基于图像识别的移动机器人自动完成充电

移动机器人正在慢慢普及到我们的生活当中,而移动机器人在移动过程中消耗电能,当电能耗尽时需要人工为移动机器人进行充电,提出一种基于图像识别的方法,让机器人自动完成充电。移动机器人通过对图像的纹理分析和对兴趣点的快速匹配SIFT算法完成移动机器人识别电源插座,实现自动充电。本文的方法充分考虑到由于移动机器人的灵活多变性对所采集到的图像造成的影响,如图像旋转,图像周围有遮挡物体等,用纹理分析与SIFT算法识别插座,当所采集的图像与机器人本身记忆的图像达到一定匹配度时,移动机器人将完成自动充电。     

一种家居监控机器人的角及平直线段匹配的组合定位方法

针对移动机器人家居环境下的定位问题,提出了一种结合平直线段匹配、角匹配和里程计的组合定位方法。该系统采用了Labview开发平台和CompactRIO控制器,得到了很好的实时性效果。机器人通过激光测距仪基于TCP／IP通讯协议得到环境点信息,由迭代适应点（IterativeEndPointFit,IEPF）算法得到环境线段,再由最小二乘法得到线段参数。在基于线段基础上,得到局部的平直线段和角特征,再与已知平直线段和角特征做匹配,通过平直线段和角匹配算法实时更新机器人位置和姿态。分析里程计定位、平直线段匹配定位和角匹配定位的误差,分配不同的权重得到优化的组合定位算法。实验表明：该组合定位算法定位稳定,位置误差在S0mm以内,角度偏差5。以内,循环的周期在120IllS以内。     

Position and Rotation Estimation for Mobile Robots in Outside of Recording Path Using Ego‐Motion

In this research, estimating the position and rotation of a mobile robot outside of a recording path is realized by applying ego‐motion to view‐based navigation. The ego‐motion is calculated based on the differences in 3D positions of SURF feature points between recording and current images obtained by a Kinect sensor. In conventional view‐based navigation, it is difficult to plan another path when people and objects are found in the recording path. By using the authors’ proposed estimation method, it is possible to realize flexible path planning in actual environments that include people and objects. Based on the results of experiments performed in actual indoor environments, the authors evaluated measurement accuracy for the robot's position and rotation estimated under their method, and confirmed the viability of their method for actual environments including people and objects.     

Stability Criterion of Biped Robot on Rough Terrain

This paper presents a stability criterion for biped robots on rough terrain. On rough terrain, which is not perpendicular to gravity, a robot may not fall even though its center of mass is out of the area of the supported polygon defined by the conventional VHP (virtual horizontal plane) method. This is because there exists a certain force opposing the torque that causes the robot to fall down. We propose a new measure for strictly judging the limit of falling considering all the forces acting on the contact points. Numerical simulations and experimental results are also shown, which verify the proposed method.     

基于改进深度强化学习的移动机器人路径规划

针对传统深度强化学习中移动机器人在稀疏奖励环境下只有在规定时间步内到达目标位置才能得到积极奖励,中间过程的每一步都是负面奖励的路径规划问题.提出了基于改进深度Q网络的路径规划方法,在移动机器人在探索过程中,对以真实目标为条件的轨迹进行采样,在经验回放过程中,把移动机器人已经到达的状态来代替真正的目标,这样移动机器人可以获得足够的积极奖励信号来开始学习.通过深度卷积神经网络模型,将原始RGB图像作为输入,通过端对端的方法训练,利用置信区间上界探索策略和小批量样本的方法训练神经网络参数,最后得到上、下、左、右4个动作的Q值.在相同的仿真环境中结果表明,该算法提升了采样效率,训练迭代更快,并且更容易收敛,避开障碍物到达终点的成功率增加40％左右,一定程度上解决了稀疏奖励带来的问题.     

