中文摘要：

针对随机结构提出了 TMD 的优化设计方法，基于 Gegenbauer 多项式逼近法将随机结构转化为等效确定性扩阶系统，选取该扩阶系统传递函数的 H ∞范数为优化目标，采用遗传算法来优化 TMD 参数。此外，应用概率密度演化理论分析比较了随机结构-TMD 系统控制前后的结构随机地震响应和动力可靠度，并用蒙特卡罗法进行了验证。仿真分析结果表明：采用所提方法优化设计的 TMD 能有效地减小随机结构响应的峰值和方差，明显改善结构在地震激励下的动力可靠度，采用该方法优化设计的 TMD 具有良好的鲁棒性。

英文摘要：

This paper presents a robust optimal design procedure of TMD parameters for stochastic structures based on H ∞ per-formance.A stochastic structure is first transformed into its deterministic equivalent system by Gegenbauer polynomial ap-proximation method.Genetic algorithm is used in order to evaluate TMD′s optimum parameters through H ∞ norm of the transfer function of the deterministic equivalent system as the objective function.The procedure of the optimal design of TMD is illustrated through a numerical example by the probability density evolution analysis.Numerical results show preliminarily that the optimized TMD can effectively suppress the undesirable vibration and improve structural dynamic reliability of the sto-chastic structure under the earthquake.