An algorithm of eliminating spike of the measured excavator load spectrum signal based on peak-valley extraction

The spike of the original measured load spectrum signal has great influence on the fatigue life prediction, so it must be identified and eliminated accurately. According to the variation law of the strain and force signals collected by the excavator load spectrum test, and the characteristics of the spikes contained therein, a spike elimination algorithm based on peak-valley extraction was proposed. In this algorithm, the spikes of digital signal is discriminated and corrected according to the changing trend of signal. The results of both the simulation test and the industrial example show that, the algorithm based on peak-valley extraction can accurately distinguish the spike from the normal signal. The principle of the algorithm is clear, the computational complexity is low, and the result is true and reliable, which lays a data foundation for the subsequent fatigue life prediction and fatigue reliability evaluation.