移动机器人的一种室内自然路标定位法

实时定位是移动机器人导航的一个基本前提。该文提出了一种利用墙棱边及墙平面(EdgeandPlane,EP)路标或者广义EP路标进行定位的方法,使用了异步数据融合的方法对移动机器人进行了定位。仅利用传感器对墙平面的距离数据进行测量就实现了机器人的定位。仿真显示这些方法能减少机器人的定位时间,提高对传感器测量数据的利用效率。     

基于单目视觉的移动机器人全局定位

提出在基于单目视觉创建的环境地图中实现移动机器人全局定位．基于KD树的最近邻搜索实现特征匹配．应用尺度不变特征变换(SIFT)方法提取特征,并用多维向量描述,保证了对图像光强变化、尺度缩放、三维视角和噪声具有不变性．提出了一种基于RANSAC的鲁棒定位方法．在实际室内环境Pioneer3机器人上进行的实验表明本文提出方法高效、可靠．     

Localization of an Autonomous Mobile Robot Based on Ultrasonic Sensory Information

Based on ultrasonic sensory information, an approach is proposed for localization of autonomous mobile robot (AMRs). In the proposed method, it will be proven that the combination of three ultrasonic transmitters and two receivers can determine both the position and the orientation of an AMR with respect to a reference frame uniquely. In this manner, since only ultrasonic sensors are used, the proposed method will be highly cost-effective and easy to implement. To show the validity and feasibility of the proposed method, the hardware configuration and a series of experiments will be given for illustration.     

Training a Vision Guided Mobile Robot

This paper presents the design, implementation and evaluation of a trainable vision guided mobile robot. The robot, CORGI, has a CCD camera as its only sensor which it is trained to use for a variety of tasks. The techniques used for training and the choice of natural light vision as the primary sensor makes the methodology immediately applicable to tasks such as trash collection or fruit picking. For example, the robot is readily trained to perform a ball finding task which involves avoiding obstacles and aligning with tennis balls. The robot is able to move at speeds up to 0.8 ms^-1 while performing this task, and has never had a collision in the trained environment. It can process video and update the actuators at 11 Hz using a single $20 microprocessor to perform all computation. Further results are shown to evaluate the system for generalization across unseen domains, fault tolerance and dynamic environments.     

Training a Vision Guided Mobile Robot

This paper presents the design, implementation and evaluation of a trainable vision guided mobile robot. The robot, CORGI, has a CCD camera as its only sensor which it is trained to use for a variety of tasks. The techniques used for train ing and the choice of natural light vision as the primary sensor makes the methodology immediately applicable to tasks such as trash collection or fruit picking. For example, the robot is readily trained to perform a ball finding task which involves avoiding obstacles and aligning with tennis balls. The robot is able to move at speeds up to 0.8 ms^-1 while performing this task, and has never had a collision in the trained environment. It can process video and update the actuators at 11 Hz using a single $20 microprocessor to perform all computation. Further results are shown to evaluate the system for generalization across unseen domains, fault tolerance and dynamic environments.     

基于递归卷积神经网络的移动机器人定位算法

移动机器人定位已成为机器人研究的重要任务。提出基于递归卷积神经网络的移动机器人定位（Recurrent Convolutional Neural Networks-Based Mobile Robot Localization,RCNN-MRL）算法。递归卷积神经网络（Recurrent Convolutional Neural Networks,RCNN）结合卷积神经网络（Convolutional Neural Networks,CNN）和递归神经网络（Recurrent Neural Networks,RNN）的特性,并依据机器人上嵌入的照相机拍摄的第一人称视角图像,RCNN-MRL算法利用RCNN实现自主定位。具体而言,先通过RCNN有效地处理多个连续图像,再利用RCNN作为回归模型,进而估计机器人位置。同时,设计双轮机器人移动,获取多个时间序列图像信息。最后,依据双轮机器人随机移动建立仿真环境,分析机器人定位性能。实验数据表明,提出的RCNN模型能够实现自主定位。     

Robust Practical Point Stabilization of a Nonholonomic Mobile Robot Using Neural Networks

A control structure that makes possible the integration of a kinematiccontroller and a neural network (NN) computed-torque controller fornonholonomic mobile robots is presented. A combined kinematic/torque controllaw is developed and stability is guaranteed by Lyapunov theory. Thiscontrol algorithm is applied to the practical point stabilization problemi.e., stabilization to a small neighborhood of the origin. The NN controllercan deal with unmodeled bounded disturbances and/or unstructured unmodeleddynamics in the vehicle. On-line NN weight tuning algorithms that do notrequire off-line learning yet guarantee small tracking errors and boundedcontrol signals are utilized.     

