TE-06: Optimization under Uncertainty and Applications II
Stream: Optimization under Uncertainty and Applications
Room: Moreau
Chair(s): Riccardo Cambini

An integrated multi-echelon robust closed- loop supply chain under imperfect quality production
Theodore Trafalis, Ismail Almaraj
In this paper, we consider a novel closed loop supply chain design consisting of multiple periods and multiple echelons. The models are considered under imperfect quality production with multiple uncertainties to provide meaningful solutions to practical problems. In addition, we assume that the screening is not always perfect, and inspection errors are more likely to take place in practice. We measure the amount of quality loss as conforming products deviate from the specification (target) value. In our study, we develop three robust counterparts models based on box, polyhedral, and combined.

Robust Two-stage Polynomial Optimization
Bissan Ghaddar
In this work we consider two-stage polynomial optimization problems under uncertainty. We combine tools from polynomial and robust optimization to provide a framework for general adjustable robust polynomial optimization problems. In particular we propose an iterative algorithm to build a sequence of (approximately) robustly feasible solutions with an improving objective value and verify robust feasibility or infeasibility of the resulting solution under a semialgebraic uncertainty set. We implement our approach for a use-case in energy systems to show the performance of the proposed approach.

Hierarchical Fleet Mix Problems with risk-aversion: a CVaR approach
Riccardo Cambini, Rossana Riccardi
In this paper a two-stage stochastic hierarchical workforce model is studied from both a theoretical and an algorithmic point of view. In the considered hierarchical model workforce units can be substituted by higher qualified ones; external workforce can also be hired to cover unfulfilled jobs. Demand for jobs is assumed to be stochastic. The results of a computational test are provided in order to validate the model.