文献类型：期刊文献

英文题名：Ship motion control using an enhanced predictive PID under model parameter perturbations

作者：Xu, Dongxing[1,2,3];Yin, Jianchuan[1,2,3]

机构：[1]Guangdong Ocean Univ, Naval Architecture & Shipping Coll, Zhanjiang 524005, Peoples R China;[2]Guangdong Prov Engn Res Ctr Ship Intelligence & Sa, Zhanjiang 524005, Peoples R China;[3]Guangdong Prov Key Lab Intelligent Equipment South, Zhanjiang 524088, Peoples R China

年份：2025

外文期刊名：SHIPS AND OFFSHORE STRUCTURES

收录：SCI-EXPANDED(收录号：WOS:001636336900001)、、Scopus(收录号：2-s2.0-105024716149)、WOS

基金：This work was supported by National Natural Science Foundation of China [grant number 52271361 and 52231014]; Natural Science Foundation of Guangdong Province of China [grant number 2023A1515010684]; Start-up Funds of Guangdong Ocean University [grant number 060302132105]; Special Projects of Key Areas for Colleges and Universities of Guangdong Province [grant number 2021ZDZX1008].

语种：英文

外文关键词：Ship motion control; online model identification; sliding data window; stochastic trainer; feedforward neural network; predictive PID

外文摘要：To address the problem of controller mismatch caused by ship maneuvering motion model parameter perturbations, an enhanced predictive PID controller for ship motion is proposed by using a sliding data window (SDW), a novel stochastic optimizer, and an improved feedforward neural network. Firstly, The SDW is utilized as a local observer to detect ship motion states changes. Then, an improved black widow stochastic trainer-based feedforward neural network is employed as a multi-step predictor to characterize the ship local maneuvering behavior through learning a mini-batch of data samples from the SDW in the predictive PID control system. Finally, the improved black widow optimizer performs online tuning of the PID controller parameters to obtain optimal control input. The effectiveness and feasibility of the proposed method are verified by the online model identification and the course tracking experiment of the 'Mariner' vessel under model parameter perturbations.