典型炼油装置数据校正技术的研究  

The study of the typical oil refining unit data reconciliation

作　　者：夏茂森[1] 房韡[2] 王锋[2] 

机构地区：[1]中国石化股份有限公司齐鲁分公司信息技术管理部,山东淄博255408 [2]中国石化股份有限公司石油化工科学研究院,北京100083

出　　处：《计算机与应用化学》2009年第7期883-888,共6页Computers and Applied Chemistry

摘　　要：化工过程测量数据作为反映装置运行状况的特征信息,是实现计算机过程控制、模拟、优化和生产管理的基本依据。研究过程数据校正技术,对实现装置优化控制与管理具有重要理论意义和现实意义。现有理论研究大都采用传统统计检验和线性化处理方法,在实际应用有较大局限性。本文在对已有数据校正技术分析的基础上,提出将修正的时间序列分析法用于测量数据校正。综合考虑数据的空间冗余和时间冗余,充分利用过程的历史数据,建立了时间序列概率模型,并针对含随机误差数据和含过失误差数据两种情况,从时序法平均值、阶跃过程模型等方面详细探讨数据校正方法。将新的数据校正方法用于典型常减压蒸馏装置,结果表明,新方法能够侦破出数据中含有的过失误差;校正值与真值的平均偏差非常小,具有足够的精度保证数据的准确性;修正的时间序列分析法用于数据校正能克服传统方法的局限性。Chemical process.measurements as the characteristic information of plant operation state, is one of the foundation of process control, simulation, optimization and manage. For the optimization control and management of chemical plant, it is very significance to study the data reconciliation. At present, there are many achievements in the data reconciliation that has Been applied in calculating material and scheduling production. However, the data reconciliation uses the statistic test and linearization in the gross errors detection, data reconciliation and parameter estimation so that there are a plenty of limitedness. To solve the problem, the modified time series method is used in the data reconciliation on the Basis of analyzing the known data reconciliation. The new technology utilizes the historical data, spatial and temporal redundancies and Builds time series probability model. At the same time, for the random errors and the gross errors, the new method is illuminated in the time series average and Bound model and so on. The new data reconciliation is used in the atmospheric and vacuum distillation unit. The result shows that the new method can effectively detect the gross errors and provide high quality data for process industries. The modified time series analysis method is very significant for the data reconciliation so that can overcome the problem in the existing data reconciliation.

关 键 词：数据校正 过失误差侦破 参数估计 修正时序法 常减压装置 

分 类 号：TE624[石油与天然气工程—油气加工工程]