平面度误差粒子群算法评定不确定度评估.PDFVIP

第39卷第 1期 中国测试 Vol.39 No.1 2013年 1月 CHINA MEASUREMENT TEST January袁2013 平面度误差粒子群算法评定的不确定度评估 姜焰鸣袁 刘桂雄 渊华南理工大学机械与汽车工程学院袁广东 广州 510640冤 摘 要院形状误差智能评定采用概率化随机搜索袁评定结果存在分散性袁针对此问题提出一种不确定度评估方法遥 以 平面度误差粒子群算法评定为例袁根据智能评定结果概率分布特性袁采用茁分布统示法拟合其概率分布袁并对平面 度误差进行区间估计袁最后选取仿真平面测量数据进行不确定度评估实验遥结果表明袁样本个数N=100尧截取百分位S 数Q=20时袁估计区间能够有效包容平面度误差值袁且平均宽度较小袁验证基于茁分布统示法的智能评定不确定度评 p 估可行性遥 关键词院平面度误差曰粒子群算法曰概率分布曰茁分布统示法曰不确定度评估 中图分类号院TB922曰O211.2曰TP391.9曰TM930.115 文献标志码院A 文章编号院1674-5124渊2013冤01-0013-04 Uncertainty evaluation of flatness error evaluation based on PSO JIANG Yan-ming袁 LIU Gui-xiong 渊School of Mechanical and Automotive Engineering袁South China University of Technology袁 Guangzhou 510640袁China冤 Abstract: Form error intelligent evaluation is using probabilistic random seeking so that evaluation result has a feature of the dispersion袁 for this problem an uncertainty evaluation method was proposed. Taking the flatness error PSO evaluation for an example袁 茁 distribution uniform expression method was used to fit intelligent evaluation results according to its probabilistic distribution features袁 and then the interval estimation of flatness error was given. Simulation measurement data was selected for uncertainty evaluation. The results showed that when N is S equal to 100 and Q is equal to 20袁 the flatness error interval estimated effectively includesp flatness error value袁 also the average range is short袁 so that the feasibility of intelligent evaluation uncertainty evaluation based on 茁 distribution uniform expression method was verified. Key w