a class of random operator equations, generalize a famous theorem, and obtain some new results. 2000 Mathematics Subject Classification: 60H25, 47H10. Let E be a separable real Banach space, (E,B) a measurable space, where B denotes the σ -algebra generated by all open subsets in E , let (Ω,U,γ) be a complete probability measure space, where γ(Ω) = 1, let D be a bounded open set in X and ∂D the boundary of D in X . Let X be a cone in E , and let “≤”, “” be partial order of E . Lemma 1. When y 1, α 0, x ∈ X , and x ≠ θ, the following inequality holds: (y − 1)α+1x y α+1x −x. (1) Proof. Letting f (y) = y α+1 − 1 − (y − 1)α+1, where α 0, then f (y) = (α + 1)yα − (α + 1)(y − 1)α (2) α α 0 = (α + 1) y − (y − 1) (since y 1, then 0 y − 1 y, and α 0, obtaining 0 (y − 1)α yα , i.e., y α − (y − 1)α 0). Therefore f (y) is a monotonous increasing function. When y 1, we have f (y) f ( 1), and f ( 1) = 0. Hence f (y) 0, that is, y α+1 − 1 − (y − 1)α+1 0, that is, (y − 1)α+1 yα+1 − 1. (3) When x ∈ X , x ≠ θ, that is, x θ , we have (y − 1)α+1x y α+1x −x . Lemma 2 (see [1]). Let X be a closed convex subset of E , D a bounded open subset ¯ in X , and θ ∈ D . Suppose that A : Ω ×D → X is a random semiclosed 1-set-