摘要
以多变量模型预测控制为主要特征的先进控制技术(APC)是比传统的PID控制更为优异的一种控制技术,代表性技术有Aspen公司的DMCplus技术和Honeywell公司的鲁棒多变量预估控制技术(Profit Controller)。连续重整装置投用先进控制系统后,主要装置运行参数更加平稳,提高了产品质量的均匀度,降低了产品不合格率;实现了各主要操作参数的自动协调控制,同时实现了脱戊烷塔底油和脱丁烷塔底油芳烃含量的在线闭环控制,有效降低了操作人员的劳动强度;脱戊烷塔和脱丁烷塔控制器投用后,塔顶温度、塔底温度和灵敏板温度平稳性提高,通过卡边操作,降低了装置的干气产率,液体收率提高了0.15%;由于反应系统和预加氢系统操作平稳性提高,降低了反应炉的燃料气消耗,节省了0.122kg燃料气/t原料,降低了装置能耗,每年产生节能效益432.07万元。在连续重整装置上应用APC技术是当前企业节能降耗、挖潜增效的有效手段。
Advanced Process Control （APC）,characterized by muhivariable model prediction control,is a con- trol technology more advanced than the traditional PID control.Typical technologies include Aspen＇s DMCplus and Honeywell＇s Robust Multivariable Predictive Control Technology （RMPCT）.After an advanced control sys- tem was introduced in a continuous reforming unit,the main operational parameters of the unit became more stable,the uniformity of product quality improved and the percentage of defective products dropped.An auto- matic coordination and control of main operational parameters was achieved.The system also helped achieve an inline close-loop control of the content of arene in depentanizer and debutanizer bottom oil, effectively re- ducing the labor intensity of operators.As a result of the introduction of depentanizer and debutanizer con- trollers ,the stability of column top temperature, column bottom temperature and sensitive plate temperature im- proved.Through optimum constraint control,the dry gas yield of the continuous reforming unit went down and liquid yield increased by 0.15%.Due to improved stability of operation of the reaction system and the prehy- drogenation system,fuel gas consumption of the reaction furnace fell by 0.122kg per ton of feedstock used. The decline in energy use saves the company 4.32 million yuan in energy cost annually.Applying the APC technology in continuous reforming units is an effective approach for a company to saving energy and tapping company potential to increase efficiency.
出处
《中外能源》
CAS
2016年第1期68-73,共6页
Sino-Global Energy
关键词
连续重整
先进控制
模型预测
技术效果
continuous reforming
advanced control
model prediction
effect of technology
