基于双侧截尾正态分布的机构运动可靠度研究  

Research on Motion Reliability of Mechanism Based on Doubly Truncated Normal Distribution

作　　者：赵瑞学[1] 钱林方[1] 陈龙淼[1] 石海军[1] 

机构地区：[1]南京理工大学机械工程学院,江苏南京210094

出　　处：《计算机仿真》2011年第2期248-251,共4页Computer Simulation

基　　金：国家部委基金项目;南京理工大学优秀博士研究生培养基金(2010)

摘　　要：研究机械运动机构可靠度优化计算,为了更准确地分析机构运动可靠度,提出了随机变量取值的有界性。从双侧截尾正态分布的基本概念出发,推导了截尾正态分布的期望均值和方差的一般形式,并运用逆变换法获得了一种双侧截尾正态分布随机数的抽样公式。同时基于蒙特卡罗法给出了含截尾分布随机变量的机构运动可靠度计算方法。并采用一个典型机构,对运动可靠度问题的随机变量,采用截尾正态分布和一般正态分布进行比较计算,前者比后者得到的运动可靠度更高。结果表明:基于截尾分布抽样的蒙特卡罗法求解机构运动可靠度比传统方法更为准确。To analysis the motion reliability of mechanism more accurately,the boundedness of random variables are considered.The general formulae of mean and variance have been derived from definition of truncated normal distribution.And a sampling formula for truncated normal distribution has been acquired via inverse transform sampling method.Analysis method for motion reliability with truncated distribution random variables is given based on Monte-Carlo method.And then,motion reliability results are compared with non-truncated distribution using a classical example,which shows that the reliability of mechanism with truncated distribution variables is higher than non-truncated distribution variables.The results indicate that the method presented in this paper for reliability analysis is more accurate than traditional method.

关 键 词：截尾正态分布 机构可靠度 逆变换抽样 蒙特卡罗法 

分 类 号：TB114.3[理学—概率论与数理统计]