摘要
小波滤波能有效降低化工过程测量数据的随机误差,但却无法识别测量数据中是否存在过失误差。为此,本文通过总结大量小波滤波数据校正实例中校正值、分解层数与过失误差之间存在的关系,提出了三者之间的关系公式,并根据此公式侦破识别过失误差。对Aspen Dynamic模拟产生的测量数据的校正结果表明,文中提出的公式准确的反映出了校正值、分解层数和过失误差的关系,并且利用该公式能够有效地侦破和识别过失误差。
Wavelet has a good performance on chemical measurements error reduction, but it cannot identify the presence of gross errors in measured data. For this reason, this paper summarizes a large number of examples of data reconciliation, and then obtains a relationship between correction value, decomposition level and gross error, and a formula is derived from this relationship, it is used to detect and identification gross error. According to the reconciliation of measurement data which is simulated by Aspen Dynamic, the formula accurately reflects the relationship between the correction value, decomposition level and gross error, and it can effectively detect and identify the gross errors.
出处
《山东化工》
CAS
2014年第2期38-40,共3页
Shandong Chemical Industry
基金
重质油国家重点实验室开放课题基金资助项目(201103004)
关键词
小波滤波
过失误差侦破和识别
数据校正
wavelet
gross error detection and identification
data reconciliation