Active seismic control using neural network

Abstract

A novel active control algorithm is proposed in this present study. It is assumed that the shape of the time history of control force is same as that of the shape of time history of ground excitation. Based on the above assumption, control force is determined for a specific target reduction in displacement, velocity and acceleration. Time histories of ground motion compatible to response spectrum are generated for seismic zone three as per Indian seismic conditions. A neural network is trained to generate the desired control force for a specific target reduction. The results of the study show that the reduction provided by the proposed control algorithm matches well with the target reduction.