【摘要】 集装箱船在泊时间是制定泊位计划的重要依据。针对在泊时间预测面向不确定性环境的特性，考虑适用模型预测性能，提出一种基于主成分分析(PCA)、非线性参数动态控制的改进粒子群算法(IPSO)和极限学习机(ELM)的在泊时间集成预测模型，以天津港集装箱船港口作业数据为预测样本进行了实证研究。对比仿真结果表明：与已有集装箱船在泊时间预测模型相比，PCA-IPSO-ELM集成预测模型具有更高的精准度和时效性，其中，改进后的粒子群算法也体现出了较强的全局探索能力及较好的稳定性。该模型可为制定泊位计划提供有力的数据支持，有助于港口制定科学高效的泊位计划。更多还原

【Abstract】 The berthing time of container ships is an important basis for making berth plans. Aiming at the characteristics of berthing time prediction facing uncertain environment and considering the prediction performance of the applicable model, an integrated berthing time prediction model based on principal component analysis(PCA), nonlinear parameter dynamic control improved particle swarm optimization(IPSO) and extreme learning machine(ELM) was proposed. Based on the container ship operation data of Tianjin port, an empirical study was carried out. The comparison of simulation results shows that compared with the existing container ship berthing time prediction model, the PCA-IPSO-ELM integrated prediction model has higher accuracy and timeliness, and the improved particle swarm optimization algorithm also reflects strong global exploration ability and good stability. This model can provide powerful data support for making berth plan, and help the port to make scientific and efficient berth plans.更多还原