摘要
提出了一种适宜于液-固挤压复合材料工艺系统的多目标优化模型.通过合理决策过程,将多目标转化为单目标优化进行求解.在对编码规则、初始种群的产生、目标函数向适配值的转换以及种群数量、交叉概率、变异概率选择等问题进行深入研究的基础上,建立了基于遗传算法的液-固挤压复合材料优化系统.理论及实验研究表明可以获得较好的优化增益.采用优化设计参数进行实验,成形出质量良好的复合材料制件,且实现了变形力小、浸渗时间较短和浇注温度较低的预期目标,为该工艺的实际应用奠定了基础.
A multi-objective optimization model which is suitable for the technical system of liquid-solid extruding composites was proposed. In order to find a solution the multi-objective problem is converted into a single objective problem through a reasonable decision process. The profound investigation is conducted about the encode rules, the generation of initial popularity, the conversion from the object function to fitness value and the selection of cross or mutation probability. On this basis, the optimization system is established for liquid-solid extrusion to prepare composites. The theoretical and experimental researches show that the optimized gain can be achieved and the high quality products can be produced with the optimized parameters. The research realized the expected goal of low forming force and founded the base for the practical application of composite preparation.
出处
《金属学报》
SCIE
EI
CAS
CSCD
北大核心
2005年第10期1025-1030,共6页
Acta Metallurgica Sinica
基金
国家自然科学基金 50175091
国防预研基金 51412010304HK0339
西北工业大学博士论文创新基金资助
关键词
遗传算法
优化
复合材料
半固态
genetic algorithm, optimization, composite, semi-solid state