炉管区是半导体制造过程中用于热处理的重要区域。在多品种不同交货期的晶圆产品连续到达的环境下,提出了优化目标为极小化E/T(提前/延迟)值的WCRH(加权成本率)算法。该算法根据产品的交货期实时更新晶圆产品的加工优先级;通过计算和比较加权成本率来确定哪一种晶圆优先加工以及投入加工的时间。仿真结果表明,该算法在减小E/T值和缩短加工完成时间方面,优于文献中提及的启发式算法。
Furnace district is one of the most important areas in semiconductor fabrication for heat- treating. To resolve the scheduling problem of the furnace where existing multi-products and different due dates during the batch processing, a dynamic scheduling algorithm was proposed named WCRH (weighted cost rate heuristic) for minimizing the E/T (earliness and tardiness) based on delivery restriction. It renews the operation priority of arriving wafer according to the due dates and decides the scheduling plan based on the comparison of weighted cost rate. Experiment results indicate that the algorithm is more efficient in decreasing E/T and completed time than heuristic algorithms.