资源与工期双重随机下的模具项目群监控

发布时间：2019-05-09 12:18

【摘要】：针对模具企业因随机事件众多且在制项目群规模庞大而导致项目生产难以有效监控的问题,考虑可用资源及任务工期的随机性,提出一种基于资源解耦的模具项目群监控方案。利用经典优先规则得到基于任务工期均值且嵌入资源缓存的确定性项目计划,据此构建非监控项目的近似负荷模型以解除项目间的约束关系。结合马尔可夫决策过程理论建立监控项目的决策模型,通过外协核心任务使项目群的总成本期望最小。为应对维数灾问题,在资源解耦极大降低问题规模的基础上,进一步提出基于工期和资源中断概率阈值的近似求解方法。通过计算实例验证了所构建的近似负荷模型、决策模型及近似方法的有效性。
[Abstract]:Considering the randomness of available resources and task duration, a project group monitoring scheme based on resource decoupling is proposed to solve the problem that it is difficult for mold enterprises to monitor project production effectively due to the large number of random events and the large scale of project clusters in manufacturing projects. Based on the classical priority rule, a deterministic project plan based on task duration average and embedded resource cache is obtained. Based on this, an approximate load model of non-monitored projects is constructed to remove the constraints between projects. Based on Markov decision-making process theory, the decision-making model of monitoring project is established, and the total cost expectation of the project group is minimized by the core task of external cooperation. In order to deal with the problem of dimension disaster, on the basis of reducing the scale of the problem greatly by resource decoupling, an approximate solution method based on the time limit and the probability threshold of resource interruption is proposed. The validity of the approximate load model, decision-making model and approximate method is verified by an example.
【作者单位】：广东工业大学广东省计算机集成制造系统重点实验室;
【基金】：国家自然科学基金资助项目(51175094,51375098) 国家科技支撑计划资助项目(2012BAF12B10)~~
【分类号】：TG76;TP277

【参考文献】