Dual SDDP for risk-averse multistage stochastic programs

TL;DR

This paper introduces a dual formulation and an SDDP algorithm for risk-averse multistage stochastic programs, providing converging bounds and enhancing solution methods for complex decision-making under uncertainty.

Contribution

It develops a dual approach and an SDDP-based method specifically tailored for risk-averse multistage stochastic problems, improving solution guarantees.

Findings

The dual formulation enables new solution techniques.

The proposed SDDP algorithm converges to deterministic upper bounds.

The method is applicable to various risk-averse stochastic models.

Abstract

Risk-averse multistage stochastic programs appear in multiple areas and are challenging to solve. Stochastic Dual Dynamic Programming (SDDP) is a well-known tool to address such problems under time-independence assumptions. We show how to derive a dual formulation for these problems and apply an SDDP algorithm, leading to converging and deterministic upper bounds for risk-averse problems.