Optimization of hybrid cable networks with dampers and cross-ties for vibration control via multi-objective genetic algorithm

For long-span cable-stayed bridges, interconnecting adjacent cables using cross-ties and connecting each cable to bridge deck become a promising solution for cable vibration mitigation, leading to hybrid cable networks. However, optimizing the system parameters and configuration for best vibration control effects is still challenging. Therefore, this study proposes an optimization method based on three important ingredients, i.e., a general and efficient numerical model of cable networks, a reasonable definition of the optimization objectives, and a multi-objective optimization method. The cable network is modeled using the component mode synthesis method with modal expansion and quasi-static correction, the optimization objectives consider both multi-mode damping effects and stiffness enhancement, and the multi-objective genetic algorithm is then employed. The proposed method is first verified by comparing with analytical methods when applying to relatively simple cable-damper systems. Subsequently, a five-cable network is studied using the method for optimal design of cross-tie and damper parameters. Furthermore, the method is applied to optimize cross-ties and dampers for the eight longest cables in the Sutong Bridge, considering different numbers, locations and anchoring conditions of the cross-ties. The results have demonstrated the effectiveness of the proposed method in dealing with real-world large-scale cable networks. It is shown that the optimization can ensure that vibrations in local modes of the cable network are dominated by cable segments with a damper installed in the span and hence are sufficiently damped. Besides, damping ratios of hybrid cable networks are shown to be sensitive to the cross-tie location and parameters while they are relatively insensitive to the damper coefficients.

Automated summary

This study presents an optimization method for cable networks considering multiple damping effects and stiffness enhancement, using a multi-objective genetic algorithm for optimization.