Abstract: For the problem that the optimization of multi-project scheduling under unknown project priority was hard to solve by traditional centralized project method,based on solving idea of combining decentralized local decision making with global coordination decision making,a decentralized project scheduling method by integrating local optimization with shared resources bidding,resources combinatorial auction and price iterative adjustment was proposed in market mechanism,which could reduce average distance between multi-project planned duration and optimal duration.Simulation results showed the advantage and usability of decentralized solution compared to centralized solution.

摘要： 鉴于传统的基于集中优化方法难以很好解决优先级未知环境下的多项目工期优化问题,基于分散式局部决策与协调决策相结合的求解思想,提出融合局部优化与共享资源竞标出价、资源组合拍卖与价格迭代调整的基于市场机制的分散式项目调度方法,以减小多项目计划工期与最佳工期的平均差距。通过仿真实验,验证了分散式解法相比于集中式解法的优势和适用性,能够解决集中方法难以求解的项目优先级未知或不确定的多项目调度问题。

关键词: 项目调度, 分散式决策, 市场机制, 竞争判优