Price decomposition in large-scale stochastic optimal control (1012.2092v2)

Abstract: We are interested in optimally driving a dynamical system that can be influenced by exogenous noises. This is generally called a Stochastic Optimal Control (SOC) problem and the Dynamic Programming (DP) principle is the natural way of solving it. Unfortunately, DP faces the so-called curse of dimensionality: the complexity of solving DP equations grows exponentially with the dimension of the information variable that is sufficient to take optimal decisions (the state variable). For a large class of SOC problems, which includes important practical problems, we propose an original way of obtaining strategies to drive the system. The algorithm we introduce is based on Lagrangian relaxation, of which the application to decomposition is well-known in the deterministic framework. However, its application to such closed-loop problems is not straightforward and an additional statistical approximation concerning the dual process is needed. We give a convergence proof, that derives directly from classical results concerning duality in optimization, and enlghten the error made by our approximation. Numerical results are also provided, on a large-scale SOC problem. This idea extends the original DADP algorithm that was presented by Barty, Carpentier and Girardeau (2010).