智能轮椅视觉SLAM仿真系统实现

运用了基于视觉的EKF同时定位与地图创建（SLAM）方法来实现智能轮椅在室内环境下的导航问题。通过对图像的特征提取及匹配、更新确定自身位姿并建立地图。有效准确的特征提取是SLAM实现的必要条件之一。仿真实验显示,所提出的算法可以实现移动机器人的视觉SLAM。     

融合特征法与直接法的RGB-D稠密视觉SLAM算法

为了保持直接法的快速性与特征法的高精度和闭环能力,提出了一种融合直接法与特征法的RGB-D同时定位与地图创建(SLAM)算法。该算法主要包含3个并行线程:跟踪线程、局部建图线程和闭环线程。在跟踪线程中对非关键帧进行跟踪,通过最小化光度图像误差来进行相机的初始位姿估计以及像素点的对应关系计算,利用最小化局部地图点重投影误差进一步优化相机位姿,实现快速准确的跟踪与定位;在局部建图线程中对关键帧进行提取并匹配ORB特征,执行局部BA(光束平差法),对局部关键帧位姿和局部地图点的位置进行优化,提高SLAM的局部一致性;在闭环线程中执行对关键帧的闭环检测和优化,从而保证SLAM全局一致性。另外,根据RGB-D图像和相机位姿信息,通过基于Octomap的建图框架,构建完整准确的3D稠密环境地图。在TUM数据集下的实验表明,所提出的方法可以得到与基于特征法相当的精度,且所需时间更少。     

量测转换卡尔曼滤波在塔康导航中的应用

塔康导航系统是一种无线电测角测距系统,结合高程数据可以确定目标的三维位置。由于塔康导航的量测误差较大,导致其对目标位置的估计精度不高。为解决此问题,用基于量测转换的滤波跟踪技术实现非线性量测下的状态估计。针对塔康导航系统量测值的特点,首先在斜距值与高度所在平面内推导出量测转换的误差统计特性,并将其推广到三维空间,进而推导出塔康导航系统中的量测转换模型。基于所推导模型的卡尔曼滤波器用于塔康导航系统中的目标跟踪,取得良好的效果。与经典滤波算法的性能对比表明,由于严格基于量测值推导量测转换误差的统计特性,算法在滤波误差、置信度和计算量上优于其它算法,可在不同量测噪声水平下稳健滤波,有较好的全面性和鲁棒性。      

Improved Rao-Blackwellized H∞ filter based mobile robot SLAM

For the problems of estimation accuracy, inconsistencies and robustness in mobile robot simultaneous localization and mapping (SLAM), a novel SLAM based on improved Rao-Blackwellized H∞ particle filter (IRBHF-SLAM) algorithm is proposed. The iterated unscented H∞ filter (IUHF) is utilized to accurately calculate the importance density function, repeatedly correcting the state mean and the covariance matrix by the iterative update method. The laser sensor’s observation information is introduced into sequential importance sampling routine. It can avoid the calculation of Jacobian matrix and linearization error accumulation; meanwhile, the robustness of the algorithm is enhanced. IRBHF-SLAM is compared with FastSLAM2.0 and the unscented FastSLAM (UFastSLAM) under different noises in simulation experiments. Results show the algorithm can improve the estimation accuracy and stability. The improved approach, based on the robot operation system (ROS), runs on the Pioneer3-DX robot equipped with a HOKUYO URG-04LX (URG) laser range finder. Experimental results show the improved algorithm can reduce the required number of particles and the operating time; and create online 2 dimensional (2-D) grid-map with high precision in different environments.     

基于扩展卡尔曼滤波器的移动机器人SLAM研究

文中阐述一种移动机器人SLAM问题的解决方法,首先利用激光测距仪得到环境中障碍物的监测图表,然后增量的构建全局地图。利用扩展卡尔曼滤波器（EKF）创建移动机器人定位计算的有界估量;最后通过仿真和物理实验验证了该方法的正确性。可为解决机器人在未知环境下的地图创建与定位问题提供理论依据,具有实际意义。     

