一种有效诊断电站热力系统故障源位置的算法  

An Efficient Algorithm to Fault Location in Power Plant Thermal System

作　　者：佟鹏[1] 曹文亮[2] 张冀[2] 王玉清[1] 赵志宏[1] 许丽卿[1] 辛宇峰 

机构地区：[1]保定电力职业技术学院,河北保定071051 [2]华北电力大学,河北保定071003 [3]阳城国际发电有限责任公司,山西阳城048102

出　　处：《系统仿真学报》2006年第7期1991-1994,共4页Journal of System Simulation

摘　　要：基于符号有向图(SignedDirectedGraph,SDG)的定性推理方法用有向图将故障传播模型化,并用节点间的符号逻辑关系定位系统的故障源,是一种完备的揭示系统故障的有效方法。为了克服其固有缺陷和提高诊断精度,通过采用扩展的有效节点和一致边对被控节点进行描述,用穷举故障模式和搜索有向图的最大强连通支的方法去处理不可测量节点,可对系统故障进行有效的诊断;然后对某些定性特征一样而定量值不同的故障模式,为了能有效区分,通过采取加入适当定量信息去构造定性定量模型的办法来进行进一步诊断。案例试验表明所给出的方法是可行的。Signed Directed Graph （SDG） is used for a mathematical model representing the influences among elements of the system, which considers the characteristic of system＇s structure and can deal with the complex process of system in a large-scale, and it is a self-contained method to diagnosis system failures. In order to eliminate carrying out the complicated and inefficient quantitative simulation, Firstly, the controlled nodes can be described by adopting the expanded effective nodes and consistent branches, and the shortcoming of the unmeasurable nodes can be overcome by enumerating all fault patterns and searching the maximal strongly-connected component in the cause-effect graph reflecting the pattern of abnormality, and then the qualitative diagnosis results of fault modes can be achieved. Secondly, the patterns with the same qualitative characteristics can be distinguished by using quantitative knowledge to form a qualitative and quantitative model The case studies show the improved SDG is valid.

关 键 词：符号有向图 故障定位 定性定量模型 算法研究 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]