【作者】 乔晋崴；

【导师】 尚建忠；

【作者基本信息】 国防科学技术大学 ， 机械工程， 2012， 博士

【摘要】 随着现代化工业技术的发展，各种粗细不均、蜿蜒迂回的管路大量存在于石油化工、制冷、核发电与居民日常生活等领域。管内机器人研究技术的发展为管道内部的检测与维修提供了新的途径。虽然管内机器人领域已有许多研究成果，但在现实复杂的管内环境中运行时仍表现出一些不足，许多研究机构正积极致力于提升各类型管内机器人的运动性能以加速其在实际生产中的工程化应用。针对核反应堆蒸发器、石油水平井等领域中存在的对大牵引力快速推进管道机器人的紧迫性需求问题，本文结合自主锁止机构的功能与特性，利用公理设计的方法进行基于自主锁止机理的伸缩式管内机器人的概念设计、参数优化、特性分析、虚拟仿真与样机研制，克服了传统管内机器人牵引能力受其自身与管壁间恒定最大摩擦力的限制、实现了机器人运动速度与牵引能力的可单独调节，其中“自主锁止”是指管内机器人在无需主动控制的情况下可实现与管壁的自动锁止。本文主要贡献和创新点归纳如下：(1)深入概括了现有各种类型的管内机器人，提出了基于接触方式的分类方法。目前管内机器人种类繁多，大小不一，难以进行不同类别间的分析与比较，研究重点集中于动力输出装置。本文以机器人与管壁间不同接触方式为划分准则将管内机器人分为三类：面接触式、点线接触式和非接触类，方便了不同类别管内机器人间运动性能对比分析。(2)深入分析了管内机器人的设计过程，首次将公理设计方法引入管内机器人的设计，提出基于自主锁止机理伸缩式管内机器人的概念。现有管内机器人牵引能力有限，无法超越其支撑机构与管壁间恒定的最大静摩擦力，机器人运动速度与牵引能力不能得到单独调节。本文采用公理设计的方法进行基于自主锁止机理管内机器人的概念设计与耦合性分析。作为对比，针对轮式与气囊伸缩式管内机器人的设计过程进行耦合性分析，结果显示前者为耦合设计，后者为解耦设计，找到了问题的根源。(3)详细设计了自主锁止式管内机器人结构参数，将概念设计具体化，提出一系列指导性设计准则与优化理论，推导出伸缩弹簧弯曲刚度的计算公式。细长、封闭且存在径向尺寸变化的管内环境对管内机器人的外形、连接单元以及安全性设计提出了全新的要求。本文围绕两种锁止结构，求出为实现自主锁止所应满足的具体关系式，优化凸轮接触轮廓，仿真计算永磁体间排斥力，求解锁止装置解锁所需最小作用力与最短时间。采用伸缩弹簧作为柔性连接单元，应用细杆弹性理论首次推导出伸缩弹簧弯曲刚度，详细设计了应用于管道内部的安全离合器。(4)深入研究了自主锁止结构的锁止特性，在完成锁止功能虚拟验证的基础上建立了自主锁止式管内机器人牵引能力与机械本体定位精度的计算公式。牵引能力不足是许多常见管内机器人无法应用于工程实际的一个重要因素，同时管内定位功能的实现则较多地依赖于各种传感器。本文以惯性冲击式为例进行常见管内机器人的牵引能力分析，针对两种锁止机构，定量地推导出牵引能力与自身定位精度的计算公式。结果显示，文中提出的自主锁止机构理论牵引能力可随外载荷需求自动增长。(5)深入研究了管内机器人在通过弯管的整个过程中所需满足的几何位形约束，提出了指导设计支撑腿弹性变形量的理论准则。合适的变形能力可以确保弹性支撑腿在整个运动过程中与管壁始终保持接触，避免通过弯管时出现的运动不连续现象，同时为机体留足尺寸空间，而现有研究仅仅得出了机器人整体外形尺寸所应满足的关系式。本文通过建立机器人在弯管中的数学模型，求得支撑腿实时变化量计算公式，通过选取不同结构参数代入Matlab程序进行分析，得出支撑腿的实时变形量及最大变形需求量，分析研究了不同结构参数对变形量的影响程度与影响规律。(6)针对自主锁止式管内机器人展开样机研制、综合试验系统搭建、原理性验证试验与性能测试试验。在完成理论设计与虚拟样机验证的基础上，以各设计准则为指导，研制了两台试验样机，开发了样机试验平台，完成了自主锁止式机构可行性验证试验，对机器人的爬升性能、牵引负载性能、运动速度性能、以及解耦性进行了试验测试与验证。首次提出驱动力空间比的概念，使不同类别管内机器人的驱动性能具备了可比性，量化了驱动能力。试验结果表明，两种类型自主锁止式管内机器人可以实现在管道内部的快速行走与定位。其中直径为Ф16mm的试验样机，可顺利在竖直管道内快速爬升，推动15.2N的重物前进，最大运动速度为13.72mm/s，驱动力空间比达到9.364（现有常见值均小于1），实现了运动速度与牵引负载的解耦。更多还原

