The two-field vibroacoustic finite-element (FE) model requires a relatively large number of degrees of freedom compared to the monophysics model, and the conventional force identification method for structural vibration can be adjusted for multiphysics problems. In this study, an effective inverse force identification method for an FE vibroacoustic interaction model of an interior fluid-structure system was proposed. The method consists of: (1) implicit inverse force identification based on the Newmark-$\beta$ time integration algorithm for stability and efficiency, (2) second-order ordinary differential formulation by avoiding the state-space form causing large degrees of freedom, (3) projection-based multiphysics reduced-order modeling for further reduction of degrees of freedom, and (4) Tikhonov regularization to alleviate the measurement noise. The proposed method can accurately identify the unmeasured applied forces on the in situ application and concurrently reconstruct the response fields. The accuracy, stability, and computational efficiency of the proposed method were evaluated using numerical models and an experimental testbed. A comparative study with the augmented Kalman filter method was performed to evaluate its relative performance.
翻译:与单物理模型相比,两侧振动定分模型需要相对数量较大的自由度,而且可针对多物理问题调整结构振动的常规部队识别方法。在本研究中,提出了内部流体结构系统FE振动声学互动模型的有效反向识别方法。该方法包括:(1) 根据新马克-$\beeta$的时间整合算法,为稳定性和效率进行隐性反向识别;(2) 通过避免产生较大自由度的状态空间形式进行二级普通差异配置;(3) 为进一步降低自由度而采用投影多物理减序模型;(4) 使Tikhonov正规化以缓解测量噪音;拟议方法可以准确确定现场应用的不计量力量,并同时重建应对场;使用数字模型和实验试验台对拟议方法的准确性、稳定性和计算效率进行了评估;与扩大的Kalman过滤法进行了比较研究,以评价其相对性能。