Intelligent active control of a benchmark cable-stayed bridge

Abstract

A neural-fuzzy control strategy for cable-stayed bridge structures subjected to earthquake excitation is presented by employing electromagnet-driven active mass dampers (EMDs). The active EMD is an innovative control device for the earthquake response mitigation of structures. Ensuring fast selection of the control voltage in accordance with the characteristics of seismically excited bridge structures is one of the most important problems for EMD systems. In the proposed control strategy, a neural-network technique is used to solve the time-delay problem, and fuzzy logic algorithms are used to determine the control voltage from the response of the EMD. The control performance of the proposed neural-fuzzy control strategy was analysed for the benchmark control problem of a seismically excited cable-stayed bridge with an EMD system. The simulated results show that the proposed active neural–fuzzy control strategy can quickly determine an accurate voltage for the EMD system, and mitigate the seismic response of cable-stayed bridges.