Dynamic scheduling I: real-time decision making using simulation

Based on a discrete-event simulation model, Simulationbased Real-time Decision-Making (SRDM) is an innovative approach to real-time, goal-directed decision-making. When applied to a flexible manufacturing system, SRDM makes better decisions than most fixed policies, such as deterministic, stochastic and manual. SRDM even improved over fixed policies optimized within a class of policies by OptQuest, in our numerical experiments. Compared to these fixed policies, SRDM shows greater improvement for more complex systems and is quite robust with respect to modeling errors. SRDM provides an improvement over fixed policies by its ability to implement adaptive policies. Since most real-time decisions in currently deployed manufacturing systems are made either manually or by using fixed policies, our results suggest that using SRDM instead could lead to significant improvement in operating performance.

[1]  Paul Glasserman,et al.  Gradient Estimation Via Perturbation Analysis , 1990 .

[2]  Stanley B. Gershwin,et al.  Manufacturing Systems Engineering , 1993 .

[3]  Jeffrey S. Smith,et al.  Simulation system for real-time planning, scheduling, and control , 1996, Proceedings Winter Simulation Conference.

[4]  E. Polak.  Introduction to linear and nonlinear programming , 1973 .

[5]  Sridhar Mahadevan,et al.  Optimizing Production Manufacturing Using Reinforcement Learning , 1998, FLAIRS.

[6]  Jeffrey S. Smith,et al.  Discrete event simulation for shop floor control , 1994, Proceedings of Winter Simulation Conference.

[7]  D. J. van der Zee.  Real-time adaptive control of multi-server bulk service processes , 2001 .

[8]  Halim Damerdji,et al.  Parametric Inference for Generalized Semi-Markov Processes , 1993, Proceedings of 1993 Winter Simulation Conference - (WSC '93).

[9]  James O. Henriksen,et al.  SLX: the X is for extensibility , 2000, WSC '00.

[10]  C. R. Glassey,et al.  Dynamic batching heuristic for simultaneous processing , 1991 .

[11]  I. McGregor.  The relationship between simulation and emulation , 2002, Proceedings of the Winter Simulation Conference.

[12]  Peter Norvig,et al.  Artificial Intelligence: A Modern Approach , 1995 .

[13]  Averill M. Law,et al.  Simulation optimization: simulation-based optimization , 2002, WSC '02.

[14]  Averill M. Law,et al.  Simulation-based optimization , 2000, 2000 Winter Simulation Conference Proceedings (Cat. No.00CH37165).

[15]  Martin A. Riedmiller,et al.  A Neural Reinforcement Learning Approach to Learn Local Dispatching Policies in Production Scheduling , 1999, IJCAI.

[16]  Kazuo Miyashita,et al.  Job-shop scheduling with genetic programming , 2000 .

[17]  Ian McGregor.  Equipment interface: the relationship between simulation and emulation , 2002, WSC '02.

[18]  Wayne J. Davis,et al.  A simulation-based controller for distributed discrete-event systems with application to flexible manufacturing , 1997, WSC '97.

[19]  J. P. Kelly,et al.  New advances and applications of combining simulation and optimization , 1996, Proceedings Winter Simulation Conference.

[20]  Sigurdur Olafsson,et al.  Simulation optimization , 2002, Proceedings of the Winter Simulation Conference.

[21]  Sigurdur Ólafsson,et al.  Simulation optimization: simulation optimization , 2002, WSC '02.

[22]  Andrew W. Moore,et al.  Value Function Based Production Scheduling , 1998, ICML.