水下滑翔机海洋特征观测控制策略研究
文献类型：学位论文
| 作者 | 张少伟
|
| 学位类别 | 博士 |
| 答辩日期 | 2013-05-31 |
| 授予单位 | 中国科学院沈阳自动化研究所 |
| 导师 | 张艾群 ; 俞建成 |
| 关键词 | 水下滑翔机 海洋观测 动力学 运动控制 跟踪策略 |
| 其他题名 | Research on Control Strategy of Underwater Gliders for Ocean Feature Observation |
| 学位专业 | 机械电子工程 |
| 中文摘要 | 本文结合国家863计划目标导向类课题“水下滑翔测量系统（2006AA092157）”，中科院知识创新工程“多水下滑翔机自适应采样策略与控制技术研究”，和国家自然科学基金重点项目“水下机器人海洋环境自主观测理论与技术(61233013)”的需求，对基于水下滑翔机的海洋观测与海洋特征跟踪问题进行了深入的研究，主要包括以下内容： 1、海洋观测系统框架的研究。将观测系统依据其功能进行划分，并界定各个模块的功能和范围。针对现有观测系统在观测平台、海洋特征特性、跟踪决策与路径规划上结合不足、海洋特征特性分析欠缺等问题，重点分析了观测平台选用和观测精度的要求；分析了海洋模型与数据同化、实时采样数据估计之间的关系；并阐述了观测数据分析结果在跟踪过程中的作用。最后，以温度场等值线近实时跟踪和剖面跃层观测为例，将海洋观测跟踪策略、实时数据分析与观测跟踪过程结合起来，设计了相应的观测流程。 2、典型海洋现象特征的跟踪策略研究。针对中小尺度海洋现象如上升流、跃层、内波、锋面、涡流等，传统的观测方法在精度上、自主性上有很大欠缺。我们分析了这些现象的成因和变化特性，并建立了这些现象的跟踪决策模型。重点讨论了这些现象在中国近海海域的发生位置、尺度、观测阈值。 3、水下滑翔机动力学建模研究。现有水下滑翔机的区别是转向机构、混合驱动模式上的区别。针对通过内置质量块转动以实现偏航的滑翔机，建立了它的动力学模型；并将仿真结果与试验数据进行对比，重点分析动力学系统的耦合性；针对大深度滑翔过程，设计了海流、盐度、深度对净浮力的补偿方法。 4、水下滑翔机稳态滑翔特性分析。考虑到水下滑翔机稳态滑翔的时间占总工作时间的比重较大，分析了稳态滑翔时在不同的控制输入下，滑翔机状态的变化规律及其影响因素。滑翔机在自主观测过程中，需要根据规划出的速度、角速度等状态反求其控制量，因此基于向量点乘的方法化简了动力学方程，并设计了一种迭代算法求解。 5、水下滑翔机剖面滑翔切换控制方法研究。采用LQR线性二次最优方法设计了不同俯仰姿态下的切换控制方法，以实时调整剖面观测过程中的空间尺度。针对潜浮切换问题，基于最优控制中的两点边值方法设计了切换控制策略，以提高位置控制精度，并抵御海流的影响。 6、海洋温度场等值线跟踪。海洋现象特征的梯度是反映其变化特性的一个重要表征。海洋温度场等值线跟踪的过程，包括海洋特征值、梯度的提取与估计、多滑翔机队形控制律的设计。对于前者，利用多个水下滑翔机的观测数据来估计观测区域的梯度和特征值；并结合跟踪目标，设计了多滑翔机群体的运动方向和速度控制律。对于后者，多滑翔机以一定的队形移动、旋转进行跟踪，因此设计控制律使其沿着等值线切线方向运动，使获得的海洋观测采样值的差最大，以提高估计的效果。最后针对实际的温度场，通过仿真验证该方法的有效性。 |
| 英文摘要 | According to the needs of National 863 Project “Underwater Glider Measurement System (2006AA092157)” , National 863 Project “Autonomous observation technology of ocean environment with underwater vehicles(61233013)”,and Open Foundation of State Key Laboratory of Robotics “research of underwater glider adaptive sampling strategy and control technology”, this dissertation conducts a deep research on control strategy of the underwater gliders for ocean observation. The main contents of this dissertation are as follows: 1. Observational architecture analysis of ocean observation system. The architecture of the observation system is based on the function, so the scope and function of each subsystem is defined. Due to the mobile platform, the control and decision making and path planning, and the lack discussion of the marine characteristics, we give discussion about the autonomous platform selection and observation precision requirements with respect to different observation mission. We give discussion about the ocean model, data assimilation, and the sampling data estimation. We expound the relationship between the data estimation results and the tracking process decision. Lastly, we take the near real time tracking of contour line of the temperature field and upwell observation，as an example to design the observation system and tracking process. 2. Analysis of tracking strategy of typical ocean phenomenon. According to micro-scale and meso-scale ocean phenomenon, such as upwelling, thermocline, internal waves, frontal and eddy, the traditional observation method has greatly lack in precision and autonomy. Therefore we analyze these phenomenon and characteristics, and give the tracking and decision strategy of these phenomena. Especially, for China offshore area, we give discussion about the location, scale, and observational threshold for some ocean phenomenon. 3. Research on the underwater glider dynamics model. The difference of the glider is mainly the steering mechanism and the hybrid mode. For the glider which is steered by the internal mass, we establish the glider dynamic model with the Lagrange equation. We compare the simulation result with the experiment data, and give a thorough discussion about the couple of glider hydrodynamic system. Also we consider the ocean current influence, and design a compensation for the net buoyancy influced by the variation of temperature, depth, salty during the glider deeply diving process. 4. Analysis of underwater glider steady gliding state characteristic. Considering that the underwater glider steady state gliding time accounted for a large proportion of the total work time, we have to analyze the relationships between the control inputs, the glider steady gliding state, and its influencing factors. During the observation process, we need to solve the control input with the specific velocity, and angular velocity that got from the path plan system, so we design the iterative algorithm to solve it. 5. Switching control research of underwater glider vertical sawtooth motion. LQR theory is adopted, so as to adjust the sample spatial density during the vertical profile observation. We give the corresponding control strategy based on the two-point boundary problem theory, so as to enhance the position control precision, and resist the ocean current effect. 6. Isoline tracking simulation and analysis of ocean temperature field. Gradient is one of the important characteristic that reflects the property of the phenomena. The tracking process includes the ocean characteristic value estimation, gradient estimation and glider formation control law design. For the former, we utilize multiple underwater glider observation data to estimate the gradient and the scalar value, then we connect it with the tracking target to design the gliders swarm movement direction and speed control law. For the latter, many gliders keep certain formation and rotation mode for tracking. Therefore we design the control law to drive the glider to move along the contour tangent direction, so as to get the biggest observation difference to improve the estimation effect. Finally, we give the simulation with the actual temperature field to verify the validation of the method. |
| 语种 | 中文 |
| 公开日期 | 2013-08-19 |
| 产权排序 | 1 |
| 页码 | 122页 |
| 分类号 | TP273 |
| 源URL | [http://ir.sia.ac.cn/handle/173321/10790]  |
| 专题 | 沈阳自动化研究所_水下机器人研究室 |
| 推荐引用方式 GB/T 7714 | 张少伟. 水下滑翔机海洋特征观测控制策略研究[D]. 中国科学院沈阳自动化研究所. 2013. |
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。