基于广义既约梯度法的水下机器人推力分配  

Underwater vehicle thrust allocation based on generalized reduced gradient method

作　　者：傅培成 吕海宁 杨海鹏 FU Peicheng;LV Haining;YANG Haipeng(School of Ocean and Civil Engineering,Shanghai Jiaotong University,Shanghai 200240,China;The 710 Research Institute of CSSC,Yichang 443003,China)

机构地区：[1]上海交通大学船舶海洋与建筑工程学院,上海200240 [2]中国船舶集团有限公司第七一○研究所,湖北宜昌443003

出　　处：《舰船科学技术》2025年第5期103-111,共9页Ship Science and Technology

摘　　要：为提高自主水下机器人动力控位系统的控制精度,延长水下作业时间,提出一种与灰狼优化算法相结合的广义既约梯度法对自主水下机器人进行推力分配。该方法通过计算目标函数的既约梯度获取其下降方向,通过灰狼优化算法在下降方向上进行一维搜索获取符合约束条件的更优解,经过多次迭代获取到目标函数的全局最优解。通过仿真模拟对算法性能进行测试。仿真结果表明,相比于传统推力分配算法,广义既约梯度法能更有效地降低能耗、减少推力饱和问题。To improve the control accuracy of the power control system of AUV(Autonomous Underwater Vehicle)and extend the underwater operation time,a GRG(Generalized Reduced Gradient)method combined with grey wolf optim-ization algorithm is proposed for thrust allocation of AUV.This method obtains the descent direction of the objective func-tion by calculating its reduced gradient,and uses the grey wolf optimization algorithm to conduct one-dimensional search in the descent direction to obtain a more optimal solution that meets the constraint conditions.After multiple iterations,the global optimal solution of the objective function is obtained.This article tests the performance of the algorithm through simu-lation.The simulation results show that compared to traditional thrust allocation algorithms,the generalized reduced gradi-ent method can more effectively reduce energy consumption and thrust saturation problems.

关 键 词：自主水下机器人 广义既约梯度法 推力分配 

分 类 号：TP242.6[自动化与计算机技术—检测技术与自动化装置]