SIA OpenIR  > 水下机器人研究室
基于ROV援潜救生自主作业方法研究
Alternative TitleResearch on Autonomous Operation of Submarine Escape and Rescue Based on ROV
李延富1,2
Department水下机器人技术研究室
Thesis Advisor封锡盛
ClassificationTP242.3
Keyword援潜救生 Rov-a系统 力控制 运动规划 水下目标定位
Call NumberTP242.3/L36/2009
Pages106页
Degree Discipline模式识别与智能系统
Degree Name博士
2009-05-20
Degree Grantor中国科学院沈阳自动化研究所
Place of Conferral沈阳
Abstract利用ROV代替潜水员进行潜艇供排气管对接作业是援潜救生的发展趋势,其中供排气管对接技术是为失事艇员提供生命保障的关键技术。由于水下作业环境比较复杂以及水下作业难度大等特点,人工遥控ROV进行援潜对接供排空气管作业过程中,还停留在只是能够打开潜艇供排气系统花甲板盖的工作上,还没有实现对类似接通供排空气管精确作业的工作。受到海流、风浪、失事潜艇周围较大涡流、供排气管的拖曳阻力、水下可视条件以及高精度定位等因素的限制,作业任务越复杂,要求ROV系统上机械手精确定位精度越高,操纵的难度也越大,由人工操作很难实现。在现有ROV人工操作的基础上,使ROV系统具有自主作业功能,更能发挥其在对接空气管作业中的能力,提高作业效率,缩短救援时间,因此本文在现有ROV基础上扩展预编程自主控制驾驶功能模块即ROV-A系统,借助预编程技术ROV-A系统在对接空气管作业中的一些特殊任务时能够实现自主作业功能。本文结合中国科学院沈阳自动化研究所水下机器人中心未来援潜救生关键技术研究内容,针对为失事潜艇接通空气管作业中的一些关键技术,开展基于具有自主作业能力的ROV-A系统自主作业方法研究,深入研究了具有自主作业能力ROV-A系统的运动规划与协调控制技术以及水下作业力控制技术,以实现高精度的机械手末端位置控制与期望力跟踪;针对潜艇供排气系统的已知结构研究了水下目标定位技术,为对接空气管的特定需求发展具有自主作业能力的水下作业技术提供理论依据。重点研究针对系统特点的运动规划与多性能指标实现归一化问题;研究基于系统动力学模型误差的系统位姿控制问题;研究基于阻抗力控制水下作业目标定位问题;在系统运动规划与控制研究的基础上,研究能实现机械手末端精确轨迹跟踪的控制策略问题以及力控制问题。本论文研究内容如下:(1)根据ROV-A系统特点,从描述系统的空间位置和姿态着手,研究了ROV-A系统的空间运动学与基于Kane方法的动力学,得出载体基座自由运动模式的系统空间运动数学模型,为论文后续研究工作奠定了基础。(2)针对对接空气管作业中的一些高精度和复杂的作业,研究了系统作业时的运动规划问题。针对系统运动学冗余、作业规划约束性能指标多,例如机械手可操作性,关节限制,载体的姿态,避障等等,结合梯度投影法和最小范数伪逆矩阵法以及任务优先逆运动学方法,引入模糊控制技术,将模糊控制的定性知识表达与任务优先逆运动学算法相结合,以水下作业末端位置控制及轨迹跟踪为前提,对系统运动分配、关节限位、避免奇异有无海流下的系统性能优化等运动规划进行了研究,仿真证明运动规划算法的有效性。(3)研究了基于动力学模型的系统基本控制问题。针对系统动力学模型的复杂性和不精确性,在基于载体输出反馈控制的基础上,设计了基于输出反馈自适应控制算法,通过自适应学习的方式直接逼近系统动力学方程状态量之间的非线性关系,在外界不停扰动下逐渐提高系统的控制精度;基于Lyapunov稳定性理论,证明了存在外界干扰和自适应逼近误差条件下ROV-A系统控制器的闭环稳定性;通过仿真实验验证控制系统的有效性,为系统的控制提供了一种新思路。同时在基于位置运动控制的基础上引入了力控制,通过对常规PID外环力控制器基础上的改进,在系统离线规划的前提下实现力与位置控制的解耦,通过仿真验证期望力的稳定跟踪。(4)研究了基于ROV-A系统阻抗力控制的水下矩形围壁环境约束的位置定位方法。通过在对具有先验知识的矩形围壁环境约束的学习理解基础上,应用外部阻抗力控制环包容内部位置运动控制环的控制策略。利用力传感器的反馈信息变化确定系统末端执行器与约束环境的接触特征点,根据环境的先验知识推理出环境定位信息。通过计算机仿真实验验证了水下具有先验知识环境的定位方法和ROV-A系统的控制策略的有效性。仿真结果证明定位方法和控制策略是可行的。(5)结合系统控制,分析了系统在为失事潜艇对接空气管作业中的恒定和时变两种期望力输入作业模式,提出了基于在线运动规划下外部力控制环包容内部位置环控制方法,利用ROV-A系统完成了为失事潜艇自主对接空气管中两个典型作业。介绍了整个控制方法的组成和执行过程,分析了综合控制策略,最后通过仿真实验分析了综合力控制方法的性能,包括恒力和时变期望力输入进行了深入研究。
Other AbstractUsing ROVs instead of diver carrying out underwater operation is the development trend of submarine escape and rescue, interfacing technique of salvage air hose is the key technique which provides life-support for the wrecked staffs. Manual remotely control of the ROV only is able to open the submarine docking during the process of interfacing technique of salvage air hose because of the complexity of the underwater environment and difficultly underwater operation, and not achieve a similar precision for interfacing technique of salvage air hose. Because of some constraints, for example, the currents, winds and waves, large eddy around the submarine, drag resistance of salvage air hose, underwater visibility conditions, as well as factors such as high-precision positioning, it is very difficult to achieve high-precision positioning by manual operation during the process of the interfacing technique of salvage air hose. Autonomous operation technique based on the existing ROV system can give fully play to the interfacing technique of salvage air hose capacity and improve operating efficiency, shorten rescue time. So, the method of the autonomous operation is presented by expansion of the existing ROV (ROV-A) functional module technology using pre-programmed in the special operations missions of the interfacing technique of salvage air hose. Combining with the future program of Shenyang Institute of Automation “Key technique research on autonomous control of underwater working system applied on operation of submarine escape and rescue”, centering on autonomous working of ROV-A, this thesis thoroughly studies on coordinated motion planning and control of autonomous ROV-A. The goal is to realize position and force control and trajectory-tracking of the manipulator’s end-effector, in same times, localization of the underwater rectangle surroundings and supply theory for developing underwater vehicle with autonomous working ability to explore deep-sea resource. The research is concentrated on motion planning and motion realization of autonomous ROV-A, control