摘要
构建沥青路面试验检测数据验证方法体系,进而探索出识别和监控虚假数据的计算流程,以防范和遏制虚假数据的蔓延。将经验判断和数学分析相结合,从试验检测数据的逻辑联系和正态性、方差可分解性等基本特征着手,综合异常数据识别、统计检验、误差分析等数据验证方法,提出了沥青路面试验检测数据验证一般流程。工程应用表明,逻辑性检查和异常数据识别可为统计检验提供较好的数据环境,而统计检验需按K-S检验、F检验和t检验3步依次进行,方能保证数据验证的有效性。误差分析作为验证方法的重要补充,有利于促进质量抽检争议的解决,保证试验检测向规范抽样、科学分析的方向发展。
For establishing test data validation methodology of asphalt pavement and explore the calculation process of false data identification and monitoring to prevent and curb the spread of false data,combined experienced judgment with mathematical analysis,some data validation methods which include abnormal data recognition,statistics test and variance analysis were integrated,and the general process of data validation of asphalt pavement test was presented considering the logical link,normality and variance decomposition of test data.Application results show that logical examination and abnormal data recognition could provide better data environment for statistical test,and statistical test needs to develop a three-tiered analysis approach consisted of K-S test,F test and T test to ensure the validity of data validation.As an important supplement to the validation method,error analysis could promote dispute resolution of quality sampling and ensure the test and inspection development goals to be standardized sampling and scientific analysis.
出处
《公路交通科技》
CAS
CSCD
北大核心
2012年第5期6-12,共7页
Journal of Highway and Transportation Research and Development
基金
交通运输部西部交通建设科技项目(2011318223710)
关键词
道路工程
沥青路面
数据验证方法
统计检验
试验检测
road engineering
asphalt pavement
data validation method
statistical test
test and inspection