Multi‐robot Mapping Using Omnidirectional‐Vision SLAM Based on Fisheye Images

This paper proposes a global mapping algorithm for multiple robots from an omnidirectional‐vision simultaneous localization and mapping (SLAM) approach based on an object extraction method using Lucas–Kanade optical flow motion detection and images obtained through fisheye lenses mounted on robots. The multi‐robot mapping algorithm draws a global map by using map data obtained from all of the individual robots. Global mapping takes a long time to process because it exchanges map data from individual robots while searching all areas. An omnidirectional image sensor has many advantages for object detection and mapping because it can measure all information around a robot simultaneously. The process calculations of the correction algorithm are improved over existing methods by correcting only the object's feature points. The proposed algorithm has two steps: first, a local map is created based on an omnidirectional‐vision SLAM approach for individual robots. Second, a global map is generated by merging individual maps from multiple robots. The reliability of the proposed mapping algorithm is verified through a comparison of maps based on the proposed algorithm and real maps.     

一种基于PL-ICP及NDT点云匹配的单线激光里程计

近年来随着人工智能的迅速发展,服务机器人在许多领域得到应用。里程计为机器人提供基础位姿信息,是机器人完成SLAM、路径规划及导航等任务的基础。基于PL-ICP及NDT点云匹配算法分别构建两种激光里程计,并基于Kalman滤波的思想对两种里程计信息进行校正融合,构建了一种适用于小场景下的二维单线激光里程计。经实验,该里程计在点云形状较完整的小场景下具有较高的定位精度和较好的鲁棒性。     

Illumination robust image transformations for feature-based SLAM using photometric and feature matches loss

Simultaneous localization and mapping (SLAM) technology becomes more and more important in robot localization. The purpose of this paper is to improve the robustness of visual features to lighting changes and increase the recall rate of map re-localization under different lighting environments by optimizing the image transformation model. An image transformation method based on matches and photometric error (name the method as MPT) is proposed in this paper, and it is seamlessly integrated into the pre-processing stage of the feature-based visual SLAM framework. The results of the experiment show that the MPT method has a better matching effect on different visual features. In addition, the image transformation module encapsulated by a robot operating system (ROS) can be used with multiple visual SLAM systems and improve its re-localization effect under different lighting environments.     

基于激光点云NDT特征的两步回环检测

针对传统激光雷达配准算法进行大规模同时定位与建图(SLAM)时,存在较大累计误差的问题,本文提出一种基于正态分布变换(NDT)的两步回环检测方法,充分利用NDT配准中点云均值与方差特征,并将所提方法加入SLAM完整框架。点云匹配中采用重叠网格,首先根据各网格特征值,构建点云外观描述,进行粗回环检测。符合粗回环条件后,计算点云网格均值到坐标原点距离的方法使点云具有旋转不变性,进行精确回环检测。本文提出算法在“小旋风”智能车平台进行验证,实验表明,所提算法可以有效减小大规模建图中的累计误差,系统的鲁棒性更强,跟踪性能更好。     

Mobile Robot Exploration in Indoor Environment Using Topological Structure with Invisible Barcodes

This paper addresses the localization and navigation problem in the movement of service robots by using invisible two dimensional barcodes on the floor. Compared with other methods using natural or artificial landmarks, the proposed localization method has great advantages in cost and appearance since the location of the robot is perfectly known using the barcode information after mapping is finished. We also propose a navigation algorithm which uses a topological structure. For the topological information, we define nodes and edges which are suitable for indoor navigation, especially for large area having multiple rooms, many walls, and many static obstacles. The proposed algorithm also has the advantage that errors which occur in each node are mutually independent and can be compensated exactly after some navigation using barcodes. Simulation and experimental results were performed to verify the algorithm in the barcode environment, showing excellent performance results. After mapping, it is also possible to solve the kidnapped robot problem and to generate paths using topological information.     

