New efficient numerical procedures for solving stochastic variational problems with a priori maximum pointwise error estimates (Q864651): Difference between revisions
From MaRDI portal
Latest revision as of 14:00, 25 June 2024
scientific article
Language | Label | Description | Also known as |
---|---|---|---|
English | New efficient numerical procedures for solving stochastic variational problems with a priori maximum pointwise error estimates |
scientific article |
Statements
New efficient numerical procedures for solving stochastic variational problems with a priori maximum pointwise error estimates (English)
0 references
12 February 2007
0 references
Consider the pathwise minimization of the functional \[ J(x,u) = \int^b_a f(t,x(t),u(t)) \,dt + \tfrac{1}{2} k [x(b)]^2 \] with piecewise-continuous control variable \(u \in H_{a,b}\) (a Cameron-Martin space of absolutely continuous functions), where process \(x\) satisfies the state-independent ordinary stochastic differential equation (SDE) \[ dx(t) = u(t) dt + \sigma (t) dW(t) \] driven by an \(m\)-dimensional Wiener process \(W\). The authors discuss some theory of pathwise optimal control problem \(\min_{u \in H_{a,b}} J(x,u) \) and related numerical algorithms. They claim that their major algorithm (borrowed from deterministic theory) has pathwise global pointwise error of order \(h^{3/2}\) along equidistant partitions with nonrandom mesh size \(h\). However, the leading error coefficients are not determined precisely. They find an approximation to the critical point solution of the variational problem using relations derived from setting to zero the directional derivative of the cost functional in the direction of simple test functions. Basicly, their examples and analysis restrict to the case of quadratic functionals \(J\) in \(x\) and \(u\) and time-shifted Brownian motion (note that the noise \(x\) is governed by an SDE with additive noise, state-dependent coefficients and explicitly known strong solution). To prove major results, the theory of deterministic Euler-Lagrange equations and Lagrange multipliers is applied to convert the constraint problem into a nonconstraint one. It is interesting that the theory of nonanticipative integration is not needed here since anticipative integrands only appear in ordinary integrals of related control problems.
0 references
stochastic variational problems
0 references
pathwise optimal control
0 references
numerical methods
0 references
Wiener process
0 references
stochastic differential equation
0 references
quadratic control
0 references
0 references
0 references