摘要
对于一个固定拓扑结构的供应链,基于设施能力和顾客需求的约束,以正向物流与逆向物流的总成本最小为目标,构造供应链网络设计模型,采用基于优先权编码的遗传算法求解。按照遗传算法的原理,将供应链按物流活动过程划分成4个阶段,运用基于优先权的编码方法对各个阶段进行编码;为加速算法收敛,利用贪婪启发式算法确定初始种群,通过遗传操作产生后代,并定义交叉算子和变异算子。以某一区域供应链网络规划为例,运用给出的模型和算法对需要设置工厂、铁路货运中转站的数量、位置及功能进行计算,并与其他算法所用计算时间比较。结果表明,采用本算法能够得到区域供应链网络规划的最优解,且所用的计算时间最少。
For a supply chain with fixed topology, a supply chain network design model is established based on the restraints of facilities capacity and customer demand with the minimum total cost of the forward logistics and reverse logistics as the target. Genetic algorithm of priority based encoding is adopted for solutions. According to the principle of genetic algorithm, the logistics activity process in the supply chain is divided into four stages, and each stage is encoded by priority based encoding method. To accelerate the algorithm convergence, the initial population is determined by greedy heuristic algorithm and the offspring is generated through genetic operation. Crossover operator and mutation operator are also defined. Taking the a regional supply chain network planning as an example, the proposed models and algorithms are used to determine the quantity, location, and function of the factories and railway freight distribution centers. The computation time of the proposed models and algorithms is compared with that of other algorithms. The result shows that the optimal solution of the regional supply chain network planning can be worked out by the proposed models and algorithms, and the computation time is the least.
出处
《中国铁道科学》
EI
CAS
CSCD
北大核心
2008年第6期116-120,共5页
China Railway Science
关键词
逆向物流
供应链网络设计
遗传算法
优先权编码
Reverse logistics
Supply chain network design
Genetic algorithm
Priority based encoding
