The research on pheromone based dynamic intelligent scheduling for semiconductor wafer fabrication

Abstract

Semiconductor wafer fabrication cries for really dynamic intelligent scheduling method, due to its uncertainty, large-scale, reentrance, mixed processing, unbalanced-workload, etc. On the basis of analysis on self-organizing behaviors of ant colony system, pheromone-based multi-agent-system (MAS) for dynamic scheduling for semiconductor wafer fabrication facilities is presented, and an algorithm is proposed. There are mainly two kinds of ant agents in the pheromone-based MAS, i.e. workpiece ant agents and resource ant agents (including machine ant agents, transportation ant agents and operator ant agents). The coordination and cooperation between ant agents through pheromone make it possible to realize dynamic intelligent scheduling for semiconductor wafer fabrication facilities.