PDF Publication Title:
Text from PDF Page: 012
Processes 2022, 10, x FOR PEER REVIEW 12 of 19 Processes 2022, 10, 812 12 of 19 changed and retrained. Lee et al. [60] trained and tested a dynamic-model ANN for H2 recovery and CO2 capture from the tail gas of hydrogen plants, and they found that the dynamic-model-based ANN could precisely predict the dynamic behavior and optimum performance of an integrated process at a low computational cost. Figure4..SchematicdiiagramofamullttiillayerffeedfforwarrdnettworrkttrraiinedussiingttheeBPaallgoorriitthm.. In addition to ANN, applications of other surrogate models, such as response surface In addition to ANN, applications of other surrogate models, such as response surface models, Kriging models, polynomial models, orthogonal configuration models, support models, Kriging models, polynomial models, orthogonal configuration models, support vector machines, and gaussian processes regression, have been reported as well. Karson [68] vector machines, and gaussian processes regression, have been reported as well. Karson proposed and tested a model-reduction-based approach that systematically generates low- [68] proposed and tested a model-reduction-based approach that systematically generates order representations of rigorous PSA models. M. Moustapha [69] performed a detailed low-order representations of rigorous PSA models. M. Moustapha [69] performed a de- comparative study of Kriging and support vector regression (SVR). Both models are consid- tailed comparative study of Kriging and support vector regression (SVR). Both models are ered with isotropic or anisotropic kernels, but the Kriging model had a higher accuracy than considered with isotropic or anisotropic kernels, but the Kriging model had a higher ac- SVR and fewer failed cases. Considering real configurations instead of the experimental curacy than SVR and fewer failed cases. Considering real configurations instead of the design, anisotropic L2-SVR with Matérn seemed to be consistently the best model. Alison experimental design, anisotropic L2-SVR with Matérn seemed to be consistently the best Cozad and Nikolaos V. Sahinidis [70] designed a software package called automated learn- model. Alison Cozad and Nikolaos V. Sahinidis [70] designed a software package called ing of algebraic models for optimization (ALAMO) to automate the proposed methodology, automated learning of algebraic models for optimization (ALAMO) to automate the pro- which was able to identify accurate, low-complexity algebraic models that approximate a posed methodology, which was able to identify accurate, low-complexity algebraic mod- variety of high-fidelity systems. Zhang et al. [71] implemented the Box–Behnken design els that approximate a variety of high-fidelity systems. Zhang et al. [71] implemented the methodology (BBD), a type of response surface method, on the operating-parameter opti- Box–Behnken design methodology (BBD), a type of response surface method, on the op- mization of hydrogen purification. The flow chart of the BBD method is shown in Figure 5. erating-parameter optimization of hydrogen purification. The flow chart of the BBD They put the data from Aspen simulated results into Design ExpertTM solved through the method is shown in Figure 5. They put the data from Aspen simulated results into Design BBD method to obtain optimized operating conditions and then validated the predictions ExpertTM solved through the BBD method to obtain optimized operating conditions and by comparison with the Aspen recalculated results. In the BBD method, desirability is an then validated the predictions by comparison with the Aspen recalculated results. In the objective function that ranges from zero (outside of the limits of the performance objectives) BBD method, desirability is an objective function that ranges from zero (outside of the to one (where the performance objectives are met). Shen et al. [72] applied the central com- limits of the performance objectives) to one (where the performance objectives are met). posite design methodology to design experiments of the VPSA process, which explained Shen et al. [72] applied the central composite design methodology to design experiments the effects of the P/F ratio, adsorption time and desorption pressure on product purity, recovery and process energy consumption.PDF Image | Numerical Research on the Pressure Swing Adsorption Process
PDF Search Title:
Numerical Research on the Pressure Swing Adsorption ProcessOriginal File Name Searched:
processes-10-00812-v2.pdfDIY PDF Search: Google It | Yahoo | Bing
CO2 Organic Rankine Cycle Experimenter Platform The supercritical CO2 phase change system is both a heat pump and organic rankine cycle which can be used for those purposes and as a supercritical extractor for advanced subcritical and supercritical extraction technology. Uses include producing nanoparticles, precious metal CO2 extraction, lithium battery recycling, and other applications... More Info
Heat Pumps CO2 ORC Heat Pump System Platform More Info
| CONTACT TEL: 608-238-6001 Email: greg@infinityturbine.com | RSS | AMP |