Prediction in Catalytic Cracking Process Based on Swarm Intelligence Algorithm Optimization of LSTM

Key Words:CARBON-DIOXIDE ABSORPTION; MOLAR HEAT-CAPACITIES; SOLUBILITY PARAMETERS; METHYLDIETHANOLAMINE; ALKANOLAMINES; PREDICTION; ABSORBENTS; SOLVENTS

Abstract:Amine solvent has attracted much attention due to its high CO2 capture level and wide application range, but its high energy consumption for recycling restricts its large-scale commercialization. In this work, a multi-objective optimization technology based on the group contribution method was used to select potential amine solvents for CO2 capture. This computer-aided molecular design method considers the thermodynamic and kinetic properties of the candidate solvent and evaluates the influence of relevant parameters on solvent performance. Compared with previous experimental methods used to optimize solvent, this method selects potential solvents from a large number of solvent databases based on group contribution. Firstly, a corresponding classification database was established for ious kinds of amine solvents. Then, the traditional experiments were used to verify and screen solvents. At the same time, the method was applied to 31 amine absorbents concerning solubility, molar volume, surface tension, heat capacity, viscosity, pKa, saturated vapor pressure, and so on, and seven solvents were found to have comparable performance to MEA, with higher absorption rates and solubility. This method provides guidance for screening CO2 capture absorbents with economic viability, high efficiency, fast absorption rates, and low regeneration energy consumption.

Volume:11

Issue:5

Translation or Not:no