GPRs搭接网络分解优化定理在流水作业中的应用

本文发现在GPRs搭接网络传统算法中，针对某些可分解的关键工序，通过工序的分解会产生分解悖论和咖啡时间悖论。通过对这些悖论现象的分析研究，发现其存在帕累托改进。对此，提出了两个分解优化定理及网络的分解优化方法，使网络的总工期和总时差的分布都得到了优化，为项目WBS和资源优化提供了更科学的，更充足的条件。并将该分解优化定理同流水作业原理相结合，用实例证明了该方法的可操作性，为流水作业中施工段的划分提供了科学的优化方法。

Application of GPRs Network Division Optimization Theorem in the Flow Process Network

In this paper, the critical activities decomposition paradox and the total floats paradox in the traditional algorithm of GPRs multi-time difference network are found. The critical activities decomposition paradox is that the critical activity which is decomposed into two activities with FTS=0 logical relation will lead to the total duration shortened. The total floats paradox is that activities which are decomposed will increase the total float. The reasons of these two paradoxes and propose critical activities decomposition optimization theorem and total float decomposition optimization theorem are analyzed. The new methods make the total project duration and the distribution of total time of the network optimized. They can also provide more scientific and sufficient conditions for project WBS and resource optimization. In addition the division optimization theorem is combined with the flow process network in order to provide a scientific optimization method for the construction section in the flow process.