Abstract

Fast (Incremental) Algorithms for Useful Classes of Simple Temporal Problems with Preferences

Fast (Incremental) Algorithms for Useful Classes of Simple Temporal Problems with Preferences

T. K. Satish Kumar

In this paper, we will provide a fast polynomial-time algorithm for solving simple temporal problems (STPs) with piecewise linear convex preference functions and a utilitarian objective function. Our algorithm is motivated by the study of the linear programming (LP)-dual of a given mincost circulation problem (MCCP). We will also show how this duality relationship between simple temporal problems with preferences (STPPs) and MCCPs leads to fast incremental algorithms for solving the former. Our algorithms bear important implications in planning, scheduling and execution monitoring scenarios where (partial) plans are subject to repeated changes, and the most preferred solutions to the underlying STPPs have to be computed and updated fast (incrementally).

URL: http://www.eecs.berkeley.edu/~tksk/PAPERS/tkskIJCAI2007.pdf