摘要
人工气候室中的参数测试存在着测量误差大和重复性差的缺点,严重影响测试结果的准确性.针对这个问题,研究了集温度、湿度、光照度、CO2含量等参数测试为一体的人工气候实验室智能测试专家系统,详细介绍了该系统所采用的基于格罗贝斯准则的疏失误差剔除方法以及数据挖掘方法.经实际应用验证,该系统消除了人工气候室参数测量的不确定性,可获得比传统的算术平均值更准确的测试结果,并能依据各参数间的关联关系对测试的手段和方法进行有效的调整.
The accuracy of the parameter testing with a phytotron is greatly reduced by errors and low repeatability. This paper introduced an intellectual testing expert system for the phytotron, involving the testing of temperature, humidity, light, carbon dioxide, etc., and described precisely the mistake error rejecting method and the data mining method adopted by the system. It has been proved by online testing that the system is free of testing uncertainty with more accurate results than traditional systems and is self - adjustable because of the association relations acquired from data mining.
出处
《湖南大学学报（自然科学版）》
EI
CAS
CSCD
北大核心
2006年第4期60-63,共4页
Journal of Hunan University:Natural Sciences
基金
国家技术创新资助项目(国经贸技术[2002]845号)
关键词
智能控制
人工气候室
智能测试
疏失误差
数据挖掘
intelligent control
phytotron
intellectual testing
mistake error
data mining
