SIA OpenIR  > 机器人学研究室
飞行机器人吊运系统的建模与控制方法研究
Alternative TitleModeling and Control of Unmanned Aerial Vehicles Transporting a Cable-Suspended Payload
易奎1,2
Department机器人学研究室
Thesis Advisor韩建达
Keyword飞行机器人吊运系统 单机吊运系统 多机吊运系统 接触过渡控制 载荷动态分配
Pages122页
Degree Discipline模式识别与智能系统
Degree Name博士
2020-11-27
Degree Grantor中国科学院沈阳自动化研究所
Place of Conferral沈阳
Abstract随着机器人技术、控制理论的发展,飞行机器人的相关研究备受关注。飞行机器人吊运系统可以到达偏远山区或灾后现场,不需要过多考虑负载的形状、体积等,装卸物资非常方便,在森林灭火、货物运送、灾后救援等方面占据很好的应用前景。但是,飞行机器人吊运系统的研究尚不成熟,还存在诸多的科学问题,如负载突变的单机吊运系统建模与控制、飞行机器人的接触过渡控制、多机吊运系统的载荷动态分配等。为了解决上述科学问题,本文基于四旋翼飞行机器人、单机吊运系统、多机吊运系统开展了研究工作,主要内容如下:1. 针对单机吊运系统的轨迹跟踪问题,基于Newton-Euler及Euler-Lagrange方法进行整体建模,进而得到单机吊运系统的动力学模型;基于滑模控制方法设计了控制器,并选取Lyapunov函数,证明了系统的稳定性;通过计算机仿真,验证了所提控制策略的有效性。仿真结果表明,所设计的滑模控制器可以使单机吊运系统跟踪上所期望的轨迹,跟踪误差较小,可以满足控制性能需求。2. 针对负载突变的单机吊运系统,对基于Newton-Euler方法建立的四旋翼飞行机器人的动力学模型进行简化处理,得到了系统的名义模型;引入模型差,从而弥补了系统的未建模部分;将模型差与名义模型的状态相结合,构成新的扩展状态,基于卡尔曼滤波,在线估计了系统的模型差;引入最优化方法,从而基于代价函数设计了基于主动建模增强的补偿控制策略,最终的控制器包括名义控制器和补偿控制器两个部分;在四旋翼飞行机器人的对称机臂悬挂2根等长的绳子,搭建负载突变的单机吊运系统实验平台,并设计新颖的实验情节,验证了所用方法的实用性和有效性。实验结果表明,主动建模的模型输出曲线与测量曲线较为重合,主动建模弥补了名义模型的失配问题,提升了模型的准确性;基于主动建模增强的补偿控制策略能够应对单机吊运系统出现负载突变的情况,系统的位置偏移量较小,并且能够较快地回到悬停点附近,其控制效果明显优于名义控制器,提升了系统的抗干扰能力。3. 针对面向柔顺调节的飞行机器人接触过渡控制,首先对四旋翼飞行机器人进行动力学分析,并分析了自由空间下的常规控制;其次,针对飞行机器人从自由空间过渡到约束空间,分析了接触阶段以及接触过渡过程;为了使飞行机器人平稳地从自由空间过渡到约束空间,设计了基于加速度反馈的控制策略;为了展示所提方法的有效性和实用性,基于实验平台设计实验情景,针对低刚度、中刚度、高刚度三种环境下开展对照实验,验证了实验性能。实验结果表明,力控制仅仅可以应对特定场景下的接触过程,当环境刚度发生改变,系统的控制器参数则需要随之改变,若不改变控制器参数,系统则会发生振荡,无法完成平稳过渡;基于加速度反馈增强的控制策略可以适应更多刚度变化的环境,能够使得旋翼飞行机器人在多种环境下从自由空间平稳地过渡到约束空间。4. 针对多机吊运系统的载荷动态分配与控制,以双机吊运系统为例开展了相关研究,首先进行动力学建模、负载分配条件分析以及阐述控制目标;为了实现双机吊运系统的载荷动态分配与控制,基于导纳控制的思想设计控制策略;为了验证所提方法的有效性,设计了2种仿真情节进行仿真,并对结果进行了分析和讨论;为了进一步验证所提方法的实用性,设计了双机吊运系统的实验,并完成了初步实验。仿真结果表明,所提出的控制框架可以使得系统从初始位置运动到期望位置,并可以实现负载的重分配,每个飞行机器人所承载的拉力可以达到期望值,进而在某一或某些飞行机器人负载能力下降甚至出现故障时仍然能完成协同运输任务。实验结果表明,目前可以实现协同运输,系统可以从初始位置到达期望位置;负载的在线重分配实验效果还需要进一步完善。综上所述,本文主要对单机吊运系统的轨迹跟踪控制、负载突变的单机吊运系统建模与抗扰控制、面向柔顺调节的飞行机器人接触过渡控制、双机吊运系统的载荷动态分配与控制四个方面展开研究,并通过相关仿真或实验展示了所提方法的有效性和实用性。
Other AbstractWith the development of robotics and control theory, the research of aerial robots is getting more and more attention. Aerial robot slung load system can arrive at remote places or the scene of disaster, and the shape or volume of the suspended load is unimportant. The suspended load can be easily loaded and unloaded, and the aerial robot slung load system can be used for transporting objects, extinguishing fires, rescuing people in dangerous situations, and so on. The system is promising, but some scientific problems have to be solved, such as the modeling and control of a quadrotor carrying a changed slung load, contact transition control for a quadrotor and the load dynamic allocation for multi-lift rotorcraft systems. To cope with the above challenges, this dissertation conducts the following research: 1. In terms of the trajectory tracking problem of the single aerial robot slung load system, this dissertation first makes the mathematical model based on Newton-Euler and Euler-Lagrange method. Then, a sliding mode controller is designed and the system stability is proven on the basis of choosing Lyapunov candidates. Finally, some simulation results are given to show the effectiveness of the proposed sliding mode controller. It can be seen from the simulation curves that the system can achieve its desired trajectory, and the tracking errors are small. That is to say, the designed strategy can meet the requirements. 2. In terms of the quadrotor carrying a changed slung load, this dissertation first obtains the nominal model of the system by simplifying the dynamic model of the quadrotor at the hovering point, which is based on Newton-Euler method. By introducing modeling errors, the unmodeled parts can be compensated. The new extended states can be obtained by combing modeling errors with the states of the nominal model. On the basis of Kalman filter, the modeling errors can be estimated. Based on optimization method, the active-model-based enhanced compensation control scheme is designed. And the final control consists of the nominal control