In view of the fixed sampling rate setting, the problems of under sampling and over sampling in the default sampling mode of Android smartphone sensors are discussed in this paper, and a smartphone sensor signal acquisition software with a custom sampling rate and sampling precision is developed, which can synchronously record the data of the sampling epoch, the time interval and the sensor values. Then, both two power spectrum analysis methods and two displacement signal reconstruction methods of analog acceleration signals are compared and analyzed, and the results show that the L-S method and the frequency-domain integral method can effectively process the non-uniform sampling signal of smartphone sensors with noise. At the end, the acceleration signals of Xuzhou Han Bridge are collected from both smartphones (NX 5 and Mi 3) and a high precision inertial integrated navigation system (SPAN-IGM-A1) used for comparison experiments, and the bridge vibration detection and analysis results verify that the portable and low-cost Android smartphone can be used to conduct the bridge vibration detection.

Android smartphone sensors  /  vibration detection  /  non-uniform time series  /  power spectrum analysis  /  frequency-domain integral