摘要
针对试验或生产过程中测试方法、工艺、材料、配方等变更,如何评价产品性能是否受到影响的问题,本论文采用对比试验方法,以板材测试方法选择、涂层表面张力变化、玻璃化学增强用原料调整为例,运用假设检验中的秩和检验、t检验、F检验的方法,依次给出了母体分布情况不知及母体服从正态分布的情况下,评价某项指标测试数据均值、方差是否有显著差异,可为研制及生产过程技术状态管理、变更决策提供支持。
There are some changes about experiment and productive process, such as test method, process, and materials, ingredient, etc. How do we evaluate the performance of the product? How to take a conclusion about yes or no? The selection of sheet testing method, the change of surface tension of the coating and the adjustment of raw materials for ion-exchange glass are taken as an example. The article used comparative experiments with the help of hypothesis test, which are rank test, and t test and F test, to evaluate significant difference of the mean value and variance, whether the parent distribution is not known or obeys the normal distribution. It can provide support for decision of configuration change.
引文
[1]盛骤,谢式千,潘承毅.概率论与数理统计(第四版)[M].高等教育出版社,2008:173-217
[2](美)John A.数理统计与数据分析[M].机械出版社,2016:228-259
[3]张雨宇,梁中耀,陆文涛,刘永,郭怀成.基于正态分布假设检验的湖泊营养盐浓度分级评价方法研究[J].环境科学学报,2017,37(6):2387-2393
[4]王学杰,董向毅,许维江.假设检验在热加工领域计数型质量控制中的应用[J].理化检验-物理分册,2016,52(9):646-649
[5]兰冲锋.统计学教学中关于假设检验问题探讨[J].湖南城市学院学报(自然科学版),2016,25(2):303-304
[6]凌树森.试验数据的统计处理和误差分析第七讲对比试验结果的统计处理[J].理化检验-物理分册,2001,37(11):502-506
[7]戚维明.全国质量管理(第三版)[M].中国科学技术出版社,2010