The aim of this study is to develop a signal decomposition method that not only overcomes the shortcomings of existing decomposition methods, but also can be further applied to complex underwater acoustic signal processing. Therefore, a novel Adaptive Complementary Ensemble Local Mean Decomposition (ACELMD) method was proposed. A simulated signal and real-world underwater acoustic signals from three marine mammals (white-sided dolphin, long-finned pilot whale, and harp seal) were employed to evaluate the decomposition performance, reliability, and practicality of the proposed ACELMD. Also, the decomposition results of the same simulated signal using ACELMD, Local Mean Decomposition (LMD), Ensemble Local Mean Decomposition (ELMD), and Complementary Ensemble Local Mean Decomposition (CELMD) methods were compared. All the results demonstrate the excellent decomposition performance of the proposed method. Moreover, compared with the other three methods, ACELMD effectively reduces the modal aliasing in the decomposition results, reduces the number of LMD executions, and is more inclusive of the white noise amplitude, all of which indicate its great potential in practical applications.

这项研究的目的是开发一种信号分解方法，该方法不仅克服了现有分解方法的缺点，而且可以进一步应用于复杂的水下声信号处理。因此，提出了一种新的自适应互补集合局部均值分解（ACELMD）方法。来自三个海洋哺乳动物（白面海豚，长鳍领航鲸和竖琴海豹）的模拟信号和真实世界的水下声信号被用来评估拟议ACELMD的分解性能，可靠性和实用性。此外，比较了使用ACELMD，局部均值分解（LMD），集合局部均值分解（ELMD）和互补集合局部均值分解（CELMD）方法对相同模拟信号的分解结果。所有结果证明了该方法的优异分解性能。而且，与其他三种方法相比，ACELMD有效地减少了分解结果中的模态混叠，减少了LMD执行的次数，并且更包含白噪声幅度，所有这些都表明了其在实际应用中的巨大潜力。