Analytical and Experimental Analysis of a Neural-Based Hybrid Vibration Control for Cable-Stayed Bridges

Abstract

Abstract Intelligent civil engineering structures which posses the capability to resist catastrophic seismic activities and high load degradation are of great interest to the scientific community and to the population at large. This research focuses on vibration reduction in large civil engineering structures using intelligent control techniques and multiple low power actuation devices. Electromagnetic shakers and hydromechanical adaptive filters are utilized as actuation devices. A complex, nonlinear civil engineering structure is considered, mathematically modeled, analyzed, and retrofitted with these actuators for experimental purposes.. A 1/150 scale model of the existing cable-stayed bridge was constructed from aluminum, steel, high tensile strength piano wire, and guitar machine heads (to facilitate cable tensioning) and is mounted on a rectangular steel base. The prototype has a standard cable-stayed bridge design with dual A-frame towers and a fan-type cable arrangement. It has an overall length of 10.5 feet and a tower height of approximately 1.5 feet. It is the goal of this paper to present intelligent control schemes that utilize a fixed controller design for disturbance rejection and uses intelligent adaptive controller to handle small plant perturbations. An adaptive control strategy that encompasses vibration reduction, system identification, hybrid actuation (i.e. the use of active, semi-active, and passive actuators in a single control system), and combined feedforward / feedback control is sought. Intelligent methods of system identification and disturbance prediction are also incorporated. Experimental results obtained from the experimental prototype are presented.