项目名称: 基于分布式传感器网络因果关系分析的复杂制造系统过程监控与诊断
项目编号: No.71471005
项目类型: 面上项目
立项/批准年度: 2015
项目学科: 管理科学
项目作者: 史建军
作者单位: 北京大学
项目金额: 62万元
中文摘要: 生产过程的监测与诊断在生产稳定运行与产品质量保障方面发挥着极其重要的作用。通过分布式传感器网络技术,复杂制造系统能够采集制造过程中各个工序的过程数据,从而形成时间和空间维度上的信息富集环境,这为复杂制造系统的生产过程监控与诊断水平的提升提供了前所未有的机遇。然而,由于传感器网络结构的复杂性和各种不确定性因素的影响,现有的监测与诊断系统普遍存在无法有效溯源和诊断的问题。这不仅严重影响了产品的质量和生产效益,甚至会造成生产设备的损害。本课题以复杂制造系统的生产过程为研究对象,结合分布式传感器网络的特点和工程背景知识,研究基于因果关系分析的过程监控与诊断方法,并开展复杂制造系统过程监控与诊断的验证研究,解决不确定性条件下的因果关系数学建模、过程监控与诊断性能指标评估、实时过程监控与诊断等关键科学问题,为解决目前复杂制造系统监测与诊断的突出问题提供理论依据与关键技术。
中文关键词: 统计过程控制;过程监控;质量工程
英文摘要: Process monitoring, diagnosis and control play an important role for product quality and process safety in manufacturing systems. Distributed sensing network (DSN), a system-wide deployment of different sensing devices embedded in manufacturing systems, has resulted in both temporally and spatially dense data-rich environment, which provide unprecedented opportunities for process and quality improvement. However, because of data errors and uncertainties in the existed in DSN, it is difficult to efficiently monitor the process and find the root causes, which greatly deteriorates the quality and productivity, and even causes equipment damages. Aiming at the complex manufacturing processes, we will systematically study the theory of causation-based process monitoring, diagnosis and control, by incorporating the engineering domain knowledge and the natures of DSNs. Our research targets are to establish the mathematical models, design the monitoring and control strategies under uncertainties, and evaluate the performance of our proposed methods by a real-world manufacturing example, and finally establish theoretical basis and provide key technologies.
英文关键词: SPC;Process control;Quality Engineering