Including robustness considerations in the search phase of Many-Objective Robust Decision Making. (July 2018)

• Record Type: Journal Article
• Authors: Eker, Sibel; Kwakkel, Jan H.
• Is Part Of: Environmental modelling & software. Volume 105(2018)
• Page Start: 201
• Page End: 216
• Publication Date: 2018-07
• Subjects: Robust decision making -- Multi-objective optimization -- Scenario selection -- Scenario diversity; Environmental monitoring -- Computer programs -- Periodicals; Ecology -- Computer simulation -- Periodicals; Digital computer simulation -- Periodicals; Computer software -- Periodicals; Environmental Monitoring -- Periodicals; Computer Simulation -- Periodicals; Environnement -- Surveillance -- Logiciels -- Périodiques; Écologie -- Simulation, Méthodes de -- Périodiques; Simulation par ordinateur -- Périodiques; Logiciels -- Périodiques; Computer software; Digital computer simulation; Ecology -- Computer simulation; Environmental monitoring -- Computer programs; Periodicals; Electronic journals; 363.70015118
• DOI: 10.1016/j.envsoft.2018.03.029
• Languages: English
• ISSNs: 1364-8152

Abstract

Abstract: Many-Objective Robust Decision Making (MORDM) is a prominent model-based approach for dealing with deep uncertainty. MORDM has four phases: a systems analytical problem formulation, a search phase to generate candidate solutions, a trade-off analysis where different strategies are compared across many objectives, and a scenario discovery phase to identify the vulnerabilities. In its original inception, the search phase identifies optimal strategies for a single reference scenario for deep uncertainties, which may result in missing locally near-optimal, but globally more robust strategies. Recent work has addressed this issue by generating candidate strategies for multiple policy-relevant scenarios. In this paper, we incorporate a systematic scenario selection procedure in the search phase to consider both policy relevance and scenario diversity. The results demonstrate an increased tradeoff variety besides higher robustness, compared to the solutions found for a reference scenario. Future research can routinize multi-scenario search in MORDM with the aid of software packages. Highlights: MORDM aims at developing robust solutions but identifies them only under a reference scenario. We propose an extension to the search phase of MORDM. We generate the candidate solutions under multiple scenarios. We select these scenarios based on diversity and policy relevance. We obtain a wider variety of robustness tradeoffs compared to the reference case.