摘要
基于适应度函数相对误差和若干参数的选择 ,提出了一种交互式的进化规划·这种交互方式利于群体初始解和变异 ,选择操作中方差参数的选取·将此算法应用在一个供应链优化问题中并进行问题的仿真应用·结果表明对于数据量庞大的非线性模型的优化求解 。
Evolutionary programming is an intelligent algorithm for optimal search. It is mainly appropriate of none line model with a great volume of data. For facilitating the selection of parameters during computing,an interactive evolutionary programming is proposed based on the relativistic error and selection of some other parameters. The programming shows an advantage for initial value determination,mutation and variance parameter selective operation. This algorithm is applied to simulation optimization of supply chains. The result shows that improved evolutionary programming is much more appropriate of none line model with a great volume of data.
出处
《东北大学学报(自然科学版)》
EI
CAS
CSCD
北大核心
2000年第5期569-572,共4页
Journal of Northeastern University(Natural Science)
基金
辽宁省自然科学基金资助项目! ( 9910 2 0 0 2 0 8)
关键词
交互式
进化规划
供应链
优化
仿真
非线性模型
evolutionary programming
supply chain
interactive
optimization
simulation
manufacture
distribution