摘要
针对参数法计算织物中甲醛暴露点浓度控制上限受取样样本量限制这一影响计算值准确性的问题,随机抽取国产织物样品并对其甲醛含量进行测定,以实际取样测定值为依据,采用Bootstrap模拟取样方法对测定原始数据集进行重复取样,建立多次模拟取样的甲醛均值含量分布,并对甲醛暴露点浓度控制上限(95%UCL)进行计算,所得到的甲醛暴露点浓度控制上限为52.73 mg/kg。这种方法克服了数据非正态分布及数据量小的缺陷,对数据分布形态未知的情况更具通用性。
To break through the limitation of sample amount in calculating the upper control limits for exposure point concentration of formaldehyde content in textiles by parametric method, a non-parametric method, Bootstrap resampling method was introduced to simulate the sampling process based on the testing data of formaldehyde of textile samples collected randomly in domestic marl^et to generate a mass of simulation dataset. Upper control limit for exposure point concentration of formaldehyde （95 % UCL） was calculated based on this bootstrap resampling dataset and got a result of 52.73 mg/kg by this method. This method can be used in wide range, even in the conditions of non-normality of data distribution and limitation of sample amounts.
出处
《纺织学报》
EI
CAS
CSCD
北大核心
2009年第6期95-98,共4页
Journal of Textile Research
基金
十一五科技支撑项目(2006BAK10B03)
国家质检总局科技项目(2008IK090)
山东省科技攻关项目(2008GG10008005)
作者简介
纪雷（1970-2008），男，高级工程师，博士。研究方向为分析化学、环境毒理学、化学计量学等。
林雨霏，通讯作者，E-mail：viennalinn@yahoo．com．cn。
