Abstract
<title>Abstract</title> <p>To improve the vibration control performance of adjacent structures connected by viscous dampers under seismic excitation, this study proposes a damper parameter optimization method that integrates the MATLAB genetic algorithm with the SAP2000-OAPI. The effectiveness of the method is systematically verified through shaking table tests and an intelligent prediction model. First, a dynamic analytical model of adjacent asymmetric twin-tower structures was established. The installation locations and damping coefficients of viscous dampers were selected as optimization variables. Based on these variables, a GA-SAP2000 integrated optimization framework was developed to determine the optimal parameters for both single-damper and double-damper connection systems under different earthquake records. Subsequently, a shaking table test involving a 15-story main structure and a 7-story substructure was conducted to validate the finite element model and the optimized schemes. The results show that the discrepancies between the natural frequencies of the numerical and experimental models for the first two modes are within 5%. In addition, the errors in floor dynamic responses are less than 10%, indicating that the proposed finite element model is reliable. Compared with the uncontrolled adjacent structure system, the optimized viscous damper-connected system significantly reduces the displacement and acceleration responses of both the main and substructures. Moreover, the double-damper configuration provides better overall control performance than the single-damper configuration. This result indicates that multi-location collaborative deployment can further enhance vibration mitigation effectiveness. Based on these findings, an SSA-BP neural network model was established to rapidly predict the maximum displacement and acceleration responses of both the main and substructures. The results indicate that the proposed model achieves high prediction accuracy for both the training and testing datasets, with coefficients of determination greater than 0.92.</p>