Although vapor recompression heat pumps(VRHP)can enhance the steady-state performance of a di-viding-wall distillation column(DWDC),they also exacerbate the coupling between the controlled variables,mak-ing the smooth operation of a vapor-recompressed dividing-wall distillation column(VRHP-DWDC)a challenging task.In this work,a novel single temperature decentralized control system based on a non-square relative gain ma-trix and in-depth analysis of closed-loop responses has been developed.The method has been evaluated using a VRHP-DWDC to separate an intermediate-component dominated benzene/tolulene/o-xylene ternary mixture.Since the top/side heat-pump compressors are used to control the sensitive stage temperatures of side/common stripping section,the transient responses of the side product can be accelerated.and the peak deviations of the bottom prod-uct can also be reduced.Since the high internal coupling and the distorted correspondence between temperature and composition are responsible for the large steady-state deviations in the side and bottom products,this work utilizes a double temperature difference structure to strengthen the control system further,thereby giving rise to an asymmetric temperature control system.Reducing the inherent non-linearities helps to reduce the steady-state deviations of side and bottom products.The closed-loop simulation results clearly demonstrate the superiority of the derived asymmet-ric temperature control system.This study highlights that although the introduction of the VRHP aggravates the in-ternal coupling of a VRHP-DWDC,it provides a variable which can potentially be manipulated,namely the com-pressor power.The better dynamic characteristics combined with the effective use of a double temperature difference structure can,to a certain extent,improve the closed-loop operation of a VRHP-DWDC.
关键词
隔离壁蒸馏塔/过程强化/蒸汽再压缩热泵/温度控制/温差控制
Key words
dividing-wall distillation column/process intensification/vapor recompression heat pumps/tempera-ture control/temperature difference control
维普
万方数据