Calibrating the noise level is an essential step in the design of vibration-based structural health monitoring(SHM)system.It is beneficial to the optimization of the test configuration and the later modal parameter identification.Based on the investigation of the overall data acquisition process,the main source and the stochastic characteristics of the measurement noise are analyzed.Based on the least squares method,an n-channel method is developed for the channel hard-ware noise quantification and a strategy is proposed to quantify the synchronization error.The former allows one to estimate the noise with an arbitrary number of sensors,utilize the information from all channels and thus yield more robust estima-tion;the latter allows one to realize the estimation of synchronization error of sub-milliseconds by using the ratio of the cross-power spectrum density based on the frequency-domain representation of signals with time lag.The correctness,ro-bustness,and feasibility of the proposed methods are verified in a practical SHM project.
关键词
振动与波/结构健康监测/拾振器/测量噪声/同步误差/最小二乘法
Key words
vibration and wave/structural health monitoring/vibration sensors/measurement noise/synchronization error/least squares estimation
维普
万方数据