6 Conclusions
The effectiveness of the proposed method has been demonstrated by virtue of the experimental investigation of a cantilever beam. The experimental results show that both excitation and response can be estimated simultaneously using the limited measurement data, and the dynamic responses at the locations where no sensors are installed can be reconstructed accurately. By comparing experimental results of different loading cases, the optimal sensor placement achieved by the proposed method shows its robustness and applicability in various loading conditions, provided that the external loading location, the mode number used in response reconstruction and environmental noise characteristics remain the same. Through comparison of the response reconstruction results of the cantilever beam between three different sensor placement schemes (OSP, SP1, SP2), the results reveal that the proposed optimal multitype sensor placement method has remarkable superiority compared with the alternative sensor placement confi gurations and a multi-type sensor system with inclinometers, accelerometers and displacement meters could provide more accurate and reliable monitoring for fl exible building structures.