The Iterative Plan Optimization computer program automatically optimizes plans with respect to preferences expressed by human planners. This program incorporates a generalization of commonly occurring plan-quality metrics to provide a language for expression of preferences. The program implements a technique of iterative optimization that is a generalization of a prior technique of iterative repair, in which conflicts are detected and addressed one at a time until either no conflicts exist or a user-defined time limit has been exceeded. During iterative optimization, low-scoring preferences are detected and addressed individually until the maximum score is attained or until a user-defined time limit has been exceeded. A preference is a quality metric for a plan variable and can be improved by modifying the plan in a manner similar to that of repairing it. Plan modifications can include moving, creating, and deleting activities. For each preference, a domain-independent improvement-expert subprogram automatically generates the subset of modifications that could potentially improve the preference score.
This program was written by Steve Chien, Barbara Engelhardt, and Gregg Rabideau of Caltech for NASA's Jet Propulsion Laboratory. For further information, access the Technical Support Package (TSP) free on-line at www.nasatech.com/tsp under the Software category.
This software is available for commercial licensing. Please contact Don Hart of the California Institute of Technology at (818) 393-3425. Refer to NPO-20922.
This Brief includes a Technical Support Package (TSP).
Software for Iterative Optimization of Plans
(reference NPO-20922) is currently available for download from the TSP library.
Don't have an account? Sign up here.