基于抽区间采样的内燃机状态监测方法研究

Internal Combustion Engine Condition Monitoring Research Based on Extract Time Interval Sampling

内燃机由于内部激励源众多,外部测得的振动信号十分复杂,不能准确分析和监测机器的磨损状态。利用各激励力存在一定的相位差进行抽区间采样,可以缩短数据长度,突出局部时间段信号特征,削弱其它干扰成分的影响。提出基于抽区间采样的内燃机状态监测方法:抽取多个工作循环特定时间段的数据,合成一个新的数据信号;对合成信号进行经验模态分解,通过分解后的本征模态函数的幅值谱比较,选取对故障敏感的本征模态函数进行信号的重构;然后对重构的信号进行时频分析,达到对内燃机状态监测的目的。

The vibration signal is very complex getting from the ICE＇s body because of many driving sources inside the machine. It is not good to estimate machine＇s condition by the direct analysis signal. Extract time interval sampling method where the different driving sources have phase differences can shorten data lengths, give prominence to signal feature of the local time segment, and weaken others disturb components affection. This paper gave that ICE condition monitoring method based on the extract time interval sampling. Time segments were extracted from several work cycles according to phase, and then, a new data signal was compoesd using emepirical mode decomposition （EMD） method. Synthesis signal was decomposed using EMD method. Several intrinsic mode functions being sensitive to the fault reconstruct signal were chosen accoding to amplitude spectrum comparison to obtain a further characteristic extraction of the signal. ICE＇s condition monitoring was attained by the time-frequency analysis to the reconstruction signal.