面向高密度聚乙烯生产过程的最优采样策略  

Optimal sampling strategy for high-density polyethylene production processes

作　　者：尹鹏 陈伟锋[1] YIN Peng;CHEN Wei-feng(School of Information Engineering,Zhejiang University of Technology,Hangzhou 310023,China)

机构地区：[1]浙江工业大学信息工程学院,浙江杭州310023

出　　处：《高校化学工程学报》2022年第5期738-747,共10页Journal of Chemical Engineering of Chinese Universities

基　　金：国家重点研发计划(2017YFE0106700);国家自然科学基金(61873242)。

摘　　要：针对需要大量采样点保证参数估计的精确度使得样本分析成本过高的问题,提出一种基于拉丁超立方体采样的最优采样策略,既能减少采样点数量又能保证参数估计精度。最优采样策略通过拉丁超立方体采样参数空间候选值并求解相应的最优采样子问题,构建各潜在采样点的选择频率,使用启发式策略和迭代增强策略选择采样点,并实现参数估计。数值仿真结果表明,提出的采样策略在保证参数估计精度情况下,能够大幅度降低采样点数量,基于随机采样的最优采样方法相比,稳定性更好。A large number of sampling points are required to ensure the accuracy of parameter estimation,which leads to high cost of sample analysis. An optimal sampling strategy was thus put forward based on Latin hypercube sampling to reduce the number of sampling points and ensure parameter estimation accuracy. By solving corresponding optimal sampling sub-problems with the candidate values of parameter space by Latin hypercube sampling, the selection frequency of each potential sampling point was constructed in the optimal sampling strategy. Sampling points were selected by virtue of heuristic strategy and iterative enhancement strategy, and parameter estimation was achieved. The numerical simulation results show that the sampling strategy proposed can greatly reduce the number of sampling points while ensuring parameter estimation accuracy. In addition, it is more stable than the optimal sampling method based on random sampling.

关 键 词：高密度聚乙烯 参数估计 最优采样 参数子集选择 拉丁超立方体采样 

分 类 号：TQ311[化学工程—高聚物工业]