基于 SA-GA 模糊熵的 VMD 算法在爆破振动信号分解中的应用

针对变分模态分解(VMD)算法预设参数选择的问题,提出了一种基于 SA-GA 模糊熵的 VMD 参数优化 算法,该算法结合模拟退火算法(SA)和遗传算法(GA)的优点,选取模糊熵( FE)为适应度函数,求解最优分解参数。 经过仿真信号分析,相比 EMD 算法,SA-GA 模糊熵的 VMD 参数优化算法有效地抑制了模态混叠和虚假分量的现象, 具有较高的分解精度。 最后利用 SA-GA 模糊熵的 VMD 参数优化算法进行爆破振动信号实测分析,结果表明:SA-GA 模糊熵的 VMD 参数优化算法可以根据不同的爆破振动自适应地选取最优解,解出来的 IMF 分量具有明确的物理意 义,频谱图能清晰地看出信号内所包含的频率成分,具有良好的适用性。

Application of VMD Parameter Optimization Based on SA-GA Fuzzy Entropy in Blasting Vibration Signal Decomposition

For variational mode decomposition (VMD) preset parameter selection problem,a VMD parameter optimization algorithm based on SA-GA fuzzy entropy is proposed,Junction and the advantages of simulated annealing algorithm (SA) and genetic algorithm (GA),the Fuzzy Entropy (FE) was selected as the fitness function,Solving the optimal decomposition parameters;After simulation signal analysis,compared with EMD algorithm,VMD parameter optimization algorithm based on SA-GA fuzzy entropy can effectively suppress the phenomenon of mode mixing and false components,and has higher decomposition accuracy. Finally,the VMD parameter optimization algorithm of SA-GA fuzzy entropy is used to conduct the actual analysis of blasting vibration signal. The results show that:The VMD parameter optimization algorithm of SA-GA fuzzy entropy can select the optimal solution according to different blasting vibration adaptive,and the IMF component solved has clear physical significance,and the spectrum diagram can clearly see the frequency component contained in the signal,which has good applicability.