of the vehicle and manipulator based on disturbance resisting control of the vehicle based dynamical model, and in the last, with the research basis of motion planning and control, strategies to realize manipulator end trajectory-tracking with high precision and direct force control are discussed. The following are main research contents of this thesis: (1) According to characteristic of ROV-A, from the description ROV-A space position and attitude, the space kinematics and dynamics of the ROV-A is derived based Kane method. The research has a good support for ROV-A mathematics model, and it establishes good foundation for the following research of the thesis. (2) Motion planning based on kinematically redundant system and handling a large number of varialbles, e.g., the manipulator manipulability, the joint range limits to avoid mechanical breaks, the vehicle roll and pitch angles for correct tuning of the proximity sensors, etc. is proposed, which is realized by combining weighted least-norm solution with gradient projection method and singularity-robust task priority redundancy resolution, and inducing the fuzzy inference technic. Computer simulation shows the validity of the method proposed. (3) Due to the complexity and imprecision of system dynamical model, based on dynamical model and output feedback, an adaptive controller is proposed. The control errors are decreasing by the controller’s self-study with input model parameter error. In the presence of outside disturbance forces error, the controller’s design and stability analysis is presented using Lyapunov theory. The correctness and validity of the proposed control method is validated by pool experiments, and the research offers a new way to control vehicle subsystem of any ROV-A. Futhernore, in view of the development of a ROV-A able to perform a completely autonomous mission the capability of the end-effctor to interest with the environment, the control of the force exchanged with the environment is investigated. By improving on PID control algorithm, which is done to avoid the need to directly access the control input, an external force control scheme is exploited to avoid dynamic coupling between vehicle and manipulator. (4) A localization of the underwater rectangle surroundings based impedance force control for ROV-A is presented. By tracking the reference trajectory the end-effectors of the underwater work system can follow the underwater rectangle surroundings surface. The position and orientation of the rectangle surroundings environments with respected to the ROV-A frame can be obtained by estimating the specific contact point acquired by changing of the feedback force. Computer simulations on an ROV-A with a 3-link underwater manipulator are conducted to verify the force tracking capability of the proposed control strategy. The method of localization of the underwater rectangle surroundings is effective. (5) Combining with system control, by analyzing two kind of operation model of pipe-interfacing technology applied on operation of submarine escape and rescue, a control method based dynamical output feedback adaptive external force is presented. The composition and the implementation of the entire control method have been introduced. Finally control method performance has been analyzed through the simulation experiment, including the constant and the time-variable expectation force input control.
Language中文
Contribution Rank1
Document Type学位论文
Identifierhttp://ir.sia.cn/handle/173321/416
Collection水下机器人研究室
Affiliation1.中国科学院沈阳自动化研究所
2.中国科学院研究生院
Recommended Citation
GB/T 7714
李延富. 基于ROV援潜救生自主作业方法研究[D]. 沈阳. 中国科学院沈阳自动化研究所,2009.
Files in This Item:
File Name/Size DocType Version Access License
10001_20051801472801(11211KB) 开放获取--Application Full Text
Related Services
Recommend this item
Bookmark
Usage statistics
Export to Endnote
Google Scholar
Similar articles in Google Scholar
[李延富]'s Articles
Baidu academic
Similar articles in Baidu academic
[李延富]'s Articles
Bing Scholar
Similar articles in Bing Scholar
[李延富]'s Articles
Terms of Use
No data!
Social Bookmark/Share
All comments (0)
No comment.
 

Items in the repository are protected by copyright, with all rights reserved, unless otherwise indicated.