平面度误差粒子群算法评定的不确定度评估

Uncertainty evaluation of flatness error evaluation based on PSO

形状误差智能评定采用概率化随机搜索,评定结果存在分散性,针对此问题提出一种不确定度评估方法。以平面度误差粒子群算法评定为例,根据智能评定结果概率分布特性,采用β分布统示法拟合其概率分布,并对平面度误差进行区间估计,最后选取仿真平面测量数据进行不确定度评估实验。结果表明,样本个数NS=100、截取百分位数Qp=20时,估计区间能够有效包容平面度误差值,且平均宽度较小,验证基于β分布统示法的智能评定不确定度评估可行性。

Form error intelligent evaluation is using probabilistic random seeking so that evaluation result has a feature of the dispersion,for this problem an uncertainty evaluation method was proposed.Taking the flatness error PSO evaluation for an example,β distribution uniform expression method was used to fit intelligent evaluation results according to its probabilistic distribution features,and then the interval estimation of flatness error was given.Simulation measurement data was selected for uncertainty evaluation.The results showed that when NS is equal to 100 and Qp is equal to 20,the flatness error interval estimated effectively includes flatness error value,also the average range is short,so that the feasibility of intelligent evaluation uncertainty evaluation based on β distribution uniform expression method was verified.