机构动作可靠性估计的自适应极值响应面法  

An adaptive extremum response surface method for mechanismaction reliability estimation

作　　者：文浩 侯保林[1] WEN Hao;HOU Baolin(School of Mechanical Engineering,Nanjing University of Science and Technology,Nanjing 210094,China)

机构地区：[1]南京理工大学机械工程学院,江苏南京210094

出　　处：《哈尔滨工程大学学报》2024年第3期581-589,共9页Journal of Harbin Engineering University

基　　金：总装备部预研项目(104010401)。

摘　　要：针对存在随机-区间混合不确定性的机构动作可靠性估计问题,本文提出了一种基于自适应极值响应面的高效计算方法,将其转化为随机不确定性下的动作可靠度上下界求解问题。使用麻雀搜索算法优化的混合核极限学习机构建从混合不确定性变量到极限状态函数响应值的初始响应面和从随机变量到极限状态函数响应极值的极值响应面;利用结合主动学习与反向学习的自适应加点策略选取极限状态曲面附近的样本点更新极值响应面以提高其精度与效率;最后结合极值响应面和蒙特卡罗仿真算得到动作可靠度上下界的近似解。通过数值案例和回转链式输送机的工程案例对所提自适应极值响应面方法的高效性与准确性进行了验证,为随机-区间混合不确定性下的机构动作可靠性估计提供了一种参考。A high-efficiency calculation method based on adaptive extremum response surface(AERS)is proposed to address the problem of mechanism action reliability estimation under random-interval mixed uncertainty.This is then transformed into the problem of solving the upper and lower bounds of action reliability under random uncertainty.The mixed kernel extreme learning machine optimized by the sparrow search algorithm is used to construct the initial response surface from mixed uncertainty variables and the extremum response surface(ERS)from the random variables and transform them into the limit state function(LSF)response value and the LSF response extremum,respectively.An adaptive infilling strategy combining active learning and opposition-based learning is then used to select the sample points near the limit state surface to update the ERS and thus improve its accuracy and efficiency.Finally,the approximate solutions of the upper and lower bounds of the action reliability are obtained by the ERS and Monte Carlo simulation.The efficiency and accuracy of the proposed method are then verified by a numerical case and an engineering case of a rotary chain conveyor.The proposed method provides a reference for the mechanism action reliability estimation under random-interval mixed uncertainty.

关 键 词：动作可靠性 混合不确定性 极值响应面 自适应加点策略 混合核极限学习机 麻雀搜索算法 主动学习 反向学习 

分 类 号：TJ307[兵器科学与技术—火炮、自动武器与弹药工程] U462.3[机械工程—车辆工程]