基于终端滑模的四旋翼飞行器非线性轨迹跟踪控制

考虑到四旋翼飞行器的传统内外环控制策略依赖时标分离假设,稳定性分析复杂,并且控制参数选取困难的缺点,提出了一种与传统内外环控制策略不同的轨迹跟踪控制器;首先将四旋翼飞行器数学模型进行相应的变换,以分解为高度、偏航角和纵横向三个级联的子系统,再使用终端滑模控制方法设计高度和偏航角子系统的控制器,使两个子系统的状态误差可以在有限时间内收敛到原点,之后基于变量非线性变换设计纵横向子系统的控制器,分析了闭环系统稳定性,证明了所设计的轨迹跟踪控制器可以保证闭环系统跟踪误差渐近稳定到原点,最后仿真实验的结果验证了所设计的控制器的有效性。     

基于强化学习的四旋翼无人机控制律设计

目前四旋翼无人机大部分都采用经典控制方法进行控制律的设计,然而控制参数的选择和对被控对象数学模型的依赖一直是经典控制方法设计中需要克服的问题;针对此问题,采用了一种基于深度强化学习算法Deep Q Network的无人机控制律设计方法,以四旋翼姿态角和姿态角速率作为三层神经网络的输入数据,最终输出动作值函数,再根据贪婪策略进行动作的选取,通过与环境的不断交互,智能体根据奖惩信息来更新神经网络的权值,使得智能体朝着获得累积回报最大值的方向选取动作;仿真结果表明在经过强化学习训练之后,四旋翼姿态角能够快速准确地跟踪上参考指令的变化,证明了基于强化学习的四旋翼无人机控制律的可行性,从而避免了传统控制方法对控制参数的选择与控制模型的依赖。     

基于模型的四旋翼无人机半实物仿真平台研究

为了降低在真实飞行器上测试新的控制策略时所存在的设备损坏风险,对四旋翼无人机的控制器和半实物仿真实验平台进行了开发和介绍;首先,分析了四旋翼无人机基本结构和飞行原理,并对其进行了动力学建模;其次,设计了对应的常规PID控制器和滑模控制器,并进行了Matlab仿真对比和分析;最后,展示了采用先进的基于模型的设计方法和代码自动等技术的半实物仿真实验平台,并详细介绍了其硬件和软件的总体架构和不同模块的配置;仿真结果表明所设计滑模控制器相比于PID控制器有更好的控制效果,并给出了其在半实物仿真平台上用来研究四旋翼无人机姿态控制的可行性。     

基于RFID技术的四旋翼无人机轨迹跟踪控制系统设计

为促进四旋翼无人机的飞行自主性,增强无人监管情况下飞行器主机所具备的避障行进能力,设计基于RFID技术的四旋翼无人机轨迹跟踪控制系统;采用RFID标签识别技术,调制处理既定控制信号,利用标签识别协议,连接微型四旋翼轨迹控制器与内环姿态控制器,通过数据通信链路,提取轨迹跟踪控制所需的传输电子量,完成轨迹跟踪控制系统硬件设计;利用动力系统中的参数辨识策略,确定与轨迹姿态控制相关的物理规律标注,实现四旋翼无人机轨迹跟踪控制;实验结果表明,与机器视觉型控制系统相比,基于RFID技术的控制系统的SSI避障行进指标数值相对较高,全局最大值达到了 79％,四旋翼无人机滚转角平均值为85°,能够有效抑制四旋翼无人机滚转角的数值上升趋势,增强无人监管情况下飞行器主机避障行进能力.     

偏振光传感器在四旋翼飞行器中的应用

为了实现偏振光传感器在实际三维空间中的导航应用,设计了一种基于仿生偏振光导航传感器、微惯性测量单元（MIMU）与全球定位系统（GPS）的组合导航控制系统,并实际应用到了四旋翼导航控制之中.本文首先介绍了偏振光传感器的功能模型和测角原理.其次采用扩展卡尔曼滤波（EKF）技术设计了偏振光传感器、MIMU、GPS的信息融合算法,通过室外飞行试验对该导航系统的性能进行了测试,并与传统MIMU/GPS/电子罗盘导航系统进行了比较.结果显示：在有磁场干扰环境下,基于偏振光的导航控制系统平均位置精度较传统导航系统提高了50.4%.实验结果表明：该导航控制系统实时性好、精度较高、抗电磁干扰能力强,且误差不随时间累积,基本满足移动载体进行自主导航时的精度和可靠性要求.     

