首页|Cascaded model predictive controller performance for the selection of robust working fluids in absorption refrigeration cycles
Cascaded model predictive controller performance for the selection of robust working fluids in absorption refrigeration cycles
扫码查看
点击上方二维码区域，可以放大扫码查看
原文链接
NSTL
Elsevier
Current working fluid selection approaches for the absorption refrigeration cycle are based exclusively on steady-state cycle performance. In real operating conditions, the cycle is subject to exogenous disturbances, with detrimental effects on their performance, resulting in increased resource utilization or even failure to meet load demands. In the present work, the problem of working fluid evaluation for the absorption refrigeration cycle is investigated, by considering both the steady state and the dynamic performance of the system. The evaluation of 13 novel, organic working fluids is performed utilizing a cascaded proportional-integral and a cascaded model predictive controller, implemented on a single-effect absorption refrigeration cycle. A set of previously validated non-linear, dynamic process models, developed in ASPEN Plus are used to create the linear models required for the model predictive controller. The closed loop dynamic performance is evaluated based on speed of response, resource utilization and deviations from the desired operating point (setpoint), while operating under a disturbance scenario involving load demand changes. Multi-criteria assessment results indicate that the cascaded model predictive control is considerably more consistent than the cascade PI controller. The novel mixture of acetaldehyde/dimethylformamide exhibits superior performance than the conventional NH3/H2O mixture, by 57% in speed of response and 76% in resource utilization. The mixture ιs also 25% and 12% better than NH3/H2O in steady-state cost per ton of cooling and coefficient of performance, indicating high economic potential and robustness for single-effect ABR systems.
NSTL
Elsevier
