中文摘要：

研究了一类从无缝钢管生产作业中提炼出的新的并行流水车间调度问题．该问题具有工件无等待、工序之间存在运输时间、设备需要调整时间等特点．这些特点使得问题变得相当复杂．建立了大规模的混合整数规划模型，通过提出的变换方法简化和降低了模型的规模．针对此模型，提出并开发了适合此问题的遗传算法．通过实验比较六种规则调度方法及遗传算法的性能．计算结果表明，六种规则调度中最好的方法是SPT，而遗传算法调度的性能优于SPT．

英文摘要：

This paper considers a class of parallel flowshop scheduling problem, which is abstracted from the production of seamless steel pipe and characterized by no-wait, transfer times between operations, machine dependent setup times, etc. These characteristics complicate the problem. We formulate it as a large scale mixed integer programming model. We present a transform method to simplify and decrease the scale of the model. A problem specific genetic algorithm is then proposed. We test the performance of six rule scheduling methods and genetic algorithm. The test results show that SPT is the best one of the six rule scheduling methods and the genetic algorithm is better than SPT.