一种基于遗传算法的无人潜水器推力分配优化方法  

An Optimization Method for Thrust Allocation of ROV Based on Genetic Algorithm

作　　者：徐蕾 宋俊辉 涂绍平 朱俊 李仁雄 XU Lei;SONG Junhui;TU Shaoping;ZHU Jun;LI Renxiong(CRRC SMD(Shanghai)Ltd.,Shanghai 201306,China)

机构地区：[1]上海中车艾森迪海洋装备有限公司,上海201306

出　　处：《控制与信息技术》2023年第6期98-104,共7页CONTROL AND INFORMATION TECHNOLOGY

基　　金：国家重点研发计划项目(2022YFC2806901)。

摘　　要：为解决无人潜水器(remotely operated vehicle,ROV)推力分配的输出饱和问题,文章提出一种基于遗传算法的无人潜水器推力分配优化方法。首先以某型微小型电动ROV的推进器空间布局为基础建立了ROV水平面与垂直面的推力数学模型,然后应用遗传算法分别对ROV的水平面与垂直面的推进器的推力分配进行优化,并与伪逆法的推力分配进行对比。仿真分析表明,与伪逆法推力分配相比,采用遗传算法优化的推力分配,不仅可以避免推力输出饱和,而且各推进器输出推力的合力或力矩与期望的合力或力矩基本吻合,误差可以保证在±1%以内。To address the problem of output saturation in the thrust allocation of remotely operated vehicles(ROVs),this paper introduces a thrust distribution optimization technique based on genetic algorithm.Initially,a mathematical model describing the horizontal and vertical thrust of an ROV is established,using the spatial configuration of a miniaturized electric ROV's propellers as foundation.Subsequently,genetic algorithm is employed to optimize thrust allocation for the ROV's horizontal and vertical propellers,with results compared to outcomes from a pseudo-inverse method of thrust allocation.Simulation analysis reveals that the thrust allocation optimized via the genetic algorithm can effectively prevent output saturation and ensures that the resultant force or torque from each propeller's output aligns almost exactly with the intended combined force or torque,maintaining an error within±1%.

关 键 词：ROV 伪逆法 推力分配 遗传算法 仿真分析 

分 类 号：TP242.3[自动化与计算机技术—检测技术与自动化装置]