时间依赖型多时间窗车辆路径问题研究

针对一类考虑城市交通拥堵情况的时间依赖型多时间窗车辆路径问题(time-dependent vehicle routing problem with multiple time windows，TD_VRPMTW)，提出一种混合离散灰狼算法(hybrid discrete grey wolf optimizer，HDGWO)进行求解。在HDGWO中，设计了新的灰狼个体更新公式，采用基于客户排列的整数编码方式，使算法可直接在离散问题解空间中执行基于标准灰狼算法个体更新机理的全局搜索；设计了基于问题性质的种群初始化策略，用于生成具有高质量和多样性的初始种群；引入头狼信息交流公式，用于探索头狼形成的优质解空间；构造具有多种局部搜索操作的自适应变邻域局部搜索策略，用于增强算法的局部搜索能力。结果表明：HDGWO可有效求解TD_VRPMTW。

Research on Time-dependent Vehicle Routing Problem with Multiple Time Windows

Aiming at the time-dependent vehicle routing problem with multiple time windows (TD_VRPMTW) that considers urban traffic congestion, a hybrid discrete gray wolf optimizer (HDGWO) is proposed. In the HDGWO, a new grey wolf individual updating formula is designed, and the integer coding method based on customer permutation is adopted, so that the algorithm can directly perform the global search based on GWO individual updating mechanism in the discrete problem solution space.A population initialization strategy based on the nature of the problem is designed to generate the initial population with high quality and diversity.The information exchange formula of the head wolf is introduced to explore the high-quality solution space formed by the head wolf. An adaptive variable neighborhood local search strategy with multiple local search operators is constructed to enhance the local search ability of the algorithm. Results show that HDGWO can effectively solve TD_VRPMTW.