Reliability based robust design optimization for tuned mass damper in passive vibration control of deterministic/uncertain structures
Reliability based robust design optimization for tuned mass damper in passive vibration control of deterministic/uncertain structures
复制标题
DOI：
10.1016/j.jsv.2012.12.014
复制
发表时间：
2013-04
影响因子：
4.7
通讯作者：
H. Yu;F. Gillot;M. Ichchou
中科院分区：
文献类型：
--
作者：
H. Yu;F. Gillot;M. Ichchou
在这项工作中，我们提出了一个基于可靠性的鲁棒设计优化（RBRDO）框架，致力于调谐质量阻尼器（TMD）的被动振动控制。与传统的随机设计优化（CSDO）方法相比，该方法具有上级优点，可以同时考虑结构的质量和安全性，适用于确定性和不确定性结构。RBRDO的主要挑战是由于优化过程中涉及大量的可靠性分析而导致的效率低下。为了克服这个问题，我们开发了一种基于可靠性灵敏度的序贯RBRDO，它将可靠性分析与优化分离，从而减少了可靠性分析的数量。序贯策略的原理是构造一个等价的确定性约束来代替原来的概率性约束，这是通过局部指数近似来实现的。关联可靠性灵敏度分析不仅可以处理随机设计变量的问题，而且可以处理确定性设计变量的问题。案例研究涉及的线性TMD系统受到随机基础加速度，这是由一个过滤平稳白色噪声过程建模。数值仿真结果表明，与CSDO相比，RBRDO是TMD优化设计的有力工具，并能提高TMD的有效性。
We propose in this work a reliability based robust design optimization (RBRDO) framework dedicated to the tuned mass damper (TMD) in passive vibration control. Superior to conventional stochastic design optimization (CSDO), in which all structural parameters are assumed to be deterministic, this framework is applicable for either deterministic or uncertain structures, and is capable to take quality and safety into account simultaneously. The main challenge of RBRDO is the low efficiency caused by large number of reliability analyses involved in the optimization procedure. To overcome this, we develop a reliability sensitivity based sequential RBRDO, which decouples reliability analysis from the optimization so that the number of reliability analysis is reduced. Principle of the sequential strategy is to construct an equivalent deterministic constraint instead of the original probabilistic one, which is realized by using the locally exponential approximation. The associated reliability sensitivity analysis can not only deal with problems with random design variables, but also tackle problems with deterministic ones. The case study is related to a linear TMD system subjected to a random base acceleration, which is modeled by a filtered stationary white noise process. Compared with CSDO, numerical simulations demonstrate that RBRDO is a powerful tool for design optimization of the TMD and also provides improvements in the effectiveness.