Stochastic Set Partitioning Methods for Operational Planning of Aircraft

摘要

In the last two years, the project has focused on developing tools for representing, mathematically and in software, complex operational problems. We have developed a formal problem class called Dynamic Resource Transformation Problems which captures in a compact way a broad range of complex applications. We have developed a simulation library that captures the specific features of this problem class, in particular the ability to make decisions within the simulation. Finally, we have developed an algorithmic metastrategy that provides an effective way of solving these problems. We consider this work to be an important breakthrough in our ability to model and solve the types of operational challenges that arise in military airlift operations, as well as a variety of civilian applications such as railroads and trucking, as well as air traffic control.