修正时序法在石化装置数据校正中的应用研究

为了有效地克服传统统计检验和线性化处理方法进行过失误差侦破、数据校正与参数估计时存在的局限性，对现有过程测量数据校正技术进行了综合分析，探讨了数据分类、过失误差侦破和过程测量数据校正等问题，以具有代表性的化工装置和炼油装置为研究对象，提出了修正时间序列分析法，给出了时间序列分析法的基本思想、概率模型和特点，以及含随机误差的数据校正方法，并将时序法用于过失误差侦破。在大型化工装置的测量数据校正结果表明，该方法是实用性强、方便有效的数据校正新策略。

Study on modified time series method of data correct in petrochemical equipment

To eliminate the limitation of the traditional method of statistical check and linear planning treatment, the existing techniques of measurement data correct in petrochemical process were analyzed firstly. Then the issues of data classification, error detection and data correction were investigated. Furthermore, the modified time series method was proposed for typical petrochemical equipments. The basic idea, probability model and characters of the method were stated. The method was applied to the large petrochemical equipment and measurement data correct, and it is shown that the shortcomings of the traditional method of data correction have been overcome.