论文部分内容阅读
以典型的醋酸乙烯化工过程模型数据和英国石油Kwinana炼油厂实际警报数据为基础,研究了化工过程中通过发掘关联警报标签集来支持合理化的报警设置,提出了依据交叉效果测试来进行事件分割和数据过滤。
Based on the typical model data of vinyl acetate chemical process and the actual alarm data of BP Kwinana Refinery, this paper studies the rationalization of alarm settings by exploring the association of alarm labels in the chemical process, and proposes a method based on the cross effect test for event segmentation and Data filtering.