中文摘要：

研究加权超前延误工件数问题．在单机存在非限制性共同宽容交货期（common due window,CDW）条件下，给出一个动态规划算法及一个近似算法；对单机限制性CDW中的某个特殊情况，给出一个多项式时间算法；对两台平行机非限制性CDW情况，构建一个伪多项式时间动态规划算法，证明其是一般意义下的NP—hard问题。

英文摘要：

There are a number of jobs with a common due window （CDW） to be processed on one certain machine facility. The objective is to minimize the total weighted number of early and tardy jobs for three different machine situations. A dynamic programming algorithm （DPA） and an approximate scheme are developed for the single machine situation with an unrestricted CDW. This DPA needs only the similar pseudo-polynomial computational time with that of knapsack problem. A polynomial approach is established for a special case of the single machine situation with a restricted CDW. For the situation of two parallel machines with an unrestricted CDW, another novel DPA still running in pseudo-polynomial time is presented, which shows that this situation is ordinary NP-hard.