摘要
从快速傅里叶变换 (FFT)离散谱线中难以恰好找到半功率点 ,通常是通过线性插值法近似确定半功率点来估计阻尼。文中采用理论分析与数值仿真相结合的方法研究基于线性插值的半功率点法的主要误差因素和各自特性 ,特别是其中窗阻尼变化规律。研究结果表明 ,①应尽量接近整周期采样。②无阻尼系统因进行FFT截断也会出现虚假的窗阻尼 ,且真实阻尼不等于表观阻尼减去窗阻尼。③阻尼的相对误差主要取决于采样周期数与阻尼比 ,其物理意义是截断窗口内的信号衰减程度。④要保证阻尼 10 %的精度 ,窗阻尼要小于表观阻尼的 2 3。
To guarantee the precision of damping estimated by the method of half power points, the main error factors involved in the estimation procedure was investigated by theoretical analysis and simulation, the characteristics of pseudo damping of window (PDW) truncation was studied, and the algorithm based on linear interpolation from discrete frequency response function, which is calculated by fast Fourier transform, was presented. The study results show that, firstly integral period sampling is preferred for improving the precision of damping ratio and frequency, secondly due to sampling truncation, the estimation damping ratio isn't zero for a system without damping, and the real damping isn't equal to apparent damping subtracted by PDW when the real damping ratio is smaller than PDW, thirdly, the relative damping error is approximately proportional to the production of multiplication between sampling period number and real damping ratio, and the production corresponds to the attenuation of the truncated signal, lastly, to guarantee the error less than 10%, the actual damping ratio should be larger than one and a half of PDW.
出处
《机械强度》
CAS
CSCD
北大核心
2002年第4期510-514,共5页
Journal of Mechanical Strength
