具有外源输入的船舶横摇运动NARX神经网络预测  

Improved prediction method for ship rolling motion via NARX neural network

作　　者：李冲 章文俊[1] 薛宗耀 张国庆[1] 赵常九 LI Chong;ZHANG Wen-jun;XUE Zong-yao;ZHANG Guo-qing;ZHAO Chang-jiu(Navigation College,Dalian Maritime University,Dalian 116026,China)

机构地区：[1]大连海事大学航海学院,辽宁大连116026

出　　处：《舰船科学技术》2022年第11期63-67,共5页Ship Science and Technology

基　　金：国家重点研发计划资助项目(2019YFB1600602);辽宁省重点研发计划资助项目(201801704);辽宁省自然基金资金项目(201801705)。

摘　　要：为了提高船舶在风浪中航行的安全性,需要精确地预测船舶在风浪中的横摇运动,以提高船舶横摇控制效果,本文通过应用一种带外源输入的非线性自回归(NARX)神经网络预测方法预测船舶横摇运动。该方法考虑了实船操纵性试验数据受风、浪、流等外界因素的的影响,将实测的风向、风速、流速、流向、浪向以及浪高的数据作为外源输入,能够有效提高船舶横摇运动的预测精度。基于“育鲲”轮,利用该方法对实际船舶海上横摇运动进行了实时预测实验,并将其实验结果与SAPSO-BP神经网络模型的预测结果进行对比。从对比结果可以看出,本文所提方法对复杂海浪环境具有良好的适应性,NARX模型的预测精度优于普通反向传播(BP)神经网络和自适应粒子群算法优化的普通反向传播(SAPSO-BP)神经网络.In order to accurately and efficiently predict the rolling motion of a ship in waves and ensure the safety of personnel,cargo and ship,a nonlinear autoRegressive xogenous(NARX)neural network prediction method based on external input is proposed in this paper.The influence of wind on the actual ship maneuverability test data is considered in this method,and the measured wind direction and wind speed data are taken as external input,which can effectively improve the performance of the ship.The prediction accuracy of ship rolling motion.Based on the scientific research practice ship Yukun of Dalian Maritime University,the real-time prediction experiment of ship rolling motion is carried out by using this method,and the experimental results are compared with the prediction results of simulated annealing particle swarm optimizationback propagation(SAPSO-BP)neural network model optimized by adaptive particle swarm optimization.From the comparison results,it can be seen that the proposed method has good adaptability to complex sea wave environment,and the prediction accuracy of NARX model is better than that of BP neural network and SAPSO-BP neural network optimized by adaptive particle swarm optimization.

关 键 词：船舶横摇运动 NARX神经网络 复杂海洋环境 模型预测 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]