Flocking control of a fleet of unmanned aerial vehicles

Current applications using single unmanned vehicle have been gradually extended to multiple ones due to their increased efficiency in mission accomplishment, expanded coverage areas and ranges, as well as enhanced system reliability. This paper presents a flocking control method with application to a fleet of unmanned quadrotor helicopters (UQHs). Three critical characteristics of formation keeping, collision avoidance, and velocity matching have been taken into account in the algorithm development to make it capable of accomplishing the desired objectives (like forest/pipeline surveillance) by safely and efficiently operating a group of UQHs. To achieve these, three layered system design philosophy is considered in this study. The first layer is the flocking controller which is designed based on the kinematics of UQH. The modified Cucker and Smale model is used for guaranteeing the convergence of UQHs to flocking, while a repelling force between each two UQHs is also added for ensuring a specified safety distance. The second layer is the motion controller which is devised based on the kinetics of UQH by employing the augmented state-feedback control approach to greatly minimize the steady-state error. The last layer is the UQH system along with its actuators. Two primary contributions have been made in this work: first, different from most of the existing works conducted on agents with double integrator dynamics, a new flocking control algorithm has been designed and implemented on a group of UQHs with nonlinear dynamics. Furthermore, the constraint of fixed neighbouring distance in formation has been relaxed expecting to significantly reduce the complexity caused by the increase of agents number and provide more flexibility to the formation control. Extensive numerical simulations on a group of UQH nonlinear models have been carried out to verify the effectiveness of the proposed method.     

Feedback linearization vs. adaptive sliding mode control for a quadrotor helicopter

This paper presents two types of nonlinear controllers for an autonomous quadrotor helicopter. One type, a feedback linearization controller involves high-order derivative terms and turns out to be quite sensitive to sensor noise as well as modeling uncertainty. The second type involves a new approach to an adaptive sliding mode controller using input augmentation in order to account for the underactuated property of the helicopter, sensor noise, and uncertainty without using control inputs of large magnitude. The sliding mode controller performs very well under noisy conditions, and adaptation can effectively estimate uncertainty such as ground effects. Recommended by Editorial Board member Hyo-Choong Bang under the direction of Editor Hyun Seok Yang. This work was supported by the Korea Research Foundation Grant (MOEHRD) KRF-2005-204-D00002, the Korea Science and Engineering Foundation(KOSEF) grant funded by the Korea government(MOST) R0A-2007-000-10017-0 and Engineering Research Institute at Seoul National University. Daewon Lee received the B.S. degree in Mechanical and Aerospace Engineering from Seoul National University (SNU), Seoul, Korea, in 2005, where he is currently working toward a Ph.D. degree in Mechanical and Aerospace Engineering. He has been a member of the UAV research team at SNU since 2005. His research interests include applications of nonlinear control and vision-based control of UAV. H. Jin Kim received the B.S. degree from Korea Advanced Institute of Technology (KAIST) in 1995, and the M.S. and Ph.D. degrees in Mechanical Engineering from University of California, Berkeley in 1999 and 2001, respectively. From 2002–2004, she was a Postdoctoral Researcher and Lecturer in Electrical Engineering and Computer Science (EECS), University of California, Berkeley (UC Berkeley). From 2004–2009, she was an Assistant Professor in the School of in Mechanical and Aerospace Engineering at Seoul National University (SNU), Seoul, Korea, where she is currently an Associate Professor. Her research interests include applications of nonlinear control theory and artificial intelligence for robotics, motion planning algorithms. Shankar Sastry received the B.Tech. degree from the Indian Institute of Technology, Bombay, in 1977, and the M.S. degree in EECS, the M.A. degree in mathematics, and the Ph.D. degree in EECS from UC Berkeley, in 1979, 1980, and 1981, respectively. He is currently Dean of the College of Engineering at UC Berkeley. He was formerly the Director of the Center for Information Technology Research in the Interest of Society (CITRIS). He served as Chair of the EECS Department from January, 2001 through June 2004. In 2000, he served as Director of the Information Technology Office at DARPA. From 1996 to 1999, he was the Director of the Electronics Research Laboratory at Berkeley (an organized research unit on the Berkeley campus conducting research in computer sciences and all aspects of electrical engineering). He is the NEC Distinguished Professor of Electrical Engineering and Computer Sciences and holds faculty appointments in the Departments of Bioengineering, EECS and Mechanical Engineering. Prior to joining the EECS faculty in 1983 he was a Professor with the Massachusetts Institute of Technology (MIT), Cambridge. He is a member of the National Academy of Engineering and Fellow of the IEEE.     

