摘要
提出了一种基于模糊模式识别的内燃机失火故障综合诊断方法。通过建立简明实用的瞬时转速诊断模型并深入分析瞬时转速信号的波形特点 ,定义并提取了 10种无量纲的诊断特征参数。在此基础上 ,利用模糊模式识别技术对这些参数进行信息融合 ,首先快速识别整机有无失火故障 ,然后在有必要时再进行故障气缸定位。实验结果表明 ,该方法可以充分利用多种特征参数的信息互补性 ,提高了诊断效率和识别正确率 。
A synthetic technique for misfire detection of internal combustion engines based on fuzy pattern recognition is proposed, in which the idea of information fusion is introduced. Under the assumption that the engine can be modeled as a system of a single degree of freedom, a simplified diagnostic model relating the gas pressure torque to the crankshaft speed fluctuations is derived. 10 dimensionless parameters are extracted by use of the model and by direct analysis of the speed waveform. By means of fuzzy pattern recognition, the parameter are integrated together as a fuzzy vector. The vector will identifies uniformly whether the engine is healthy or not first, then locate the position of a misfiring cylinder or cylinders if necessary. The experimental results showed that the strategy was able to use reasonably the complementary information of all the parameters and led to better diagnostic reliability and efficiency. The technique is simple in implementation and may provide practical utility for on line monitoring and real time diagnosis, including misfire detection on individual cylinders.
出处
《振动工程学报》
EI
CSCD
2000年第1期37-45,共9页
Journal of Vibration Engineering
基金
国家"九五"攀登计划资助项目! (PD95 2 190 8)
中国船舶工业总公司重点预研资助项目 !(2 3.8.1.3)
关键词
内燃机
故障诊断
模糊模式识别
失火
转速波动
internal combustion engine
fault diagnosis
fuzzy pattern recognition
misfire
speed fluctua-tions
