论文部分内容阅读
乙烯裂解炉管在石化行业中至关重要,其炉管质量问题对整个化工过程安全生产也有很大影响,它由离心铸造过程获得。离心铸造过程属于间歇过程,但它在每个批次中也包含离散数据信息,而非时间序列。为了实现对炉管质量的监测,本文利用离心铸造过程中可以获得的参数,提取时间序列中的关键统计信息,形成每个批次的多变量样本信息,用主元分析方法建立炉管质量监测模型,所得到质量监测模型的目前正确率为80%,误报率为10%。通过贡献图的方法对预报结果进行了分析,剖析了引起炉管质量问题的原因,为工业炉管的正常生产提供了指导。
Ethylene cracking furnace tube in the petrochemical industry is crucial, the quality of the tube of the entire chemical process safety also has a great impact, it is obtained by the centrifugal casting process. Centrifugal casting process is a batch process, but it also contains discrete data information in each batch, rather than the time series. In order to monitor the quality of the tube, this paper uses the available parameters in the centrifugal casting process to extract the key statistical information in the time series to form the multivariate sample information for each batch. The principal component analysis method is used to establish the tube quality monitoring Model, the current correct rate of quality monitoring model obtained is 80% and the false alarm rate is 10%. Through the method of contribution graph, the forecast results were analyzed, the reason of the quality problems caused by the furnace tubes was analyzed, which provided guidance for the normal production of the industrial furnace tubes.