基于奇异值分解降噪的移动载荷识别

模型梁振动响应测试信号噪声成分较强，使得根据响应信号识别梁上移动载荷无法取得理想效果，利用测试响应的时间序列重构吸引子的轨迹矩阵，对该矩阵进行奇异值分解以剔除噪声信号，根据残余信号进行载荷识别，结果表明本方法优于以前的识别方法。

Moving-Force Identification Based On Noise Control By Singular Value Decomposition

Because of the strong noise signal in measured response of the simple supported model beam caused by a moving model bar, it is difficult to identify the forces moving on the beam based on the measured responses. The track matrixes of an attractor are constructed by the time series of the measured responses, and singular value decomposition about these matrixes is performed to eliminate the noise signal and extract the remnant signal, and the moving forces are identified by the remnant signal. The analysis shows that the force identification result is better than that of the previous research.