Abstract: To enhance the operational efficiency of container terminals, the continuous berth allocation problem considering tidal factors and berth preferences was studied. A mixed-integer linear programming model was established with the objectives of minimizing ship waiting, delayed departure, berth deviation, as well as minimizing fuel consumption costs during the port period, by introducing non-overlapping ship spatiotemporal rectangles and tidal time window constraints. According to model features, CPLEX solving software, fish swarm algorithm, cuckoo search algorithm, and the hybrid cuckoo search-fish swarm algorithm were used for solution. The arrival data of 20 ships of different sizes with a planning cycle of 36 hours was utilized as a case study. A berth allocation plan that met the requirements of tidal time window and berth preference was obtained by solving the case study. The solution for the case study shows that for smaller instances, CPLEX can find the optimal berth allocation scheme in a relatively short time. However, for larger instances, CPLEX requires longer time for solution, while the cuckoo-fish swarm hybrid algorithm can obtain suboptimal solutions with a difference from 0.39% to 4.2% from CPLEX within an average of 3 minutes. Comparison between different algorithms demonstrates that the hybrid cuckoo search-fish swarm algorithm has better solving ability. In the berth allocation scheme obtained by the proposed hybrid algorithm, the entry and exit times of tidal vessels are all during high tide periods, and more than 85% of vessels receive loading and unloading services within 200 meters of their preferred berthing points.

Key words: port and waterway engineering; cuckoo-fish swarm hybrid algorithm; continuous berth allocation; mixed integer linear programming model; tidal factors; berth preferences

摘要： 为提升集装箱港口运营效率，笔者研究了考虑潮汐因素与泊位偏好的连续型泊位分配问题。引入了船舶时空矩形不可重叠约束和潮汐时间窗约束，构建以最小化船舶等待、延迟离港、泊位偏离以及在港期间油耗费用和最小为目标的混合整数线性规划模型；根据模型特征，采用CPLEX求解软件、鱼群算法、布谷鸟搜索算法和布谷鸟鱼群混合算法进行求解，以计划周期为36 h的20个不同规模的船舶到港数据为研究算例，通过算例求解得到符和潮汐时间窗、泊位偏好等要求的泊位分配方案。算例求解表明：算例规模较小时，CPLEX可以在较短时间内求出最优泊位分配方案；算例规模较大时，CPLEX求解时间较长，布谷鸟鱼群混合算法可以在平均3 min内求出与CPLEX差距为0.39%~4.20%的次优解；不同算法间的对比表明，布谷鸟鱼群混合算法求解能力更优，混合算法所得泊位分配方案中，乘潮船舶的进出港时刻均在潮汐高水位时段，且85%以上的船舶在偏好泊靠点200 m内接受装卸服务。

关键词: 港口与航道工程；布谷鸟鱼群混合算法；连续型泊位分配；混合整数线性规划模型；潮汐因素；泊位偏好