CASIA OpenIR  > 毕业生  > 博士学位论文
基于视觉伺服的无人直升机地面目标跟踪方法研究
Alternative TitleGround Target Tracking Methods Based on Visual Servo for Unmanned Helicopters
邓海波
Subtype工学博士
Thesis Advisor侯增广
2011-05-25
Degree Grantor中国科学院研究生院
Place of Conferral中国科学院自动化研究所
Degree Discipline控制理论与控制工程
Keyword无人直升机 飞行控制 视觉伺服 地面目标跟踪 Unmanned Helicopter Flight Control Visual Servo Ground Target Tracking
Abstract无人直升机有着广阔的应用前景,近年来成为了研究热点。飞行控制系统是无人直升机系统的核心,机载视觉系统是扩展无人直升机功能的有效方式。本文围绕无人直升机飞行控制器设计方法和视觉伺服系统设计方法开展研究,主要研究工作和创新点如下: 第一,在分析直升机动力学特性的基础上,设计了一种基于H∞回路成形鲁棒控制的无人直升机飞行控制器。针对该控制器的特点,提出了一种简洁、有效的分层参数整定方法。该方法综合考虑了系统的稳定裕度、频域指标、时域指标以及对舵机的要求,使所设计的飞行控制器性能比较均衡。为提高该控制器的时域性能,提出了对影响时域性能的关键参数进行增益调度的方法,有效地改善了飞行控制器的时域性能。通过仿真实验,验证了所设计的无人直升机飞行控制器的有效性。 第二,基于DSP嵌入式技术建立了无人直升机机载视觉系统,设计并实现了基于DSP的地面目标快速检测方法和算法。针对摄像机固连于机体上的结构设置和运动学特点,提出了一种结合图像信息和机载惯性导航信息的视觉伺服控制方法,该方法克服了摄像机固连于机体上时直升机姿态变化对视觉伺服的影响。飞行实验验证了该视觉系统能够控制微小型无人直升机平滑地跟踪运动目标。 第三,对一种两自由度云台进行了逆运动学分析,推导了云台运动的雅可比矩阵,在此基础上提出了H∞回路成形和前馈控制相结合的云台控制方法。该方法利用机载IMU所测量的机体角速度信息进行前馈控制来解决直升机姿态变化影响机载摄像机视轴稳定的问题。在本文所设计的虚拟现实仿真环境中验证了该方法的有效性。 第四,在上述云台运动控制器的基础上,设计了无人直升机跟踪地面运动目标的视觉系统,提出了一种无人直升机地面运动目标跟踪控制方法,基于该方法给出了跟踪控制律。所提出的方法利用图像信息、直升机姿态信息以及云台关节角信息来测量目标与无人直升机之间的相对位置,利用卡尔曼滤波来估计直升机与目标之间的相对位置和相对速度。在本文建立的虚拟现实仿真环境中验证了该视觉系统的有效性。
Other AbstractUnmanned helicopters have broad application prospects and have become a hot research area. The flight control system is the key to achieve the autonomous flight of unmanned helicopters. The computer vision system onboard is a effective way to enhance the functional of unmanned helicopters. This thesis focuses on flight controller design and visual servo system design methods for unmanned helicopters. The main context and contributions of this thesis are as follows: Firstly, according to the analysis of the helicopter's dynamics, a flight controller based on H∞ loop shaping robust control has been designed. A concise and effective parameters tuning method is proposed for this flight controller. This method takes the stability margin, frequency field index, time field index and the limitation of actuators into consideration. The flight controller tuned by this method achieved a balanced performance. To enhance the controller's performance over wide flight envelope, a gain scheduling method which focuses on the key parameters is proposed. The flight controller is validated by simulation. Secondly, a vision system onboard is built using a DSP embedded system. A fast target detection method is designed and implemented on this DSP embedded system. Since the camera is fixed on the helicopter, the varying attitude of the helicopter can bring difficulty to visual servo. A new visual servo method which combines the image information and the inertial navigation information is proposed to deal with the varying attitude of the helicopter. The vision system has successfully controlled an unmanned helicopter to follow a moving target in experimental flights. Thirdly, based on the inverse kinematical analysis of a two degree-of-freedom pan-tilt camera and the Jacobian matrix of the pan-tilt camera, a motion controller which combines the H∞ loop shaping robust control and feed-forward control is proposed to control the pan-tilt camera. The helicopter's angular velocity measured by IMU onboard is used by the feed-forward control to solve the influence of the helicopter's varying attitude on the stabilization of the camera's line-of-sight. The motion controller is validated in simulation. Fourthly, based on the motion controller of the pan-tilt camera, a vision system onboard for tracking ground targets is designed. A ground moving target tracking method for the helicopter is proposed and a tracking control law is designed based this method. The proposed method utili...
shelfnumXWLW1655
Other Identifier200718014628004
Language中文
Document Type学位论文
Identifierhttp://ir.ia.ac.cn/handle/173211/6337
Collection毕业生_博士学位论文
Recommended Citation
GB/T 7714
邓海波. 基于视觉伺服的无人直升机地面目标跟踪方法研究[D]. 中国科学院自动化研究所. 中国科学院研究生院,2011.
Files in This Item:
File Name/Size DocType Version Access License
CASIA_20071801462800(5848KB) 暂不开放CC BY-NC-SAApplication Full Text
Related Services
Recommend this item
Bookmark
Usage statistics
Export to Endnote
Google Scholar
Similar articles in Google Scholar
[邓海波]'s Articles
Baidu academic
Similar articles in Baidu academic
[邓海波]'s Articles
Bing Scholar
Similar articles in Bing Scholar
[邓海波]'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.