SCHEDULING AND RESCHEDULING WITH ITERATIVE REPAIR

被引:73
作者
ZWEBEN, M
DAVIS, E
DAUN, B
DEALE, MJ
机构
[1] NASA,AMES RES CTR,ARTIFICIAL INTELLIGENCE BRANCH,MOFFETT FIELD,CA 94035
[2] LOCKHEED SPACE OPERAT CO,LOCKHEED AI CTR,PALO ALTO,CA
来源
IEEE TRANSACTIONS ON SYSTEMS MAN AND CYBERNETICS | 1993年 / 23卷 / 06期
关键词
D O I
10.1109/21.257756
中图分类号
TP3 [计算技术、计算机技术];
学科分类号
0812 [计算机科学与技术];
摘要
The paper describes the GERRY scheduling and rescheduling system being applied to coordinate Space Shuttle Ground Processing. The system uses constraint-based iterative repair, a technique that starts with a complete but possibly flawed schedule and iteratively improves it by using constraint knowledge within repair heuristics. In this paper we explore the tradeoff between the informedness and the computational cost of several repair heuristics. It is shown empirically that some knowledge can greatly improve the convergence speed of a repair-based system, but that too much knowledge, such as the knowledge embodied within the MIN-CONFLICTS look-ahead heuristic, can overwhelm a system and result in degraded performance.
引用
收藏
页码:1588 / 1596
页数:9
相关论文
共 32 条
[1]
BIEFELD E, 1991, P IJCAI 91 SYDNEY
[2]
Bresina J., 1990, P AAAI 90
[3]
CHAPMAN D, 1987, ARTIF INTELL, V32
[4]
CURRIE K, 1991, ARTIF INTELL, V52
[5]
DAVIS E, 1987, ARTIF INTELL, V32
[6]
ESKEY M, 1990, P AAAI 90 BOSTON
[7]
FIKES RE, 1972, ARTIF INTELL, V3
[8]
FOX M, 1984, EXP SYST, V1
[9]
FOX MS, 1987, CONSTRAINT DIRECTED
[10]
FREUDER EC, 1982, J ACM, V29