7. Conclusions
This paper presents the structural sensing method using the accurate artificial vibrator and the wireless sensor network. The accurate artificial vibrator generates very accurate harmonic forces and the wireless sensor network measures the acceleration response as time-synchronizing with the vibrator. Therefore, the force function and the displacement time series, which is converted from the data of wireless sensor node, are described as an ARX model. By solving the ARX equations, Green functions are accurately estimated. The experimental results show that the residual error rate of eigen frequency, damping ratio, and peak amplitude of Green function obtained by the proposed method are less than 0.01%, 3.5%, and 3.0%, respectively. The numerical simulation also supports that the Green function is accurately obtained by the proposed method. Because of the accuracy, the Green functions or the raw time series data obtained by means of the present method can be used as input data of the existing global-damage detection methods described in Section 2. In a future work, the existing global-damage detection method will be applied to the measurement of the present approach. Application to real civil infrastructures is also important future work. Since the power of prototype vibrator is very small and the wireless communication range is short, it is difficult to apply the prototype system to a large civil infrastructure such as a suspension bridge. Therefore, we currently try to apply the preset method to the health monitoring of road signs as described in Fig. 1. Development of powerful vibrator for a large civil infrastructure is also a needed future work.
