Huang Rui, Seiki Takafumi, Dong Qinxi, Yamaoka Satoshi, Aydan Ömer
School of Architecture and Civil Engineering, Xihua University, Chengdu, 610039, China.
School of Regional Development and Creativity, Utsunomiya University, Tochigi, 321-8585, Japan.
Sci Rep. 2024 Oct 16;14(1):24214. doi: 10.1038/s41598-024-74850-2.
A time domain rock mass parameter identification method considering the Rayleigh damping is proposed in this paper. Starting with a set of initial parameter values, the algorithm solves the finite-element (FE) equations updating the parameters in each iterative step until the result converges to a local optimum. Numerical FE analysis, displacement sensitivity analysis, and the truncated singular value decomposition method are integrated into the solution algorithm for parameter identification. The technique introduced is verified in numerical studies with in-situ experiments of rock mass structures. Our results show that the approach, compared with conventional methods, can reduce ambiguities caused by inadequate data and provide more accurate insights into the subsurface for rock mass quality evaluation.
本文提出了一种考虑瑞利阻尼的时域岩体参数识别方法。该算法从一组初始参数值开始，求解有限元（FE）方程，在每个迭代步骤中更新参数，直到结果收敛到局部最优值。数值有限元分析、位移灵敏度分析和截断奇异值分解方法被集成到参数识别的求解算法中。所引入的技术在岩体结构现场试验的数值研究中得到了验证。我们的结果表明，与传统方法相比，该方法可以减少数据不足引起的模糊性，并为岩体质量评价提供更准确的地下洞察。
