融合技术在柴油机故障诊断中的应用

导致柴油机同一故障的各相关检测信号使其融合，应用在神经网络故障识别系统中，可以快速，准确地诊断柴油机.主主要融合气缸振动信号，喷油信号及水温信号，设计了通过撮以信号持征值，应用神经网络识别系统诊断柴油机功率下降故障的方法。信号特征值分别按各自信号的特点提取，其中气缸振动信号以振动信号的各子带功率谱为特征，即在频域中选取特征；喷油信号以信号波形形状识别为特征，即在时域中选取特下；水温信号因其为慢速变化信号，取不同阶段的水温信号为特征。

Application of Fusion Technology to Diesel Engine Fault Diagnosis

In fault diagnosis system of neural net, applying fusion technology to analyzing synthetically the different kinds of related signals of diesel engines, can rapidly and accurately diagnose some faults. The paper illustrates the diagnosis way by fusing vibration signal of cylinder, fuel injection signal and cool water tempature signal to diagnose power decline fault of diesel engines. The diagnosis is finished by choosing characters of the three kinds of signals in the diagnosis system of neural net. On the characters, vibration signal of cylinder chooses power spectrum of every belt, those are chosen in the frequency field; fuel injection signal chooses the characters of shape recognition, those are chosen in the time field time; cool water tempter signal chooses the volt different phase because of its slow variation.