摘要
针对订单选择问题,考虑订单具有不同制程的特征,建立了以总收益为目标的混合整数非线性规划模型。以量子粒子群优化算法为框架进行求解,采用基于排序的粒子编码方案表达0-1变量和整数变量,提出四种种群初始化策略以便提高求解质量,并在迭代过程中对不可行解进行修复。通过对比验证,结果表明模型和算法可行、有效。
Considering the characteristic of multiple process routes, this paper focuses on the order selection problem. A mixed integer non-linear programming model is built with objective of maximizing the total profits. A quantum-behaved PSO-based algorithm is employed as solution mainframe in which a sequence-based particle coding solution is designed to represent decision variables and integer variables and four population initiation strategies are proposed to improve the solution quality. A repair mechanism for infeasible solutions is adopted in iteration. By computational comparison tests, it can be proven that the model and algorithm are feasible and effective.
出处
《计算机工程与应用》
CSCD
北大核心
2015年第15期236-239,254,共5页
Computer Engineering and Applications
关键词
订单选择
多制程
量子粒子群
order selection
multiple process routes
Quantum-behaved Particle Swarm Optimization（QPSO）
