中文摘要：

提出了一种基于数字化的生产模型，使用控制图、故障树分析和专家知识，能够进行制造过程实时监控的诊断，该模型提高了故障诊断系统的可靠性，并提供了可实际操作的可视化建模工具。所开发的在线统计过程控制系统能够根据生产事件的监测，动态响应制造过程变化。该系统运用可视化建模工具，根据专家经验进行故障树建模，通过故障树自动生成专家系统诊断规则库，实现诊断知识的自动获取。将该系统应用于汽车变速箱装配过程的检测与故障诊断，验证了方法的有效性。

英文摘要：

A fault diagnosis method based on digitized production model with control charts, fault analysis tree （FTA） and expert knowledge was introduced, which could conduct real-time monitoring and control over manufacturing process. This event-based digitized production model enhanced the reliability of fault diagnosis system with visual modeling software, which provided feasible tools for system modeling. An on-line Statistic Process Control （SPC） system was developed, which could be used to detect production event and dynamically respond to the changes of manufacturing process. By using visual modeling tools and experts＇ experience to construct fault trees, rule base of expert diagnosis system could be obtained automatically, and the diagnosis knowledge could be acquired automatically. The system was applied in detection and fault diagnosis of automobile gear-box assembly process, and the effectiveness of the proposed method was verified.