基于实时反馈强化学习神经网络的船舶艏摇智能控制研究  

Research on Intelligent Control of Ship Yaw Based on Real-time Feedback Reinforcement Learning Neural Network

作　　者：宋伟伟 徐跃宾 段学静 巩方超 崔英明 SONG Weiwei;XU Yuebin;DUAN Xuejing;GONG Fangchao;CUI Yingming(Research Center of Shandong Province Ship Control Engineering and Intelligent System Engineering and Technology,Weihai 264300,China;Weihai Ocean Vocational College,Weihai 264300,China)

机构地区：[1]山东省船舶控制工程和智能系统工程技术研究中心,山东威海264300 [2]威海海洋职业学院,山东威海264300

出　　处：《现代信息科技》2024年第8期83-88,共6页Modern Information Technology

基　　金：山东省船舶控制工程与智能系统工程技术研究中心科研专项(SSCC-2021-0006)。

摘　　要：文章提出了一种基于实时反馈强化学习神经网络控制的船舶艏摇智能控制方法。该方法将神经网络的非线性建模和强化学习的自适应控制技术相结合,能够实现对船舶航行过程中舵角的精确控制。并将PID控制算法、模型预测控制算法和实时反馈强化学习神经网络控制算法进行对比分析,仿真实验结果表明,后者在控制效果和稳定性方面均优于前两种方法,能够有效地提高船舶航行过程中舵角的控制精度和鲁棒性。This paper proposes an intelligent control method for ship yaw based on real-time feedback reinforcement learning neural network control.This method combines nonlinear modeling of neural networks with adaptive control technology of reinforcement learning to achieve precise control of rudder angle during ship navigation.And the PID control algorithm,model prediction control algorithm,and real-time feedback reinforcement learning neural network control algorithm are compared and analyzed.The simulation experiment results show that the latter is superior to the previous two methods in control effectiveness and stability,and could effectively improve the control accuracy and robustness of the rudder angle during ship navigation.

关 键 词：实时反馈 强化学习 神经网络 船舶艏摇 

分 类 号：TP39[自动化与计算机技术—计算机应用技术] TP18[自动化与计算机技术—计算机科学与技术] U664.82[交通运输工程—船舶及航道工程]