Optimization of sieving process parameters for paper-based reconstituted tobacco based on Kriging model
In order to fully consider the random variation of sieving process parameters,this paper constructs the reliability-based optimization model for reconstituted tobacco process parameters,which taking the percentage of fibers below 12 mesh as the optimization objective,water flow rate and pulp concentration as the design parameters,and the percentage of fibers over 50 mesh as the probabilistic constraint.Firstly,the current process parameters are used as inheritance points for Latin hypercube sampling and screening experiments.Based on this,an initial Krig-ing model is constructed to fit the implicit relationship between water flow velocity,smoke slurry solid content,and the proportion of fiber mass below 12 mesh and above 50 mesh.And local adaptive sampling(LAS)was carried out to update the Kriging model to ensure the approximate accuracy of key positions at optimization iteration.Finally,Monte Carlo simulation(MCS)-Sequential quadratic programming(SQP)was used to solve the optimiza-tion model of screening process parameters.The results indicated that the optimized water flow rate and pulp con-centration were 6.31 L/min and 0.53%,respectively.Under this process parameter,the proportion of fiber below 12 mesh was 13.72%,which was 7.19%higher than before optimization.The proportion of fibers over 50 mesh was 17.65%,which was a decrease of 4.13%compared to before optimization.This method could meet the optimi-zation requirements of screening process parameters for paper making reconstituted tobacco leaves.
paper-based reconstituted tobaccoKriging modelinheriting latin hypercube samplingprobabilistic constraintlocal adaptive sampling
国家科技期刊平台
NSTL
万方数据