and the compensation control. To validate the effectiveness and practicability of the utilized method, some novel experimental scenes are designed by using 2 cables to carry the payload on the basis of the self-built platform. It can be seen from the experimental results that the active model outputs with the modeling errors enhancements meet the measurements very well even though there exists the abrupt change of the slung load, which improves the accuracy of the model. In addition, the deviations of the measurements by the active model enhanced control are much smaller than those by the normal PID control. The active model based control obtained a better result in comparison with the one by utilizing the nominal control, which improves the ability of disturbance rejection. 3. In terms of contact transition control for a quadrotor, this dissertation analyzes the dynamic model of the quadrotor and the normal control in free-motion phase. Then, in terms of the quadrotor flying from the free-motion phase to the contact phase, the contact phase and the contact transition are analyzed. To achieve a stable transition from free-motion to contact phase, the acceleration feedback enhanced control is proposed. To illustrate the effectiveness and practicability of the proposed scheme, some novel experimental scenes are designed. In terms of the environment in low stiffness, middle stiffness and high stiffness, many comparable experiments are conducted. It can be seen from experimental results that the acceleration feedback control serves as extra damping that can help to dissipate the kinematic energy of the UAV due to the non-zero approaching velocity, and maintain a stable contact despite the different stiffness environment. However, the force control only cannot guarantee the stable contact transition when the environment is changed. 4. In terms of the load dynamic allocation for multi-lift rotorcraft systems, this dissertation studies the twin-lift rotorcraft system. Firstly, this dissertation introduces the dynamic model, the load distribution condition, and the control objects. In terms of the load dynamic allocation and control of the twin-lift rotorcraft system, the admittance-based control strategy is utilized. To check the strength of the proposed method, 2 simulation scenarios are designed and the results are analyzed. To further illustrate the practicability of the used scheme, some flight experiments are conducted. It can be seen from the simulation results that the system can move from the initial position to the desired one and the cables’ forces can achieve the desired values to realize the load dynamic allocation when one or more quadrotors are confronted with actuator faults resulting in decreasing load-lifting capacities. In addition, the experimental results show that the system can achieve coordinated transportation from the initial position to the desired one and the load dynamic allocation needs to be improved. To sum up, the dissertation mainly focuses on the trajectory tracking control of a single aerial robot slung load system, the modeling and control of a quadrotor carrying a changed slung load, contact transition control for a quadrotor and the load dynamic allocation for multi-lift rotorcraft systems. What's more, some novel experiments are carried out to validate the effectiveness of the utilized strategies.
Language中文
Contribution Rank1
Document Type学位论文
Identifierhttp://ir.sia.cn/handle/173321/27979
Collection机器人学研究室
Affiliation1.中国科学院沈阳自动化研究所
2.中国科学院大学
Recommended Citation
GB/T 7714
易奎. 飞行机器人吊运系统的建模与控制方法研究[D]. 沈阳. 中国科学院沈阳自动化研究所,2020.
Files in This Item:
File Name/Size DocType Version Access License
飞行机器人吊运系统的建模与控制方法研究.(5684KB)学位论文 开放获取CC BY-NC-SAApplication Full Text
Related Services
Recommend this item
Bookmark
Usage statistics
Export to Endnote
Google Scholar
Similar articles in Google Scholar
[易奎]'s Articles
Baidu academic
Similar articles in Baidu academic
[易奎]'s Articles
Bing Scholar
Similar articles in Bing Scholar
[易奎]'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.