一种利用神经网络对真空断路器故障诊断的方法

A New Method Based on Neural Network for Vacuum Circuit Breaker Fault Diagnosis

文章首先对22CBDL型真空断路器的工作原理和故障机理进行了研究,选取断路器中四组比较容易测试的测试量,分析了这四组测试量与真空断路器几种常见故障之间的对应关系。然后从大量实验数据中提取具有代表特征的数据作为学习样本,用BP神经网络构建了一个22CBDL型真空断路器故障诊断模型,实现了通过简单易测试信号对复杂不易测信号的故障诊断。最后,MATLAB仿真结果验证了其正确性。

The relation between the fault of the 22CBDL Vacuum Circuit Breaker and four groups of easily getting measurements are found based on the study and analysis of the Vacuum Circuit Breaker operating principle and failure mechanism when there is a fault happened. Complex fault diagnosis is carried through the 22CBDL Vacuum Circuit Breaker fault diagnosis model which is built by BP neural network and 21 groups of easy getting representative data of the fault diagnosis experiment selected from the crowd as samples. Finally the simulation result of MATLAB shows its validity.