Experimental quadrotor flight performance using computationally efficient and recursively feasible linear model predictive control

A new linear model predictive control (MPC) algorithm in a state-space framework is presented based on the fusion of two past MPC control laws: steady-state optimal MPC (SSOMPC) and Laguerre optimal MPC (LOMPC). The new controller, SSLOMPC, is demonstrated to have improved feasibility, tracking performance and computation time than its predecessors. This is verified in both simulation and practical experimentation on a quadrotor unmanned air vehicle in an indoor motion-capture testbed. The performance of the control law is experimentally compared with proportional-integral-derivative (PID) and linear quadratic regulator (LQR) controllers in an unconstrained square manoeuvre. The use of soft control output and hard control input constraints is also examined in single and dual constrained manoeuvres.     

Concrete Defects Inspection and 3D Mapping Using CityFlyer Quadrotor Robot

The concrete aging problem has gained more attention in recent years as more bridges and tunnels in the United States lack proper maintenance. Though the Federal Highway Administration requires these public concrete structures to be inspected regularly, on-site manual inspection by human operators is time-consuming and labor-intensive. Conventional inspection approaches for concrete inspection, using RGB image-based thresholding methods, are not able to determine metric information as well as accurate location information for assessed defects for conditions. To address this challenge, we propose a deep neural network (DNN) based concrete inspection system using a quadrotor flying robot (referred to as CityFlyer) mounted with an RGB-D camera. The inspection system introduces several novel modules. Firstly, a visual-inertial fusion approach is introduced to perform camera and robot positioning and structure 3D metric reconstruction. The reconstructed map is used to retrieve the location and metric information of the defects. Secondly, we introduce a DNN model, namely AdaNet, to detect concrete spalling and cracking, with the capability of maintaining robustness under various distances between the camera and concrete surface. In order to train the model, we craft a new dataset, i.e., the concrete structure spalling and cracking (CSSC) dataset, which is released publicly to the research community. Finally, we introduce a 3D semantic mapping method using the annotated framework to reconstruct the concrete structure for visualization. We performed comparative studies and demonstrated that our AdaNet can achieve 8.41% higher detection accuracy than ResNets and VGGs. Moreover, we conducted five field tests, of which three are manual hand-held tests and two are drone-based field tests. These results indicate that our system is capable of performing metric field inspection, and can serve as an effective tool for civil engineers.      

Robust dynamic surface trajectory tracking control for a quadrotor UAV via extended state observer

In this paper, we present an extended state observer–based robust dynamic surface trajectory tracking controller for a quadrotor unmanned aerial vehicle subject to parametric uncertainties and external disturbances. First, the original cascaded dynamics of a quadrotor unmanned aerial vehicle is formulated in a strict form with lumped disturbances to facilitate the backstepping design. Second, based on the separate outer‐ and inner‐loop control methodologies, the extended state observers are constructed to online estimate the unmeasurable velocity states and lumped disturbances existed in translational and rotational dynamics, respectively. Third, to overcome the problem of “explosion of complexity” inherent in backstepping control, the technique of dynamic surface control is utilized for trajectory tracking and attitude stabilization, and with the velocity and disturbance estimates incorporated into the dynamic surface control, a robust dynamic surface flight controller that guarantees asymptotic tracking in the presence of lumped disturbances is synthesized. In addition, the stability analysis is given, showing that the present robust controller can ensure the ultimate boundedness of all signals in the closed‐loop system and make the tracking errors arbitrarily small. Finally, comparisons and extensive simulations under different flight scenarios are performed to validate the effectiveness and superiority of the proposed scheme in accurate tracking performance and enhanced antidisturbance capability.