An improved trajectory tracking decentralized adaptive controllerfor robot manipulators

A decentralized adaptive nonlinear controller for a robot manipulator is presented in this paper. Based on the promising results obtained by the decentralized adaptive PID control algorithms proposed by Seraji and other researchers, the authors redesigned the Lyapunov function, and as a result, achieved a further simplification of the control algorithm and better trajectory tracking performance. The main advantages of the proposed controller over similar controllers are the considerably faster convergence of tracking error, relatively simpler structure, and smoother control activity. Another advantage of this controller is that it only requires local position and velocity measurements, and it does not make use of the exact centralized mathematical model of the robot manipulator. Finally, the authors demonstrate through computer simulation the robustness of their controller against parameter variations and disturbances     

一种新型的INS与UWB融合的室内行人跟踪方法

针对单一室内定位系统定位精度低、鲁棒性差的问题,提出了一种新型的基于动态鲁棒容积卡尔曼滤波的超宽带(Ultra-wideband,UWB)与惯性导航系统(Inertial Navigation System,INS)融合的定位方法.首先建立了一种易于实现的UWB-INS融合定位框架,然后提出了一种动态鲁棒容积卡尔曼滤波算法以处理多源数据的融合.提出的滤波算法可将M估计理论、强跟踪算法、动态增强策略与传统的容积卡尔曼滤波算法结合,以此缓解外界噪声和系统模型误差对状态估计的不利影响.在UWB-INS组合定位框架内采用动态鲁棒容积卡尔曼滤波,可实现对室内行人运动轨迹的精确稳定跟踪.实际数据测试和Matlab仿真验证了所提方法在复杂环境下其定位精度和鲁棒性均优于单一依赖UWB或INS技术的定位系统.     

基于激光雷达SLAM的三维点云自适应算法

SLAM(Simultaneously Localization And Mapping)同步定位与地图构建作为移动机器人智能感知的关键技术。但是,大多已有的SLAM方法是在静止环境下实现的,当环境中存在移动频繁的障碍物时,SLAM建图会产生运动畸变,导致机器人无法进行精准的定位导航。同时,激光雷达等三维扫描设备获得的三维点云数据存在着大量的冗余三维数据点,过多的冗余数据不仅浪费大量的存储空间,同时也影响了各种点云处理算法的实时性。针对以上问题,本文提出一种SLAM运动畸变去除方法和一种基于曲率的点云数据分类简化框架。它通过激光插值法优化SLAM运动畸变,将优化后的点云数据分类简化。它能在提高SLAM建图精度,同时也很好的消除三维点云数据中特征不明显区域的冗余数据点,大大提高计算机运行效率。     

基于在线估计的视觉SLAM低光照图像增强算法

为了提升基于特征点的双目视觉定位算法在低光照环境下定位的准确性,提出一种基于在线估计的视觉同步定位与地图构建(simultaneous localization and mapping,SLAM)低光照图像增强算法.通过在线估计图像亮度值,实时更新图像增强算法的参数,解决了基于固定参数的图像增强算法在图像较亮、较暗等情况下的不适用性问题.首先,通过ORB-SLAM2系统寻找定位准确度的影响因素,并通过在线估计参数的方法实时更新相关参数.其次,利用低光照图像增强算法(low-light image enhancement,LIME)改善图像效果.最后,根据增强后的图像进行特征点提取,提升了特征匹配准确度,进而提升了定位的准确度.在公开EuRoC数据集上,通过与目前广泛使用的ORB-SLAM2算法进行对比实验,结果表明本文提出的视觉SLAM系统,具有更好的定位准确性及鲁棒性.     

Dead-reckoning sensor system and tracking algorithm for 3-D pipeline mapping

A dead-reckoning sensor system and a tracking algorithm for 3-D pipeline mapping are proposed for a tap water pipeline for which the diameter is small and the inner surface is rough due to pipe scales. The goals of this study are to overcome the performance limitations of small and low-grade sensors by combining various sensors with complementary functions and achieve robustness against a severe environment. A dead-reckoning sensor system consists of a small, low-cost micro electromechanical system inertial measurement unit (MEMS IMU) and an optical navigation sensor (used in laser mice). A tracking algorithm consists of a multi-rate extended Kalman filter (EKF) to fuse redundant and complementary data from the MEMS IMU and the optical navigation sensor and a geometry compensation method to reduce position estimation error using the end point of the pipeline. Two sets of experimental data have been obtained by driving a radio-controlled car equipped with the sensor system in a 3-D pipeline and on asphalt pavement. Our study can be used to estimate the path of a 3-D pipeline or mobile robots.     

