基于PSO-LSTM模型的舰船轨迹预测研究  

Research into Ship Trajectory Prediction Based on PSO-LSTM Model

作　　者：陈磊 李安然 罗寿超 CHEN Lei;LI Anran;LUO Shouchao(The Eighth Research Institute of CSSC,Yangzhou 225101,China)

机构地区：[1]中国船舶集团有限公司第八研究院,江苏扬州225101

出　　处：《舰船电子对抗》2025年第2期42-45,共4页Shipboard Electronic Countermeasure

摘　　要：船舶轨迹预测研究能够帮助提高海上交通效率,保障交通安全。利用长短时记忆(LSTM)模型擅长处理长序列时序数据的特性,采用改进的粒子群优化(PSO)算法,基于PSO-LSTM模型对舰船轨迹进行预测。使用船舶自动识别系统(AIS)数据作为实验数据,在完成实验数据的预处理之后,通过对历史轨迹数据进行分析,预测未来舰船的航行路径。对比分析了LSTM模型和反向传播(BP)神经网络模型之后,PSO-LSTM模型的精度最高,其均方根误差和平均绝对百分比误差均为最小,表明PSO-LSTM模型的精度较高,具有一定的应用价值,能够为舰船交通流管理和航行安全提供一定的决策支持。The research into ship trajectory prediction can help to improve the maritime traffic efficiency and ensure the traffic safety.Utilizing the long short-term memory(LSTM)model’s expertise in handling long sequence time series data,this paper predicts the ship trajectories based on particle swarm optimization(PSO)-LSTM model by using an improved PSO algorithm,and using automatic identification system(AIS)data as experimental data,after completing the preprocessing of the experimental data,analyzed the historical trajectory data to predict the future navigation path of the ship.The LSTM model and back propagation(BP)neural network model are compared and analyzed,and the PSO-LSTM model has the highest accuracy,with the smallest root mean square error and average absolute percentage error,which shows that the PSO-LSTM model has high accuracy and a certain application value,and can provide a certain decision support for ship traffic flow management and navigation safety.

关 键 词：舰船轨迹预测 长短时记忆网络 粒子群优化算法 自动识别系统 

分 类 号：U675.7[交通运输工程—船舶及航道工程]