【摘要】 利用离散小波变换将时变结构的时变参数在多尺度上展开为概貌信号和细节信号,忽略高频细节信号,仅由低频概貌信号估计时变参数,将时变结构识别问题转化为时不变结构识别问题。基于时变结构的输入和输出信号用最小二乘法识别低频尺度展开系数,从而重构得到原始时变参数。对于噪音引起的方程不适定性问题,采用Tikhonov正则化方法减小其影响。通过1个具有时变刚度和阻尼的2层框架结构的数值模拟验证该方法的有效性。识别结果表明:该方法可以有效识别结构时变刚度,时变阻尼识别结果对噪音较敏感。更多还原

【Abstract】 A method of identifying physical parameter of time-varying structure based on discrete wavelet transform was presented.A time-varying parameter was expended on multi-scale as approximate and detail signal by discrete wavelet basis.By ignoring detail signal,the time-varying parameter was evaluated only using approximate signal,so the identification problem of time-varying was transformed to time-invariant problem.Based on the input and output signal of structure,the scale expand coefficients of low frequency can be identified using least square method,and then the original time-varying parameter can be reconstructed.For the ill-posed problem of identification equation caused by noise,Tikhonov regularization method was used to decrease its influence.The effectiveness and accuracy of the proposed method were validated via a numerical simulation of a two story frame structure with time-varying stiffness and damping.The results show that the presented algorithm can be used to identify the time-varying stiffness,the identified result of time-varying damping is very sensitive to noise.更多还原