基于场景识别的移动机器人定位方法研究

提出了一种基于场景识别的移动机器人定位方法．对CCD采集的工作环境的系列场景图像,用多通道Gabor 滤波器提取场景图像的全局纹理特征,然后通过SVM分类器来识别场景图像,实现机器人的逻辑定位．在移动机器人CASIA-I 上对该算法进行了实验．实验结果表明,该定位方法可达到91.11%的定位准确率,对光照、对比度等因素有较强的鲁棒性,并且满足机器人实时定位的要求．     

改进强化学习算法应用于移动机器人路径规划

为了解决传统的强化学习算法应用于移动机器人未知环境的路径规划时存在收敛速度慢、迭代次数多、收敛结果不稳定等问题,提出一种改进的Q-learning算法。在状态初始化时引入人工势场法,使得越靠近目标位置状态值越大,从而引导智能体朝目标位置移动,减少算法初始阶段因对环境探索产生的大量无效迭代;在智能体选择动作时改进[ε]-贪婪策略,根据算法的收敛程度动态调整贪婪因子[ε],从而更好地平衡探索和利用之间的关系,在加快算法收敛速度的同时提高收敛结果的稳定性。基于Python的Tkinter标准化库搭建的格栅地图仿真结果表明,改进的Q-learning算法相较于传统算法在路径规划时间上缩短85.1%,收敛前迭代次数减少74.7%,同时算法的收敛结果稳定性也得到了提升。     

A New Low Cost System for Autonomous Robot Heading and Position Localization in a Closed Area

A low cost system for the localization of mobile indoor robots is presented. The system is composed of an emitter located on a wall and a receptor on top of the robot. The emitter is a laser pointer acting like a beacon, and the receptor is a cylinder made out of 32 independent photovoltaic cells. The robot's position and orientation are obtained from the moments when the laser crosses each cell.     

An Active System for Three-Dimensional Localization of Mobile Robots

This paper presents an active system, which is composed of a laser range finder and four artificial reflectors, for the three-dimensional (3D) localization of a mobile robot. In this system, it will be proved that the position and the orientation of a mobile robot in a 3D space with respect to a reference frame can be determined provided that the four artificial reflectors are not installed in the same plane. Since the artificial reflectors cannot be treated as points in practice, the proposed localization procedure will be formulated as a nonlinear programming problem to account for actual sizes of the artificial reflectors. To show the validity and feasibility of the proposed method, a series of experiments will be given for illustration.     

Real Time Map Building by Means of an Ellipse Spatial Criterion and Sensor-Based Localization for Mobile Robot

In the present paper, map building and localization problems are examined. A new algorithm is proposed for each task. The map building algorithm is based on measurements derived by ultrasonic sensors. It has small memory requirements and can distinguish between parallel edges and edges-corners. The discrimination between edges and corners is achieved based on the physical restrictions imposed by the ultrasonic sensors and the way they are fired. The parallel edges discrimination is carried out on the basis of an ellipse spatial criterion. It is of low computational complexity and, therefore, can be applied on line. The application of the proposed algorithm does not require tracking of a continuous path. The method is accurate. It converges to the existing map characteristics. The mean inclination error is equal to 0.78 degrees while the mean distance error of the mid point of the chartographed edges from the actual edges is equal to 4.11 cm. The robustness of the algorithm was verified by applying it in noisy environments. The localization algorithm reduces the accumulated odometry error by utilizing readings obtained from ultrasonic sensors. Its application does not require a priori knowledge of the statistical characteristics of the noise that affects the measurements, nor the exact robot starting position. It is of low computational complexity. The application of the method was tested on a track with length equal to 25.85 m. The experiment was repeated 50 times. The mean position error was equal to 26 cm, while a dramatic reduction of the mean position error was achieved in regions for which the detected obstacles were parallel to only one of the reference coordinate axes. The mean error in such regions was reduced to 30 cm from 48 cm. The mean heading error was equal to 6.8°.     

