Abstract: To solve the problem of inventory imbalance in the chain enterprises, this paper studies the multi-commodity vehicle routing problem with split pickupand delivery and fuzzy demand (MCVRPSPDFD). The question considers factors such as mixed loading of multiple goods, multiple visits, unmatched supply and demand, non-unique demand, and uncertain demand. This paper constructs the MCVRPSPDFD mathematical model which aims at minimizing operating costs. The model uses credibility measurement theory to deal with uncertain factors in the decision-making environment and solves it through an improved genetic tabu algorithm. To adapt to model requirements and improve computing efficiency, the algorithm designs special initial population, encoding and decoding methods and selects appropriate parameters through parameter testing. The results of the calculation example show that the paper can effectively solve the problem of chain enterprises’ inventory imbalance. The change in the preference value of decision-makers will have an impact on operating costs.

Key words: vehicle routing problem, pickupanddelivery, fuzzydemand, heuristic algorithm

摘要： 为解决连锁企业库存不平衡问题,本文研究了考虑多商品多批次取送货的模糊需求车辆路径问题。该问题综合考虑了多货混装、多次访问、供需未匹配、客户需求不唯一以及需求不确定等因素。本文以运营成本最小为目标,构建MCVRPSPDFD数学模型,模型利用可信测度理论应对决策环境中的不确定因素,通过改进的禁忌搜索算法进行求解。为适应模型需求和提升运算效率,算法设计了合理的初始种群形成过程及编码解码方式,并通过参数测试选取合适的参数。算例结果显示,本文成果能有效解决连锁企业库存不平衡问题,决策者偏好值的变动会对运营成本产生影响。

关键词: 车辆路径问题, 取送货问题, 模糊需求, 启发式算法