Event Type:

Applied Mathematics and Computation Seminar

Date/Time:

Friday, October 23, 2015 - 12:00 to 13:00

Location:

GLK 115

Event Link:

Guest Speaker:

Institution:

OSU Industrial and Manufacturing Engineering

Abstract:

The strategic design of transportation and logistics networks consists of determining facility locations and selecting paths for flows in a network. These strategic decisions are affected by constraints imposed at tactical and operational levels of decision making. This problem is usually modeled using a multi-stage approach where solutions to a mathematical model for a decision problem at a higher level of decision making are used as input to mathematical models for lower level decision problems. However, this affects optimality and applicability of the solutions obtained. Also, any decisions made based on deterministic models have no guarantee of performance for any future realization of any uncertain design parameter. We present an integrated modeling and solution approach that explicitly considers parameter uncertainty to obtain solutions that can be implemented in practice with minimal reconfiguration. A robust optimization counterpart for a large-scale network design optimization modeling approach is developed to obtain a more realistic strategic network design optimization model that balances practicality and solvability. The application of this method for the design of full truckload relay networks is presented.