基于模式识别的柴油机故障诊断

Fault Diagnosis of Diesel Engine Pattern Recognition-based

提出一种基于模式识别的故障诊断方法 ,利用传感器检测柴油机燃油压力波形 ,并对波形进行数学建模和特征提取 ;讨论分析了最小距离判别法 ,选用Euclidean距离 ,将其应用到柴油机燃油压力波形的故障诊断中 ,得到相当理想的诊断结果 ,最后给出两个实验例子。此方法具有运算简单、运算速度快、判别精度高等特点 ,具有广阔的发展空间。

Diesel engine's faults centralize in fuel system. At present, desk experiments are usually used in diesel gengine's fault diagnosis, which consumes long time, have large machinery and cost high. A new fault diagnosis method pattern recognition-based is brought forward,where diesel engine's fuel pressure wave shape is measured using sensor. Furthermore, we found its mathematics model and distilled its characters. Discriminance of least distance is discussed and discriminance of Euclidean distance is chosen to be used in fault diagnosis of diesel engine's fuel pressure wave shape,which acquired realistic diagnosis results in two real examples. This method has the performances of simple operation, fast speed and high precision, and will be used widely.