摘要
介绍了组合统计检验法的基本思想,提出了区间拓展法用于实际化工过程稳态检验的新策略。实例考核表明,新方法显著地降低了第一类错误的概率,可有效地用于实际化工过程的稳态检验。
The detection of changes in steady states is important in quality control, data reconciliation, process simulation, and monitoring applications. A composite statistical test was developed by Nedsimhan(1986) for this purpose, but this method has some defects in practical applications. There are contradiction between improving the power of test and reducing the probability of type I error in using this method. In this paper we present a new strategy called period extending method to solve this problem. A key element of this approach is to distinguish the true changes from the false changes. In this note we propose that for three successive periods, when the change between period one and period two is detected, one more test should be carried out between period one and period three to determine whether the change is true or not. Applicational studies were carried out to compare the performance of this methods with the composite statistic test method. The results show that the probability of type I error by the new method decreases significantly, while the powers of test by the two methods are almost the same. Thus the new strategy is an attractive method for detecting changes of steady states.
出处
《高校化学工程学报》
EI
CAS
CSCD
北大核心
2000年第1期71-76,共6页
Journal of Chemical Engineering of Chinese Universities
关键词
稳态检验
统计检验
化工过程计算
过程稳态
steady state
changes detection of steady states
statistical test