Uncertainty characterization and propagation analysis for pneumatic soft acoustic metamaterial system

Pneumatic soft acoustic metamaterials have gradually attracted attention inspired by pneumatic soft robots. However, current researches ignore the ubiquitous uncertainty factor, which may cause the designed pneumatic soft acoustic metamaterials to fail to achieve the expected performance. In this paper, the influence of uncertainty on pneumatic soft acoustic metamaterial system is investigated. To quantify uncertainties for the system input based on available data, two different uncertainty characterization methods are utilized. By integrating the bootstrap method with kernel density estimation, the input distribution of bounded random model can be determined based on the limited experiment data. For cases with even less experiment data, an unbiased estimation method is introduced to construct interval model. Then, an uncertainty propagation method based on Kriging model and an improved active learning strategy is developed for the pneumatic soft acoustic metamaterial system with bounded hybrid uncertain parameters. Finally, we experimental demonstrated the effectiveness of the uncertainty analysis on the deformation and acoustic property of the pneumatic soft acoustic metamaterial system. The results show the necessity of regarding uncertainties in pneumatic soft acoustic metamaterial system. The study provides a feasible and practical method to model and propagate uncertainty for pneumatic soft acoustic metamaterials systems, which can promote their application in industrial sectors.