Research on data-driven plunger pump shock signal separation model
For the problem that it is difficult to separate the shock signals of multi-cylinder plunger pumps and it is difficult to judge mechanical faults,a data-driven shock signal separation model for plunger pumps was proposed.The model first performed adaptive band-pass filtering on the original signal according to the principle of maximum spectral cliffs;then automatically searched the plunger pump speed data according to the peak of the secondary envelope spectrum combined with the optimal harmonic energy and algorithm;secondly,the single cycle major cycle of the plunger pump was segmented according to the speed to extract the single shock signal of each seal valve in each major cycle;the starting moment of every single shock was judged,and the cycle time variation was designed,the filter set was used to filter the original signal to obtain the shock signal of a single cylinder,and the corresponding characteristic index was calculated;finally,the accuracy of the above algorithm model was verified by using the actual plunger pump vibration data obtained from experiments.The results show that for the three-cylinder pump vibration signal with evident shock,its rotational frequency is 3.756 6 Hz and the signal period is 5.12 s through analysis by the separation model,and 6-period time-varying filter groups were designed to process the original signal,which achieved effective separation of shock signals of each sealing valve,and facilitated the subsequent independent monitoring and diagnostic analysis of each valve body.The research results can provide technical support and reference for the separation and extraction of multi-source shock signals.
国家科技期刊平台
NSTL
万方数据
