论文部分内容阅读
基于模型的SDG(Signed DiGraph,符号有向图)故障诊断方法因其具有完备性好、推理深度高等优点在过程工业安全工程中具有十分重要的意义,已成为安全工程中的1种关键技术。本文在以前研究的基础上,提出了1个综合性SDG故障诊断架构,以期能够实现在生产过程中及时发现故障并判明故障源。该综合性故障诊断架构按模型、推理和应用3个层次搭建,以传统定性SDG及概率SDG理论为基础,包含了从模型建立到故障诊断推理,从定性SDG方法到结合统计监控的SDG方法再到概率SDG方法等一系列实施方案。该综合性SDG故障诊断架构由于引入了多元统计监控模型,使得在系统没有表现出明显的故障征兆时就能够及时敏感地检测到异常变化,进而触发SDG及PSDG推理来实现对故障源的查找,并给出各故障源发生故障的概率值,以指导使用者按照概率值的大小顺序采取处理措施。以某石化公司的气体分馏装置为实际背景,利用该装置实时数据库中的实际生产工艺数据对该综合性诊断架构进行了实例验证,其故障诊断结果与实际发生的故障相吻合,证明了该综合性故障诊断架构的有效性。
Because of its completeness and high reasoning depth, the fault diagnosis method of model-based SDG (Signature Directed Graph) is of great significance in process industrial safety engineering and has become a key technology in safety engineering. Based on the previous research, this paper presents a comprehensive SDG fault diagnosis framework, in order to achieve the timely detection of faults in the production process and identify the source of the fault. Based on the traditional qualitative SDG and probabilistic SDG theory, the integrated fault diagnosis architecture is built on the three levels of model, reasoning and application. It includes from model building to fault diagnosis reasoning, from qualitative SDG method to SDG method combined with statistical monitoring To probabilistic SDG method and a series of implementation plans. Due to the introduction of multivariate statistical monitoring model, the comprehensive SDG fault diagnosis framework can detect abnormal changes promptly and sensitively when the system shows no obvious fault symptoms, and then triggers SDG and PSDG reasoning to find the fault source. And gives the probability value of each fault source failure to guide the user to take measures in accordance with the size of the probability value. Taking a petrochemical company’s gas fractionation unit as the actual background, the comprehensive diagnostic framework is verified by the actual production process data in the real-time database of the petrochemical company. The fault diagnosis result is consistent with the actual fault, which proves that the synthesis The effectiveness of the sexual diagnostic framework.