Robotics‐based telepresence using multi‐modal interaction for individuals with visual impairments

Although many assistive devices have been developed and utilized to aid in daily living, the most general assistive means for individuals with visual impairments are the walking cane and guide dogs. These assistive means are effective in assisting the user in navigating within an environment; however, the navigation space is limited to the proximal environment of the user. Thus, in this paper, we discuss a method to increase the range of accessibility to a remote environment through robotic embodiment that enables teleoperation and teleperception through multi‐modal feedback. In order to transform remote spatial information into a non‐visual modality, we present a framework for utilizing an RGB‐D‐based depth camera, a mobile robot, and a haptic interface for 3D haptic rendering to accomplish the goal of haptic exploration of a remote environment. Experiments with three different control methods for robot interaction are designed for users with and without visual impairments. Several hypothesis are built to study the correlation between control/feedback modality and performance in telerobotic operations. Results show that users performed best when combining semi‐autonomous navigation with 3D haptic exploration and also rated their experience with our system as fairly good. Copyright © 2014 John Wiley & Sons, Ltd.     

Control of Autonomous Mobile Robot Using Map Matching with Optimized Search Range

Accurate estimation of self‐position is indispensable for autonomous mobile robots. Map‐matching‐based self‐positioning is a promising method; however, adequate matching area selection is an issue to be solved. In this paper, we propose an adjustable map matching method that remedies a defect of the conventional template matching technology. The proposed method was evaluated experimentally, and the results confirmed the advantages of the proposed method.     

自适应网络结构的故障行波定位方法

输电网中环路的存在会影响故障初始行波到达时间与行波传输路径的匹配,影响故障点位置的计算.为此,提出了一种自适应网络结构的故障行波定位方法.对于故障线路不包含于环路的情况,利用Floyd算法搜索故障点到网络中各节点的最短路径;对于故障线路包含于环路的情况,确定环路中初始行波到达时间最大的节点为解环点,并提出了剔除无效路径的时间判据,将复杂故障网络简化为辐射形网络.在辐射形网络中,运用基于整个电网的故障行波定位算法,实现全网综合定位.仿真结果表明,该方法实现简单,适用于任何结构的电网,能够有效弥补现有行波网络定位算法的缺陷和不足.     

一种无需参数整定的混合线路故障行波定位方法

为解决混合线路故障行波定位中行波波速与线路长度参数整定的难题,论文分析了混合线路中故障行波的传输特性,提出了一种无需参数整定的故障行波定位方法。该方法在线路不同位置设置多个模拟故障点,依次试验测试或仿真测试线路上各个模拟故障点产生的故障行波传输到该线路首端与末端的时间差,构建基准行波时差数组;当线路实际发生故障时,由行波采集装置测量行波信号到达线路首端与线路末端的时间差,与基准行波时差数组中各元素比较,判别故障区段;并计算故障点在故障区段线路的相对位置,根据各杆塔或电缆接头的位置确定故障点准确位置。仿真结果表明该方法无需整定线路长度和行波波速,且定位结果不受线路弧垂的影响,具有算法简单、实用性强、定位精度高的优点。     

一种新颖的虹膜识别算法

提出一种新颖的虹膜识别算法.该算法利用二维Log-Gabor滤波器对归一化后的虹膜图像进行滤波,滤波后的中高频子带图像被分割为若干子块,提取各子块局部极大值点并对其进行相位粗量化编码来生成虹膜特征码;特征匹配时使用虹膜特征码和干扰掩模来生成汉明距离,以此衡量两个虹膜之间的相似程度.该方法具有平移、放缩和旋转不变性,并且生成的虹膜特征码仅288 B.实验结果和数据分析表明该方法的合理性和可行性.     

