|
|
|||||||||||||
|
|
|
|
|
|
|
|
|
||||||
|
|
|||||||||||||
     
|
The problem of optimal spacecraft scheduling is both important and difficult. Efficient utilization of spacecraft resources is essential, but the accompanying scheduling problems are often computationally intractable and are difficult to approximate because of the pressures of numerous interacting constraints. Although the fundamental requirements of optimal spacecraft scheduling are similar in many ways to those of other scheduling problems, encountered in commercial and industrial settings, the problems have been found to be notoriously difficult to solve in practical settings. Spike is an activity-oriented scheduling system developed at Space Telescope Institute for scheduling the Hubble Space Telescope. The overall approach adopted in Spike was inspired by advances in artificial intelligence research. Spike also incorporates several novel features including an innovative constraint representation and reasoning mechanism, and a new type of search technique motivated by research on artificial neural networks. This new approach is intended to tackle the four notable features of spacecraft scheduling that make it a difficult problem: interacting constraints, uncertainty, optimization criteria, and search. For more information on Spike and the Space Telescope Institute, please visit http://www.stsci.edu/spike.
© 2008 Franz Inc - Privacy Statement |
