摘要
本文介绍了把喷油器顶部的振动信号和喷油管泵端的固体声信号经A/D 转换,预处理和选通处理得到的数字信号作为时间序列,采用自回归(AR(p))模型和混合自回归(AR(p,q))模型抽取故障特征,再经特征比较,状态识别和参数修正等步骤,在IBM-PC/XT 微机上成功地实现了对柴油机喷油系故障的快速、自动诊断。并对两种不同时序模型的诊断结果作了比较和分析。
The vibration signal at the top of an injector and the sound signal ofthe injection pipe near the pump are measured and digitized.After being tr-eated by preprocessing and gating technique,they are taken as time-series.Signatures of engine faults are obtained by applying the autoregressivemodel AR(p)and mixed autoregressive model AR(p,q).They are compared,discriminated and amended if necessary.When fed into an IBM-PC/XTmicrocomputer,quick and effective autodiagnosis is accomplished to detectthe fault of the injection system of a diesel engine.Similarly,the twotime series models are also being compared and analysed in regard to theirdiagnostic results.
出处
《内燃机学报》
EI
CAS
CSCD
北大核心
1989年第4期363-368,共6页
Transactions of Csice
