中文摘要：

提出了一种类似于flow-shop但又区别于flow-shop的semi-flow-shop生产调度问题,即根据各自的工艺要求,在同一生产线上以批为单位加工的工件可以跳过生产线上的一些工序,直接进入下道工序。根据实际需求,其调度目标不仅要考虑产品的提前/拖期,而且还要考虑设备的空闲。针对该问题,设计了一种改进的遗传算法,基因信息熵的概念被用于共享函数、自适应交叉概率和变异概率的计算,遗传算法的性能得以进一步改善。

英文摘要：

A semi-flow-shop scheduling problem,similar to flow-shop,was presented herein.The jobs which were processed as batches on identical production line may pass certain working procedure,and go into the next directly,according to the respective technological requirements.This scheduling problem was known from the flow-shop scheduling problem.According to the actual requirements of production scheduling,the objectives considered the earliness and tardiness,as well as the idle of machine tools.An improved genetic algorithm was proposed to solve this scheduling problem.The concept of gene entropy was used for the calculation of sharing function,adaptive probabilities of crossover and mutation.These measures improve the performance of the algorithm.