基于改进NSGA-Ⅱ的涤纶长丝熔体输送过程工艺优化  

Multi-objective optimization of polymer melt convey process in polyester fiber production using improved NSGA-Ⅱ

作　　者：徐楠[1] 丁永生[1,2] 郝矿荣[1,2] 

机构地区：[1]东华大学信息科学与技术学院,上海201620 [2]数字化纺织服装技术教育部工程研究中心,上海201620

出　　处：《计算机与应用化学》2012年第7期825-828,共4页Computers and Applied Chemistry

基　　金：国家自然科学基金重点项目(61134009);国家自然科学基金(60975059);教育部高等学校博士学科点专项科研基金(20090075110002);上海市优秀学术带头人计划项目(11XDl400100);上海领军人才专项资金;上海市科学技术委员会重点基础研究项目(11JCl400200;10JCl400200)

摘　　要：涤纶纺丝生产过程中熔体输送环节具有机理复杂、受诸多因素影响、拥有多种产品性能指标等特点,对其进行工艺优化较为困难,目前往往凭借生产经验,缺乏一定的理论指导。熔体输送环节工艺优化是1个多目标优化问题,为此提出了1种智能多目标工艺优化方法:。该方法:采用以Pareto多目标最优理论为优化导向的带精英策略非支配排序遗传算法(NSGA-Ⅱ),并对其进行了一定的改进。改进之处在于,传统的NSGA-Ⅱ算法选择用于计算个体拥挤度的参考点是按各个目标分别进行选择,本文采用了多个目标综合选择法,使拥挤度的计算与筛选更为准确。本文以熔体出口处压强、温度、特性粘度3个指标为优化目标,依据工业现场数据,进行了仿真实验。实验结果:表明,该方法:运算速度较快,优化结果:准确,分布均匀,能够对实际生产的工艺优化起到一定的指导作用。The polymer melt convey process of polyester fiber production was complicated and dominated by lots of factors. Besides, its optimization object was not unique. So it was hard to optimize the polymer melt convey process. Current ways of the polymer melt convey process optimization was based on produce experience so that it was lack of proper theoretical guiding. Due to these features, an intelligent multi-objective optimization method was presented to solve such multi-objective problem. This method is guided by Pareto-optimal theory, which adopts and modifies the NAGS-Ⅱ method. To improve the NSGA-Ⅱ method, a new way, which combined all the objects to select the reference points of crowding distance, was adopted instead of the previous way, which select the reference points separately by each object. This method produced a better result of crowding distance computing. Using industrial measured data, we conduct a simulation experiment which using Pressure, Temperature and Ⅳ as the optimal objects. Simulation results demonstrate that the proposed method can obtain optimal results rapidly and precisely, and the result is comprehensive and well-distributed. The method is of guiding value to actual production process.

关 键 词：NSGA-Ⅱ PARETO最优解 多目标优化 熔体输送 

分 类 号：TQ015.9[化学工程] TP391.9[自动化与计算机技术—计算机应用技术]