基于EFSM切片的化工过程故障诊断方法  

Chemical process fault diagnosis based on EFSM slice

作　　者：朱群雄[1] 隋英丽[1] 史晟辉[1] 徐圆[1] 

机构地区：[1]北京化工大学信息科学与技术学院,北京100029

出　　处：《计算机与应用化学》2013年第7期725-729,共5页Computers and Applied Chemistry

基　　金：国家自然科学基金资助项目(61104131)

摘　　要：针对复杂化工过程故障分析和理解困难的问题,应用计算机领域中的EFSM切片技术和知识库方法,提出一种新的应用于化工过程故障诊断的解决方案。本文基于前人建立的扩展有限状态机(Extended Finite State Machine,EFSM)模型,选取系统操作流程中的异常对象作为切片准则。然后,利用邻接依赖图EFSM切片算法,给出求解故障对象相关变量数据依赖图的过程,通过宽度优先搜索变量数据依赖图考查各节点是否为故障源,建立故障诊断所需知识库。最后,列出基于EFSM切片的故障诊断推理的具体过程,以及根据构建的知识库进行的诊断工作。以双容水槽液位控制系统为例,计算EFSM切片,其规模有效约减为原模型的70%左右;进而得到相关变量数据依赖图,分析异常状态和故障源,使得基于EFSM切片的化工过程故障诊断方法的可行性得以验证,为化工过程故障诊断提供新的思路。For the fault diagnosis problems of complex chemical process, a new solution used EFSM slice and expert system in the computer technology field is put forward. In this paper, the exception object in the operation process is firstly selected as the slicing criterion. Secondly, based on the EFSM model introduced by the predecessor, the adjacent data dependence graph of fault object is conducted by EFSM slicing algorithm. Furthermore, in order to find the source node of fault, the data dependence graph is breadth-first searched, and then the knowledge base of fault is established. Finally, the diagnosis process is described. As an example, the EFSM slice for data dependence graph of the double water tanks level control system is calculated. And the size of the EFSM slice is reduced to about 70 % of the original model. Through analyzing data dependence graph of relevant variables, the source of fault is found. The feasibility of the fault diagnosis method based on EFSM slice in the chemical process is verified in the experiment. Moreover new ideas for fault diagnosis of chemical process are provided.

关 键 词：EFSM切片 邻接依赖图 故障诊断 

分 类 号：TQ015.9[化学工程] TP391.9[自动化与计算机技术—计算机应用技术]