【Abstract】 With the development of modern industrial technology, plenty of pipelines withvarious diameters and shapes have formed a complex system which successfullyoccupied every corner of human daily life, even in petrochemical industry, refrigeratingindustry and nuclear power plant. The development of in-pipe robot technique providesin-service pipelines with a potential possibility to perform inside surveillances andmaintenances. Many achievements have been reported in the in-pipe robot researchdomain, but there are still many unacceptable shortages remained as the robot worksunder tough environment. Therefore, many researchers are devoting themselves toimproving the in-pipe robot performance, aiming at accelerating its application toreality.The request of powerful and fast moving in-pipe robot is urgent from theapplication areas like the nuclear reactor evaporator and the horizontal petroleum well.By utilizing functions and properties of self-locking mechanism, this thesis makesresearch on the telescopic in-pipe robot based on self-locking principle. It performedconcept design, parameter optimization, property analysis, virtual simulation andprototype developing, which as a result breaks through the traditional constraint ofmaximum friction imposed on in-pipe robot traction ability, and makes the movingvelocity and transaction ability individually adjustable. The referred self-locking can beexplained as, the in-pipe robot could get locked with the pipeline automatically, withoutany help of control. Main contributions and innovations of this thesis are summed up as:(1) Present various in-pipe robots are summarized, and a contact style basedclassification method is proposed. It is not easy to make contrast and analysis amongdifferent in-pipe robots with diverse styles and varying dimensions, while relatedresearches are focusing on actuating systems. The thesis divides in-pipe robots intothree groups depending on different contact styles with pipeline, which makes it moreconvenient to compare different in-pipe robots. One group has faces contacted withpipeline, one has lines or points, and the left one has nothing.(2) For the first time the Axiomatic Design (AD) method is introduced into thedesign process of in-pipe robot that has been explored, and the concept of telescopicin-pipe robot based on self-locking principle is proposed. The traction ability of presentin-pipe robot is limited by the constant maximum friction between pipeline and therobot supportive mechanism, and can’t be adjusted without putting influence on itsmoving velocity. The concept design and coupling analysis are carried out following theAD theory. As a contrast, the wheeled and pneumatic telescopic in-pipe robots are bothmade coupling analysis with the results of coupled design and decoupled design,separately. (3) The in-pipe robot construction is deliberately designed, which embodies theproposed concept. A series of guide specifications and optimization theories arepresented, as well as the bending rigidity of compression spring. With varying diameter,the long and thin pipeline raises new requirement for in-pipe robot construction design,linkage design and security design. Focusing on two proposed mechanisms, it derivesthe self-locking requirement relationship, optimizes the cam contact profile, simulatesand gets the repelling force between two sets of permanent magnets, calculates theminimum required force and minimum action time. As the flexible linkage of thein-pipe robot, the spring is investigated and its bending rigidity calculation formulas arefound out based on the thin elastic rod theory. The security clutch has been developedfor the application in pipelines.(4) The property of self-locking mechanism is explored. After successful validityof partial assembly and the whole robot, the formulation relationships for calculating themaximum driving ability and the mechanically localization accuracy are built up. Theshortage of traction ability is one of the main reasons that blocks the application ofin-pipe robots in reality, meanwhile its local ability heavily depends on various sensors.The thesis takes inertia impulse in-pipe robot as an example to make analysis on itstraction ability, and then deduces related calculation formulas about the maximumtraction ability and localization accuracy for the proposed two mechanisms. The resultshows that the traction ability of the self-locking mechanism can increase automaticallywith the rise of outside payload.(5) The geometry constraint is explored as the in-pipe robot passes through anelbow, and theoretical rules are proposed to guide the design of supportive legdeflection. With a proper deflection, the elastic legs remain contact with the pipeline allthe time when moving in a pipeline, thus avoiding discontinuous motions and leavingrobot body with enough space. However, reported researches have merely establishedthe constraint relation about the outline dimension. The thesis obtains the formulas ofreal-time deflection of supportive legs by setting up the mathematical model for in-piperobot inside pipelines, gets the maximum and needed deflection with the help of Matlab,and reveals how the deflection effected by different fabric parameters.(6) About the self-locking in-pipe robot, the prototype is developed, thecomprehensive test platform is set up, and the principle test is conducted, as well as theperformance test. Guided by the deduced design rules, two prototypes are developed.On the specialized test platform, the validity of self-locking mechanism is carried out,as well as its performance of climbing, traction ability, moving velocity and decoupling.The concept of ratio of traction ability to space is invented for quantifying the tractionabilities of different in-pipe robots. The experimental results indicate that two types ofself-locking mechanism can help in-pipe robot achieve a fast and stable moving ability,and a accuracy localization ability inside pipelines. The prototype with diameter of 16mm is proved to be capable of climbing vertically set pipelines, carrying a maximumload of15.2N, speeding up to13.72mm/s, pushing the ratio of traction ability to spaceto9.364with normal ones under1, and realizing the uncouple design between movingvelocity and traction ability.更多还原