电站故障诊断系统中信号缺失处理的粗糙集方法

Rough set method for invalid and absent signal treatment in power station fault diagnosis system

采用粗糙集理论分析电站故障诊断决策表属性的重要性 ,在设计阶段用于指导诊断系统对相关输入信号的取舍 ,是分析解决诊断系统输入信号缺失问题的有效方法。根据粗糙集对条件属性的分析 ,在故障诊断专家系统中选择冗余规则 ,使诊断系统在运行中具有对单一非核属性信号失效的容错能力。以 30 0MW机组凝汽器故障诊断为例 。

The rough set theory is an effective method to analyze and solve invalid and absent signals of diagnosis system since it can be applied to analyze the significance of the property of power plant fault diagnosis decision table and can be used to guide the diagnosis system to select the relevant input signals in the designing stage. According to the analysis of the conditional property conducted through rough set theory, the redundant rules were selected in the fault diagnosis expert system to achieve the fault-tolerant ability with single signal with non-core properties during operation. The fault diagnosis for the condenser of 300 MW unit was set for an example to expatiate the practical application of the rough set theory.