DFG Project Robust Optimization Algorithms SPP 1307
(first funding period)
In praxis exact solutions are often useless if constraints can only be approximated or have to be assumed at the time when the calculation is done. Sometimes an exact solution can not be realised with the same exactness or unexpected disruptions can occur. Within the DFG Priority Programme "Algorithm Engineering" the Research Group Optimization deals with the development of robust algorithms that can find applicable solutions also in these cases.
To do this one part is to develop the theory of robust optimization further to applicability in praxis and on the other hand to evaluate experimental results to integrate them again in the theoretical considerations. In particular instances of the following fields are analysed:
Arising from this are the following five fields of work:
Task 1: A bicriteria approach for robustness
The search for pareto-solutions in relation to objective value and robustness.
Task 2: Multi-stage robustness
The analysis of successive arising disruptions.
Task 3: Locating a robust solution in solution space
The treatment of robustness as a problem of location theory.
Task 4: Generic robust algorithms for robust optimization
The development of generic algorithms.
Task 5: Classification, complexity and price of robustness
The classification and systematic collection of the theoretical and practical results reached thus far in the field of robust optimization.
The project started in November 2009 and finishes in October 2011.
