混合自适应遗传算法优化多目标染色配方模型  

Optimal design of multi-objective dyeing process model by hybrid self-adaptive taguchi-genetic algorithm

作　　者：汪岚[1] 金福江[2] 谢振辉 

机构地区：[1]黎明职业大学机电工程系,福建泉州362000 [2]华侨大学信息科学与工程学院,福建厦门361021 [3]建国科技大学自动化工程系暨机电光系统研究所,台湾地区彰化50094

出　　处：《计算机与应用化学》2012年第7期873-876,共4页Computers and Applied Chemistry

基　　金：国家自然科学基金(601143005);福建省产学研重大项目(2011H6019);黎明职业大学校级课题(LZ2011101)

摘　　要：从提高染色产品质量和效益的角度出发,综合考虑如染料浓度、温度、时间和助剂浓度等因素影响,构建了多目标染色工艺配方优化模型。针对传统遗传算法普遍存在的问题和缺陷,提出基于正交试验设计、自适应交叉操作及局部搜索等技术进行算法改进,并利用改进后的算法获得配方模型最优解的解决方法:。实践结果:证明,混合自适应遗传算法使种群更具有代表性和全面性,最大程度的继承了父代的优良特性,改善了算法的早熟现象并增强其寻优性能。最终以较少的计算量和较高的收敛速度对全局进行快速的搜索,比传统遗传算法得到的优化目标值降低了l0.8%左右。该方法:可推广应用于其他复杂过程的优化求解问题中。To improve quality and benefit for dyeing production, a multi-objective optimization for dyeing process model is developed under comprehensive consideration of factors of concentration of dye-stuff, time, temperature and concentration of textile assistant and so on. For the problems and defects of traditional genetic algorithm, key technologies such as Taguchi design, self-adaptive crossover operator and local search are applied, a novel determination method based on hybrid algorithm is proposed and the optimal model is calculated. The real experimental results show that Hybrid Self-Adaptive Taguchi-Genetic Algorithm （HSTGA） have advantages such as population diversity and comprehensiveness are maintained, good properties of parents are inherited, the phenomenon of algorithm premature is improved and optimization performance is enhanced. Finally, the global search with less computation and high convergent speed, and the function value is reduced 10.8%. This approach can be popularized to solve other complex process optimization problems.

关 键 词：多目标染色工艺配方优化模型 正交试验设计 自适应交叉操作 局部搜索 混合自适应遗传算法 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]