CAD-Vision-Range-Based Self-Localization for Mobile Robot Using One Landmark

Localization is the process of determining the robot's posture within its environment including its current position and heading direction (or orientation). The process is of utmost importance for the autonomous navigational functions of a service robot. This paper describes a new localization method for service robots operating in a building based on a CAD model of the indoor environment in reasonable details. Only one specific landmark pasted within a specific region on the wall is needed. The camera with pan/tilt/zoom functions mounted on the robot first searches for this identification landmark and starts to conduct measurements using a laser rangefinder. With the polar coordinates of few measurement points on the wall and an accurate local CAD model, the exact position and orientation of the robot can be identified. This method has five distinctive advantages. First, the position of the landmark does not need to be precise. Second, each localization exercise is independent and no previous history of the moving track of the robot is required but the computational speed is still high. Third, the method is very robust with good fault-tolerance because it makes use of the reliable Hough transform. Fourth, the resolution is automatically adjusted because the panning resolution of the camera is based on the first effective measurement representing the distance of the robot from the landmark. Fifth, only the local CAD model of the room at the vicinity of the landmark needs a high precision because this model is used for localization. The system does not demand a highly accurate CAD model of the entire built environment. CAD models at other places are for navigation and path planning only.     

室内移动机器人的视觉定位方法研究

针对地图未知的室内环境下的定位问题,提出了一种基于特征跟踪的视觉里程计方法．利用单目摄像头提取和跟踪环境特征点集,进而根据观测模型利用扩展卡尔曼滤波算法估算出机器人的位姿．办公室环境中的定位实验证明了方法的有效性．     

嵌入式移动机器人红外路标定位模块研究

针对室内环境移动机器人的自定位问题,提出一种嵌入式移动机器人红外路标定位模块。采用基于单应矩阵的初始标定算法和陒机初始标定方法,补偿由于实际使用中的安装误差所引起的定位偏差。实验结果表明,该模块易于嵌入式系统实现,定位模块位置精度可达厘米级别,角度定位精度雓于6°。     

基于直线和单特征点的移动机器人视觉推算定位法

提出了一种基于直线和单特征点的视觉推算定位方法．将摄像机固定安装在移动机器人上,并使摄像机的方向垂直于天花板．提取天花板图像中的两条相交直线及其交点作为特征,利用直线的方向估计机器人的方位,再结合相交点的像素坐标变化量推算出机器人的位置．机器人运动过程中,用作特征的交点接近图像边缘时,更换新的交点作为特征点．室内定位实验结果验证了该方法的有效性．     

基于运动视的移动机器人定位方法

提出了一种新颖的基于运动视的定位方法．该方法从机器人接近的目标上选取两个特征点并根据它们的图像坐标确定机器人运动前后相对于目标的位姿．本文还提出了一种能够更准确地找出特征点图像坐标的搜索算法．试验结果表明该方法有较高的定位精度和较强的鲁棒性．     

采用双重采样的移动机器人Monte Carlo定位方法

移动机器人Monte Carlo定位效率受限于大量粒子的权值更新运算. 本文提出一种实现粒子集规模自适应调整的双重采样方法: 第一层基于粒子权重的固定粒子数重采样, 有效减轻粒子权值退化并保证预测阶段粒子多样性; 第二层粒子稀疏化聚合重采样, 基于粒子空间分布合理性将粒子加权聚合, 从而减少参与权值更新粒子数. 该方法通过提高粒子预测能力保证滤波精度, 通过减少权值更新运算提高了粒子滤波效率. 仿真实验表明, 双重采样方法能够有效实现粒子集规模自适应调整,采用双重采样的移动机器人Monte Carlo定位方法是高效、鲁棒的.     

Monocular Vision for Mobile Robot Localization and Autonomous Navigation

This paper presents a new real-time localization system for a mobile robot. We show that autonomous navigation is possible in outdoor situation with the use of a single camera and natural landmarks. To do that, we use a three step approach. In a learning step, the robot is manually guided on a path and a video sequence is recorded with a front looking camera. Then a structure from motion algorithm is used to build a 3D map from this learning sequence. Finally in the navigation step, the robot uses this map to compute its localization in real-time and it follows the learning path or a slightly different path if desired. The vision algorithms used for map building and localization are first detailed. Then a large part of the paper is dedicated to the experimental evaluation of the accuracy and robustness of our algorithms based on experimental data collected during two years in various environments.