Intelligent prediction of mass transfer during natural gas decarburization in high gravity reactor
The high-gravity reactor has excellent mass transfer performance and has broad application prospects in the field of natural gas decarbonization.In order to effectively predict the natural gas decarburization performance of the high-gravity reactor,a set of natural gas decarburization experimental system based on high-gravity reactor was built,and the influences of operating parameters(high gravity factor,spray density of absorbent,mixture volume flow rate and intake composition,etc)on the CO2 removal performance were investigated.Then a dimensionless method was adopted to establish the mapping relationship between operating parameters and mass transfer coefficient.Finally,based on the least square support vector machine(LSSVM)algorithm,the intelligent prediction model was established.The results show that both the high gravity factor and absorbent spray density have optimal values,which are 57.62 and 2.04 m3/(m2·h),respectively.In addition,it is found that the CO2 removal effect decreases while the mass transfer coefficient increases with the increases of mixture volume flow rate.Under the condition of optimal model parameter([γ,t,d]=[13557.2021,9.5876,4]),the determination coefficient(R2)and average relative error(MRE)of prediction results of intelligent prediction model are 0.9519 and 0.0949,respectively,and the relative error of prediction results is within±20%.It demonstrates that the intelligent prediction model has high accuracy.
high gravity reactornatural gas decarburizationLSSVM algorithmmass transferintelligent prediction model
国家科技期刊平台
NSTL
万方数据
