PDEM-BASED STOCHASTIC OPTIMAL CONTROL OF STRUCTURES

摘要：

In this paper,the performance evolution,as gaged by probability density of state and control force,of controlled systems is firstly investigatied from the perspectives of stochastic events description,as indicated in the probability density evolution method (PDEM). Based on this,a physical approach for stochastic optimal control of structures is further proposed,which involves two step optimizations. A family of optimal control criteria used in the second step optimization for designing the controller parameter is elaborated,including those based on statistics of system responses,based on exceeding probability of single system quantity,and based on exceeding probability of multiple system quantities. A base-excited single-story structure installing active control devices is studied for illustrative purpose. It is noted that the effectiveness of response control hinges on the physical meansings of the optimal control criteria,and an accurate control of structural performance is reached using the developed physical stochastic optimal control methodology.

关键词： Probability density evolution method Stochastic optimal control LQG control Control criterion Exceedance probability Active control

作者: Jie Li Yong-Bo Peng

作者单位: State Key Laboratory of Disaster Reduction in Civil Engineering & School of Civil Engineering,Tongji State Key Laboratory of Disaster Reduction in Civil Engineering & School of Civil Engineering,Tongji

会议类型: 国际会议

会议名称: The Eleventh International Symposium on Structural Engineering(第十一届结构工程国际研讨会)

会议地点: 广州

会议语种:英文

页码: 811-816

在线出版日期: 2010-12-01（万方平台首次上网日期，不代表论文的发表时间）