基于递归神经网络的全海深载人潜水器控制分配研究  

Control Allocation of Full-Ocean-Depth Manned Submersible Using Recursive Neural Network

作　　者：邓超 赵洋[1,2,3] 赵兵[1,2,3] 陈帅华[1,2,3,4] DENG Chao;ZHAO Yang;ZHAO Bing;CHEN Shuai-hua(State Key Laboratory of Robotics,Shenyang Institute of Automation,Chinese Academy of Sciences,Shenyang 110016,China;Institutes for Robotics and Intelligent Manufacturing,Chinese Academy of Sciences,Shenyang 110169,China;Key Laboratory of Marine Robotics,Liaoning Province,Shenyang 110169,China;University of Chinese Academy of Sciences,Beijing 100049,China)

机构地区：[1]中国科学院沈阳自动化研究所机器人学国家重点实验室,沈阳110016 [2]中国科学院机器人与智能制造创新研究院,沈阳110169 [3]辽宁省水下机器人重点实验室,沈阳110169 [4]中国科学院大学,北京100049

出　　处：《自动化与仪表》2022年第6期22-27,48,共7页Automation & Instrumentation

基　　金：国家重点研发计划项目(2016YFC0300604)。

摘　　要：为了更好地控制全海深载人潜水器,对全海深载人潜水器“奋斗者”的控制分配进行研究,该文提出一种基于递归神经网络的控制分配方法。首先根据推进器的空间布置,建立六自由度的推力模型;其次设计代价函数,将控制分配问题转化为凸优化问题;最后设计递归神经网络求解优化问题,并采用FPGA实现该递归神经网络。仿真结果表明,该方法能有效地处理推进器约束且实时性较好。In order to control the full-ocean-depth manned submersible better,the control allocation method of full-ocean-depth manned submersible“Fendouzhe”is studied,and a control allocation method based on recursive neural network is proposed.At first,according to the space arrangement of the thruster,a six-degree-of-freedom thrust model is established.Then,the cost function is designed to transform the control allocation problem into a convex optimization problem.Finally,recursive neural network is designed to solve the optimization problem,and implemented with FPGA.Simulation results show that this method can effectively deal with propeller constraints and has good real-time performance.

关 键 词：载人潜水器 控制分配 递归神经网络 

分 类 号：TP24[自动化与计算机技术—检测技术与自动化装置]