【作者】 吴波；

【导师】 王广欣；

【作者基本信息】 大连交通大学 ， 机械工程， 2025， 硕士

【摘要】 水下捕捞机器人的研发在海生物捕捞领域具有重要意义,可以大幅度地促进海生物领域以及人工智能领域的发展。其中海参作为人们最喜欢的海珍品,海参捕捞机器人的研发尤为关键。针对传统人工捕捞方式效率低、风险高以及破坏海洋生态等问题,设计了一种基于柔性机械手的开架式海参捕捞ROV(缆控水下机器人),围绕其结构设计、动力学建模与水动力分析、视觉系统与控制系统设计以及实验验证展开研究。本文的主要工作内容如下:(1)根据机器人的设计指标和实际工作工况,对其结构进行了模块化设计。确定了机器人主要由整体框架,抓捕机构,电子密封舱,照明与摄像装置,机械臂,推进装置以及浮力材料组成。对整体框架,电子密封舱等装置进行了优化设计,借助ANSYS Workbench进行水下静力学仿真与拓扑优化,验证了结构强度的可靠性,并对整体框架进行了轻量化设计。详细阐述了抓捕机构,推进装置,照明与摄像装置以及浮力材料的设计,以保证ROV机器人能够在水下平稳高效地进行抓捕工作。(2)建立惯性坐标系E-(?)η(?)与本体坐标系O-xyz两种坐标系,构建了包含惯性矩阵、科氏力与向心力矩阵、恢复力矩阵、驱动力向量的六自由度动力学模型。针对机器人的工作环境和结构特征简化运动学模型,并借助ANSYS Fluent软件对水下机器人进行水阻力特性分析,获取相关水动力系数,进一步完善机器人的动力学模型。(3)分别对机器人的视觉系统和控制系统进行了设计与研究。视觉系统以YOLOv5s目标检测算法为核心。针对真实水下环境,引入直方图均衡技术与自动色彩均衡技术,改善传统YOLOv5s算法的同时增强海生物的检测识别效果,实现机器人对海生物的实时检测,为后续精准抓捕作业提供可靠的视觉感知。控制系统以水上远程监控与水下自主控制的双重模式为基础,结合PID双闭环控制算法,实现姿态调节与定深控制,保证了机器人多自由度的正常运动。(4)通过设计推进器推力实验以及目标检测系统实验,分析理论值与实验值的差异,验证了推进装置以及视觉系统的可行性并进一步优化和改善了机器人的系统设计。更多还原

【Abstract】 The development of underwater fishing robots is of great significance in the field of marine creature fishing,which can significantly promote the development of both the marine biology and artificial intelligence fields.Among them,as sea cucumbers are one of the most favored marine delicacies,the development of sea cucumber fishing robots is particularly crucial.Aiming at the problems of low efficiency,high risk,and damage to the marine ecosystem caused by traditional manual fishing methods,an open-frame remotely operated vehicle(ROV)for sea cucumber fishing based on a flexible manipulator was designed.The research focused on its structural design,dynamic modeling and hydrodynamic analysis,visual system and control system design,as well as experimental verification.The main contents of this thesis are as follows:(1)According to the design indicators of the robot and the actual working conditions,the overall structural design scheme of the ROV robot is given.It was determined that the robot is mainly composed of the overall frame,the capture mechanism,the electronic sealed cabin,the lighting and camera device,the mechanical arm,the propulsion device and the buoyancy material.The overall frame,electronic sealed cabin and other devices were optimized and designed.Underwater static simulation and topological optimization were carried out through ANSYS Workbench to verify the reliability of the structural strength,and the lightweight design of the overall frame was also conducted.The design of the new capture mechanism,propulsion device,lighting and camera device,as well as buoyancy material was elaborated in detail to ensure that the ROV robot can carry out the capture work smoothly and efficiently underwater.(2)Two coordinate systems,namely the inertial coordinate system E-and the bulk coordinate system,were established,and a six-degree-of-freedom dynamic model including the inertial matrix,the Coriard force and centripetal force matrix,the restoring force matrix,and the driving force vector was constructed.The kinematic model was simplified for the working environment and structural characteristics of the fishing robot.The water resistance characteristics of the underwater robot were analyzed through ANSYS Fluent software to obtain the relevant hydrodynamic coefficients,and the establishment of the dynamic model was completed.(3)A visual system was developed with the YOLOv5s object detection algorithm as the core.In response to the real underwater environment,histogram equalization technology and automatic color equalization technology were introduced to improve the traditional YOLOv5s algorithm and enhance the detection and recognition effect of marine creatures,realizing real-time detection of marine creatures by the robot and providing reliable visual perception for subsequent precise capturing operations.A control system was developed based on the dual-mode of remote monitoring on the water surface and autonomous control underwater.Combined with the PID dual-closed-loop control algorithm,attitude adjustment and depth-keeping control were realized,ensuring the normal movement of the robot in multiple degrees of freedom.(4)By designing the thruster thrust experiment and the target detection system experiment,the differences between the theoretical values and the experimental values were analyzed.The feasibility of the propulsion device and the vision system was verified,and the system design of the robot was further optimized and improved.更多还原