水面移动机器人建模与控制
文献类型：学位论文
| 作者 | 熊俊峰; 孙庆刚
|
| 学位类别 | 博士 |
| 答辩日期 | 2018-05-19 |
| 授予单位 | 中国科学院沈阳自动化研究所 |
| 授予地点 | 沈阳 |
| 导师 | 韩建达 ; 何玉庆 |
| 关键词 | 水面移动机器人 动力学建模 线性变参数系统 变增益控制 扰动估计 |
| 其他题名 | Modeling and Control of Unmanned Surface Vehicles |
| 学位专业 | 模式识别与智能系统 |
| 中文摘要 | 由于具有伤亡风险低、隐蔽性好、机动灵活、成本低廉、续航时间长等优点，水面移动机器人（USV，Unmanned Surface Vehicle）在军事勘察、海洋测绘、灾难救援等领域扮演着重要的角色。然而，通常USV靠喷水推进泵（或主螺旋桨）和方向舵（或两个不可调的螺旋桨）与水相对运动产生反作用力来为其提供动力，导致USV具有强非线性、欠驱动、强耦合的特点，并且容易受到风、浪和水流等干扰的影响。这给动力学建模与控制器设计带来了巨大的挑战。为此，本文围绕USV的建模与控制两大关键问题展开，主要工作如下：论文第二章，提出了USV的准LPV（Linear Parameter Varying）模型。针对USV动力学容易受到Froude数影响的特点，将Froude数作为变参数，用来构建一系列具有线性结构形式的非线性模型。一方面，USV的准LPV模型具有线性模型的简洁结构形式；另一方面，变参数Froude数可以反映原系统的非线性行为特性，提高系统的建模精度。论文第三章，提出了基于准LPV模型的鲁棒控制框架，将USV的控制器分解为航速控制器与航向控制器。一方面，实现了对USV控制器的部分解耦，简化了控制器设计；另一方面，航速控制器中的前向速度作为航向控制器的变参数，提高了控制性能。同时，针对USV容易受到风、浪和水流等干扰以及水动力学辨识不精确的影响，引入了鲁棒控制策略，验证了控制器的稳定性与鲁棒性。论文第四章，在LPV鲁棒控制框架的基础上引入了主动增强控制技术。由于鲁棒LPV控制器考虑的是最差情况下的稳定性，增加了系统保守性，因此需要在控制器中实时考虑当前扰动对控制器的影响。为此，将风、浪和水流等未建模误差进行在线实时估计并反馈给控制器修正，用以提高控制器的控制性能。论文第五章，研制了USV实验系统，介绍了软硬件组成，为模型参数辨识与控制器设计提供了实验基础。最后，在结论与展望部分，对论文所进行的研究作了简要总结与展望。 |
| 英文摘要 | Unmanned Surface Vehicle (USV) plays an important role in military investigating, marine surveying, and disaster rescuing due to its low risk of casualties, good concealment, flexibility, low cost and long endurance. However, USV is usually powered by the reaction force generated by the relative motion among the water-jet propulsion pump (or main propeller), the rudder (or two non-adjustable propellers) and the water. This leads the fact that its model is strong nonlinear, non-holonomic, under-actuated, strong coupling and sensitive to external disturbances such as wind, wave and current. Therefore, the dynamic modeling and controller design of USV pose enormous challenges. To solve the two major key problems, this paper focused on the modeling and control of the USV. The main works are as follows: The second chapter of the paper proposes a quasi-LPV (Linear Parameter Varying) model of USV. Since the dynamics of USV are susceptible to the Froude number, the Froude number is used as a varying parameter to construct a nonlinear model with a linear structure at each working point. On the one hand, the LPV model of USV has a simple linear structure; On the other hand, the Froude number varying parameter can reflect the nonlinear behavior of the original system and improve the modeling accuracy. The third chapter of the thesis proposes a robust control framework based on the quasi-LPV model, which decomposes the controller of USV into speed controller and heading controller. On the one hand, the heading controller of the USV can be simplified by decoupling from of the full state controller partly; one the other hand, the control performance is improved by using the forward speed as the varying parameter. Meanwhile, a robust control strategy is introduced in the controller above to attenuate the influences from wind, wave, current, and the inaccuracy of hydrodynamic identification. Finally, the feasibility and robustness is verified. The fourth chapter of the thesis introduces the active enhanced control technology based on the LPV robust control framework. Since the robust LPV controller considers the worst-case stability and increases system conservatism, it is necessary to consider the current disturbances on the controller online. Therefore, unmodeled errors such as wind, wave, and current are estimated online and feedback to the controller to improve the control performance. Chapter 5 of the dissertation develops the USV experimental system, and introduces the hardware and software components which provides experimental basis for model parameter identification and controller design. Finally, in the conclusion and the outlook, the paper summarizes the research and describes the possible future research direction briefly. |
| 语种 | 中文 |
| 产权排序 | 1 |
| 页码 | 122页 |
| 源URL | [http://ir.sia.cn/handle/173321/21789]  |
| 专题 | 沈阳自动化研究所_海洋信息技术装备中心 |
| 推荐引用方式 GB/T 7714 | 熊俊峰,孙庆刚. 水面移动机器人建模与控制[D]. 沈阳. 中国科学院沈阳自动化研究所. 2018. |
入库方式： OAI收割
来源：沈阳自动化研究所
浏览0
下载0
收藏0
其他版本
除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。