摘要
为了解决微致动器结构功能函数不能明确表示的可靠度分析问题,以BP神经网络、响应面法为基础,提出了求解可靠度的几何分析新方法:对神经网络响应面法加以改进并利用其拟合极限状态函数,得到显示的结构功能函数表达式;利用均值波动法对微致动器局部应力最大处进行可靠度分析,求得结构可靠度因子β为3.353。与JC法比较,方法简便可行,为结构可靠度计算问题提供了新的有效的思路和方法。
In order to solve the problem that performance function is not expressed explicitly, a new geometrical analysis method of the flexural micro -actuator structural reliability is presented based on BP neural network and response surface method. The BP neural network is employed to improve the response surface method and get the limit state function. The reliability index 13 is evaluated by the mean fluctuation algorithm on the maximum local stress place and the result is 3. 353. Compared with JC method, the proposed algorithm is feasible and effective, and provides a new effective idea and approach to the calculation of structural reliability index.
出处
《计算机仿真》
CSCD
北大核心
2009年第7期175-177,181,共4页
Computer Simulation
基金
辽宁省智能信息处理重点实验室开放课题(2005-5)
关键词
几何原理
神经网络
微致动器
均值波动法
可靠度
Geometrical principle
Neural network （ NN ）
Flexural electro - thermal actuator
Mean fluctuation algorithm
Reliability
