摘要
分析瞬时转速信号在发动机发生失火故障时的波动特性,提出利用发动机瞬时转速诊断电站系统中原动机的失火故障。在提取瞬时转速波形特征参数的基础上,提出基于BP(back propagation)神经网络的故障诊断方法,并对该方法进行实验验证,成功诊断出某型发动机单缸及双缸失火故障,效果明显。
The fluctuation character of the transient rotational speed is analyzed when misfire occurred in motor. The diagnosing method, via transient rotational speed, for the misfire fault of the prime motor in the power station is developed. On the basis of extracting transient rotational speed waveform characteristic parameter, the fault diagnosis method based on neural network is put forward. The experiment shows the method diagnoses the single-cylinder and double-cylinder engine misfire fault of certain motor successfully.
出处
《机械强度》
CAS
CSCD
北大核心
2006年第z1期56-59,共4页
Journal of Mechanical Strength
基金
军械工程学院创新基金项目(CXJJ0504)~~
关键词
电站
发动机
瞬时转速
失火故障
神经网络
Power station
Engine
Transient rotational speed
Misfire fault
Neural network
