Multi-Level Robust Optimization: Theory, Algorithms and Practice
Abstract
In practice, decision-problems which concern planning horizons spanning several months or years will typically involve various sources of uncertainty, and in many cases these uncertainties will impact decisions over differing time-scales and in different levels of the problem. Accurately accounting for these uncertainties in an optimization model is challenging with traditional approaches, as all uncertainties would generally be treated simultaneously. A good practical example of this type of problem appears in the simultaneous expansion planning of power generation sources and expansion planning of power transmission networks, where the short-term stochastic properties of production and demand will interact with long-term uncertainties on the price of fuel, future peak-demand, construction costs and policy uncertainty. The complexity is compounded further when both of these problems are considered jointly as a co-optimization problem for a holistic approach to energy-system planning, the so-called generation and transmission expansion planning problem. In this case, the resulting problem involves uncertainties, the different time-scales on which these feature, and often multiple levels with potentially different objectives. The proposed project will develop novel optimization theory and methodologies to tackle this general class of problems, namely multi-level robust optimization problems (and potentially its extension to multi-objective problems), in a PhD project over a period of 3 years. Efficient algorithmic approaches to tackling this new class of optimization problem will then be investigated.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Sep 19, 2018
- Source ID
- FA95501817003
Entities
People
- Kerem Akartunalı
Organizations
- Air Force Office of Scientific Research
- United States Air Force
- University of Strathclyde