Abstract: Aiming at the vehicle routing problem with fuzzy demands, the probability of vehicle service failure was analyzed based on fuzzy credibility theory, and a fuzzy expected value model was formulated. A parallel particle swarm optimization with double layers tabu search was proposed for routes optimization. In this algorithm, two different types of tabu search space were inserted, and two neighborhood arithmetic operators were designed by using new iterative formula and roulette strategy. The effectiveness of proposed method was verified by simulation contrast tests.

Key words: vehicle routing problem, fuzzy demands, fuzzy credibility, particle swarm optimization algorithm, tabu search, roulette strategy

摘要： 针对模糊需求的车辆路径问题，基于模糊可信性理论对车辆服务失败事件进行可能性分析，建立了一个模糊期望值模型,提出一种带双层禁忌搜索的并行粒子群算法。该算法引入两种不同的禁忌空间，采用新的粒子迭代公式，并利用轮盘赌策略设计了两类邻域算子。通过仿真对比实验表明了该算法的有效性。

关键词: 车辆路径问题, 模糊需求, 模糊可信性, 粒子群优化算法, 禁忌搜索, 轮盘赌策略