多变量预测控制在异构脱蜡装置中的应用

Application of multivariate predictive control in heterogeneous dewaxing plant

介绍了潞安公司异构脱蜡工艺及控制现状,针对煤基油各单元物料平衡采用常规PID控制效果不理想的问题,采用多变量预测控制(MPC)技术对其进行改造,结合工艺人员实操经验进行建模和控制器设计,实现了对异构脱蜡装置全流程的精细化控制。工业应用效果表明:MPC系统投用后,可稳定加热炉出口温度,优化塔内温度分布和稳定控制关键设备液位,关键被控工艺参数标准偏差降低35%以上,各基础油的产量、质量均得到有效提升。

The heterogeneous dewaxing process and control status of Lu′ an Company were introduced.In view of the problem that the conventional PID control effect was not ideal for the material balance of each unit of coal-based oil,the multivariate predictive control(MPC) technology was used for the modification.Modeling and controller design were carried out combined with the practical experience of the process personnel,to achieve the whole-process fine control of heterogeneous dewaxing plant.The industrial application effect showed that after the MPC system was put into use,it could stabilize the outlet temperature of the heating furnace,optimize the temperature distribution in the tower and stably control the liquid level of key equipment,reduce the standard deviation of key controlled process parameters by more than 35%,and the output and quality of each basic oil had been effectively improved.