多参量结构响应重构中信息融合与参数优化研究

针对现有信息融合响应重构方法中过程参数需要经验取值的问题，开展了基于信息融合与过程参数优化的多参量结构响应重构方法研究。通过卡尔曼滤波算法实现应变、位移和加速度信号的多源信息融合，并引入粒子群算法对过程噪声方差矩阵进行参数优化。利用梁结构和半潜式平台立柱结构数值模型分别进行验证，研究结果表明：所提方法能够准确重构这两类结构的全场加速度、位移和应变参量；相比经验取值方法，所提方法对三种参量的幅值重构误差整体降低了 11.1%~35.9%。该成果为结构全局力学指标监测提供了技术参考。

英文摘要:

To address the issue of empirical parameter selection in existing information fusion-based response reconstruction methods, this study developed a multi-parameter structural response reconstruction method integrating information fusion with process parameter optimization. The Kalman filter algorithm is employed to achieve multi-source information fusion of strain, displacement, and acceleration signals, while the Particle Swarm Optimization (PSO) algorithm is introduced to optimize parameters in the process noise covariance matrix. Numerical validations using beam structure and semi-submersible platform column models demonstrate that the proposed method can accurately reconstruct full-field acceleration, displacement, and strain parameters for both structural types. Compared with conventional empirical parameter selection approaches, the proposed method achieves significant error reduction in amplitude reconstruction: overall decreases of 11.1%~35.9% are observed across the three parameters. These findings provide valuable references for global mechanical indicator monitoring in structural health assessment.

查看全文查看/发表评论下载PDF阅读器