Communication-Aware Formation Control of AUVs With Model Uncertainty and Fading Channel via Integral Reinforcement Learning  

作　　者：Wenqiang Cao Jing Yan Xian Yang Xiaoyuan Luo Xinping Guan 

机构地区：[1]the Institute of Electrical Engineering,Yanshan University,Qinhuangdao 066004,China [2]IEEE [3]the Institute of Information Science and Engineering,Yanshan University,Qinhuangdao 066004,China [4]the Department of Automation,Shanghai Jiao Tong University,Shanghai 200240,China

出　　处：《IEEE/CAA Journal of Automatica Sinica》2023年第1期159-176,共18页自动化学报（英文版）

基　　金：supported in part by the National Natural Science Foundation of China(62222314,61973263,61873345,62033011);the Youth Talent Program of Hebei(BJ2020031);the Distinguished Young Foundation of Hebei Province(F2022203001);the Central Guidance Local Foundation of Hebei Province(226Z3201G);the Three-Three-Three Foundation of Hebei Province(C20221019)。

摘　　要：Most formation approaches of autonomous underwater vehicles(AUVs)focus on the control techniques,ignoring the influence of underwater channel.This paper is concerned with a communication-aware formation issue for AUVs,subject to model uncertainty and fading channel.An integral reinforcement learning(IRL)based estimator is designed to calculate the probabilistic channel parameters,wherein the multivariate probabilistic collocation method with orthogonal fractional factorial design(M-PCM-OFFD)is employed to evaluate the uncertain channel measurements.With the estimated signal-to-noise ratio(SNR),we employ the IRL and M-PCM-OFFD to develop a saturated formation controller for AUVs,dealing with uncertain dynamics and current parameters.For the proposed formation approach,an integrated optimization solution is presented to make a balance between formation stability and communication efficiency.Main innovations lie in three aspects:1)Construct an integrated communication and control optimization framework;2)Design an IRL-based channel prediction estimator;3)Develop an IRL-based formation controller with M-PCM-OFFD.Finally,simulation results show that the formation approach can avoid local optimum estimation,improve the channel efficiency,and relax the dependence of AUV model parameters.

关 键 词：Autonomous underwater vehicles(AUVs) communication-aware formation reinforcement learning uncertainty 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]