Effective redundant constraints for online scheduling

  • Authors:
  • Lise Getoor;Greger Ottosson;Markus Fromherz;Björn Carlson

  • Affiliations:
  • CS Dept., Stanford Univeristy, Stanford, CA;CS Dept., Uppsala University, Uppsala, Sweden;Xerox Palo Alto Research Center, Palo Alto, CA;Xerox Palo Alto Research Center, Palo Alto, CA

  • Venue:
  • AAAI'97/IAAI'97 Proceedings of the fourteenth national conference on artificial intelligence and ninth conference on Innovative applications of artificial intelligence
  • Year:
  • 1997

Quantified Score

Hi-index 0.00

Visualization

Abstract

The use of heuristics as a means to improve constraint solver performance has been researched widely. However, most work has been on problem-independent heuristics (e.g., variable and value ordering), and has focused on offline problems (e.g., one-shot constraint satisfaction). In this paper, we present an online scheduling problem for which we are developing a real-time scheduling algorithm. While we can and do use generic heuristics in the scheduler, here we focus on the use of domain-specific redundant constraints to effectively approximate optimal offline solutions. We present a taxonomy of redundant domain constraints, and examine their impact on the effectiveness of the scheduler. We also describe several techniques for generating redundant constraints, which can be applied to a large class of job shop scheduling problems.