Non-parameter system identification of isolated bridge by neural networks algorithm

The objective of this theme is to non-parameter identify and to compare the dynamic properties of the Lion-head river bridge located at Chia-I. The east bound of the bridge is designed and constructed as conventional, and the west bound as bridge isolated by lead-rubber bearings. Signals collected from the accelerometers installed on the bridge by the Central Weather Bureau will be processed. Back-propagation algorithm of Neural networks will be adopted and the nonlinear behavior of lead-rubber bearings will be simulated.