2332. Integrated optimization of multi-objective sequential processes
Invited abstract in session FA-5: Multiobjective Optimization 4: Complex Systems, Scalarizations, and Related Problems, stream Decision Theory and Multi-criteria Decision Making.
Friday, 8:45-10:15Room: H7
Authors (first author is the speaker)
| 1. | Jonas Hürter
|
| Department of Mathematics, University of Kaiserslautern-Landau | |
| 2. | Anita Schöbel
|
| Department of Mathematics, University of Kaiserslautern-Landau | |
| 3. | Philine Schiewe
|
| Department of Mathematics and Systems Analysis, Aalto University |
Abstract
Complex optimization problems are often decomposed as sequential processes of n stages: Firstly, a group of variables x1 is chosen optimally for a problem P1, then x1 is used to determine the feasible set of a subsequent problem P2(x1) for which an optimal solution x2 is chosen. Problem P3(x1, x2) depends on x1 and x2 and determines the variables x3 and so forth; up to a final problem Pn(x1, ..., x(n-1)). This approach is a heuristic for the optimization of a corresponding integrated problem.
Sequential processes and their integrated counterparts have been considered in the literature in the single-objective setting. Here, both the stages Pi of the sequential solution process and the integrated problem have only one objective function each.
In this talk we extend the framework of sequential processes and the corresponding integrated problem to the multi-objective context, where the sequential stages can have multiple objectives and the integrated problem considers all sequential objectives. We show that all solutions of such a process are always weakly efficient and identify conditions for them being (strictly) efficient. We also investigate lexicographic solutions and use such a process for deriving a cutting plane algorithm. We demonstrate the setting with regard to the sequential planning process in public transport.
Keywords
- Multi-Objective Programming
- Mobility
Status: accepted
Back to the list of papers