MTMD控制下随机结构的动力可靠度分析  

Dynamic reliability analysis on stochastic structure with MTMD control

作　　者：卜国雄[1] 谭平[2] 周福霖[2] 

机构地区：[1]哈尔滨工业大学深圳研究生院,深圳518055 [2]广州大学教育部工程抗震减震与结构安全重点实验室,广州510405

出　　处：《振动与冲击》2009年第5期119-123,共5页Journal of Vibration and Shock

基　　金：国家自然科学基金资助项目(50608021);国家教育部博士点专项基金资助项目(20061078001)

摘　　要：考虑到实际工程结构的不确定性,基于遗传优化的神经网络响应面法,进行了MTMD控制下随机结构的动力可靠度分析,并对TMD及MTMD控制下结构的动力可靠度进行对比。该方法不仅具有传统神经网络响应面法的特性,而且引进了遗传算法的全局随机搜索的优点,可以精确地逼近随机结构的功能函数表达式,有效地减少用JC法求解可靠度指标的迭代次数,节省时间。算例分析表明了本文方法的有效性和准确性,对于随机结构,MTMD比TMD能更好地提高结构的动力可靠度。Considering the uncertainties of practical civil engineering structures,the neural network response surface method(NNRSM) combined with genetic optimization algorithm(GOA) was proposed to calculate the dynamic reliability of stoch-astic structure with TMD or MTMD.The method has the advantages of both the NNRSM and the GOA,which is provided with the global searching ability and can lead to optimum solution accurately of the performance function.And the number of iteration steps of JC method can be reduced effectively.The results of instances show that the method is effec tive and accurate.For the stochastic structure the MTMD can greatly improve the dynamic reliability more than TMD.

关 键 词：TMD MTMD 遗传优化 响应面法 随机结构 动力可靠度 

分 类 号：TU311[建筑科学—结构工程]