摘要
针对间歇过程所具有的非线性特性,提出了一种基于核偏最小二乘(KPLS)模型的最终产品质量控制策略。利用初始条件、批次展开后的过程数据以及最终产品质量建立了间歇过程的KPLS模型;采用基于主成分分析(PCA)映射的预估方法对未知的过程数据进行补充,实现了最终产品质量的在线预测。为了解决最终产品质量的控制,利用T2统计量确定KPLS模型的适用范围,并作为约束引入产品质量控制问题,提高控制策略的可行性;采用粒子群优化(PSO)算法实现了优化问题的高效求解。仿真结果表明,与基于偏最小二乘(PLS)模型的控制策略相比,所提出的方法具有更高的预测精度,且能有效解决产品质量控制中出现的各种问题。
To deal with the nonlinear characteristics of batch processes,a final product quality control strategy based on kernel partial least squares(KPLS)was proposed.KPLS model of batch process was calibrated with initial conditions,batch-wise unfolding of process data and final product quality.An estimate method based on principal component analysis(PCA)mapping was used to supplement the unknown process data,and online prediction of product quality was achieved.To cope with the final product quality control,T2 statistic was used to determine the field that KPLS model was applicable,and the model was introduced into product quality control problem as constraint to enhance the feasibility of the control strategy.Particle swarm optimization(PSO)algorithm was used to solve the optimization problem efficiently.Simulation results showed that comparing with partial least squares(PLS)model based control strategy,higher prediction accuracy was obtained with the proposed method,and various problems in product quality control could be effectively solved.
出处
《化工学报》
EI
CAS
CSCD
北大核心
2013年第4期1332-1339,共8页
CIESC Journal
基金
国家高技术研究发展计划项目(2011AA060204)
国家自然科学基金项目(61203103)
中央高校基本科研业务费项目(N110304006)~~
关键词
间歇过程
质量控制
核偏最小二乘
主成分分析
优化
batch process
quality control
kernel partial least squares(KPLS)
principal component analysis(PCA)
optimization