摘要
通过仿真机进行凝汽器故障诊断,选择合适的故障判断基准,判断凝汽器系统真空是否发生故障。首先在确定真空应达值时建立了Pc-M模型,在其后的真空故障判断中使用该Pc-M模型表示机组的真空应达值,当阈值取0.5%时可以诊断出真空故障,最后分别采用了线性相关度法和BP神经网络法对单故障工况和多故障并发工况进行诊断。结果表明BP神经网络法相比线性相关度法无论是速度还是精度都有明显的优势,而且在多故障并发的情况下同样结果精确。
According to the simulation machine to do condenser fault diagnosis. Choosing the appropriate fault judg- ment benchmarks to judge whether thevacuum of condenser failed. Firstly we established the Pc-M model to judge the vacuum value, and secondly the Pc-M model was used to judge the vacuum. When the threshold was 0. 5% , the vacu- um failure can be diagnosed. Finally the linear correlation method and the BP neural network method were used to diag- nose the single fault condition and the muhi-fauh concurrency conditions. The final results showed that the BP neural network method was better than the linear correlation method and has obvious advantages both the speed and accuracy, and the same exact is accurate in the case of multiple failures.
出处
《电力科学与工程》
2017年第5期70-78,共9页
Electric Power Science and Engineering
基金
国家自然科学基金资助项目(51276059)
关键词
凝汽器
故障诊断
真空应达值
线性相关度
BP神经网络
condenser
fault diagnosis
vacuum target value
linear correlation
BP neural network （BPNN）