Neural network impedance force control of robot manipulator

The performance of an impedance controller for robot force tracking is affected by the uncertainties in both the robot dynamic model and environment stiffness. The purpose of this paper is to improve the controller robustness by applying the neural network (NN) technique to compensate for the uncertainties in the robot model. NN control techniques are applied to two impedance control methods: torque-based and position-based impedance control, which are distinguished by the way of the impedance functions being implemented. A novel error signal is proposed for the NN training. In addition, a trajectory modification algorithm is developed to determine the reference trajectory when the environment stiffness is unknown. The robustness analysis of this algorithm to force sensor noise and inaccurate environment position measurement is also presented. The performances of the two NN impedance control schemes are compared by computer simulations. Simulation results based on a three-degrees-of-freedom robot show that highly robust position/force tracking can be achieved in the presence of large uncertainties and force sensor noise     

复杂场景下基于多特征融合的视频跟踪

为了解决常见视频跟踪方法在复杂场景中难以有效跟踪运动物体的难题,研究了在粒子滤波框架下基于多特征融合的判别式视频跟踪算法.首先分析了特征提取和跟踪算法的鲁棒性和准确性的关系,指出融合多种特征能有效地提升算法在复杂场景中的跟踪效果,然后选择提取HSV颜色特征和HOG特征描述目标表观,并在线训练逻辑斯特回归分类器构造判别式目标表观模型.在公开的复杂场景视频进行测试,比较了使用单一特征和多种特征的实验效果,并且将所提算法和经典跟踪算法进行了比较,实验结果表明融合多种特征的视频跟踪更具鲁棒性和准确性.     

Fast finite-time tracking control for a 3-DOF cable-driven parallel robot by adding a power integrator✰

Cable-driven parallel robots (CDPRs) usually suffer from kinematic and dynamic uncertainties, which makes it difficult for traditional trajectory tracking control algorithms to achieve high precision, fast response time, and robustness. In this study, we present a novel fast finite-time tracking control (FFTTC) algorithm which solves this problem to a large extent. Specifically, we firstly used a function of exponential errors with fractional power combined with API technique, to deal with the key difficulty of the convergence rate degradation which exists in traditional finite-time tracking control (TFTTC) when system states are far from the equilibrium point. Simultaneously, the API technique was used to avoid the problem of the explosion of complexity. To facilitate algorithm evaluation, the finite-time stability of the close system consisting of the proposed FFTTC algorithm and the CDPRs was proved mathematically and the settling time was estimated correspondingly. The trajectory tracking experiments were performed on a 3-DOF CDPR driven by 4 cables. Simulation and experimental results show that the proposed FFTTC algorithm can cope with external disturbances, variable load, and inaccurate model parameters. The comparison experiment indicates that the proposed FFTTC algorithm is superior to the model predictive control and TFTTC algorithms in precision, response speed and robustness.     

Adaptive neural network based controller for robots

A new Adaptive Neural Network (ANN) controller for robot trajectory trackingproblem is developed. A novel and efficient training algorithm for the proposed controller ispresented in this paper. The proposed training algorithm is based on updating the weights of thenetwork each step by minimizing the quadrant tracking errors and their derivatives.A simulation study is carried out on a polar robot manipulator to assure the effectivenessof the proposed trajectory tracking robot control system. The effects of the new controllerparameters and noisy external load disturbances on the control performance are studied. Thesimulation results of the proposed adaptive ANN controller are compared with those of aconventional ANN controller. The obtained results assured the robustness of the proposed ANNcontroller for: (i) uncertainties of the robot arm dynamic model and/or parameters, (ii) variousnoisy external load disturbances. Also, the simulation results assure the effectiveness of theproposed adaptive ANN controller against the conventional ANN one.     

Mobile Robot Navigation Using Wireless Sensor Networks Without Localization Procedure

In this paper, algorithms for navigating a mobile robot through wireless sensor networks are presented. The mobile robot can navigate without the need for a map, compass, or GPS module while interacting with neighboring sensor nodes. Two navigation algorithms are proposed in this paper: the first uses the distance between the mobile robot and each sensor node and the second uses the metric calculated from one-hop neighbors’ hop-counts. Periodically measuring the distance or metric, the mobile robot can move toward a point where these values become smaller and finally come to reach the destination. These algorithms do not attempt to localize the mobile robot for navigation, therefore our approach permits cost-effective robot navigation while overcoming the limitations of traditional navigation algorithms. Through a number of experiments and simulations, the performance of the two proposed algorithms is evaluated.