Simultaneous design of the sensory morphology and controller of mobile robots

This paper proposes a method for automatic design of the sensory morphology of a mobile robot. The proposed method employs two types of adaptations, ontogenetic and phylogenetic, to optimize the sensory morphology of the robot. In ontogenetic adaptation, reinforcement learning searches for the optimal policy, which is highly dependent on the sensory morphology. In phylogenetic adaptation, a genetic algorithm is used to select morphologies with which the robot can learn tasks faster. Our proposed method was applied to the design of the sensory morphology of a line‐following robot. We performed simulation experiments to compare the design solution with a hand‐coded robot. The results of the experiments revealed that our robot outperformed the hand‐coded robot in terms of the following accuracy and learning speed, although our robot had fewer sensors than the hand‐coded one. We also built a physical robot using the design solution. The experimental results revealed that this physical robot used its morphology effectively and outperformed the hand‐coded robot. © 2010 Wiley Periodicals, Inc. Electr Eng Jpn, 172(1): 48–57, 2010; Published online in Wiley InterScience ( www.interscience.wiley.com ). DOI 10.1002/eej.20965     

Adaptive‐search template matching technique based on vehicle acceleration for monocular visual odometry system

One of the important tasks of an autonomous mobile vehicle is the reliable and fast estimation of its position over time. This paper presents the development of an adaptive technique to hasten and improve the quality of correlation‐based template matching for monocular visual odometry systems that estimate the relative motion of ground vehicles in low‐textured environments. Moreover, the factors that can affect the maximum permissible vehicle driving speed were determined and the related equations were derived. The developed system uses a single downward‐facing monocular camera installed at an optimum location to avoid the negative effect of directional sunlight and shadows which can disturb the correlation. In addition, the normalized cross‐correlation method is implemented to calculate the pixel displacement between image frames. Although this method is highly effective for template matching because of its invariance to linear brightness and contrast variations, it incurs high computational cost. Thus, the optimal sizes of image template and matching search area are selected and their locations are dynamically changed according to vehicle acceleration, in order to achieve a compromise between the performance and the computational cost of correlation. The proposed technique increases the allowable vehicle driving speed and reduces the probability of template false‐matching. Moreover, compared to traditional full search matching techniques, the adaptive technique demonstrates high efficiency and accuracy and improves the quality and speed of the correlation with more than 87% of reduction in computational cost. © 2016 Institute of Electrical Engineers of Japan. Published by John Wiley & Sons, Inc.     

Dynamic optimal route search algorithm for car navigation systems with preferences by dynamic programming

Optimal route search to the destination is one of the most important functions of car navigation devices. The development of road traffic infrastructure has made it possible to receive real‐time information of the traffic situation. Route search algorithms for car navigation devices make use of this information to avoid the traffic congestions. Such algorithms should find the new optimal route efficiently when the traffic situation changes. Usually, the minimum traveling time or distance is considered to define the optimal route. However, the minimum traveling time or distance is not always what the user is looking for. The user may prefer to travel on a certain route even at the cost of traveling time or distance. Car navigation devices should consider such preferences when finding the optimal route. In this paper, we propose a dynamic programming algorithm to find the optimal route considering that it should deal with the changes of the traffic situation and multiple criteria. The proposed method uses the information from the previous computation to find the new optimal route considering user preferences when the traveling time of the road section changes. The proposed method was applied to a real road network to find the optimal route. Results show that the proposed method can find the user‐preferred optimal route. Simulation results also show better calculation time of the proposed method compared to the Dijkstra algorithm. Copyright © 2010 Institute of Electrical Engineers of Japan. Published by John Wiley & Sons, Inc.     

Development and Control of Multi‐Degree‐of‐Freedom Mobile Robot for Acquisition of Road Environmental Modes

The acquisition of information about the environment around a mobile robot is important for purposes such as controlling the robot from a remote location and in situations such as that when the robot is running autonomously. In some research, audiovisual information has been used. However, the acquisition of information about force sensation, which is included in environmental information, has not been well investigated. The mobile‐hapto, which is a remote control system with force information, has been proposed, but the robot used for the system can acquire only the horizontal components of forces. For this reason, in this research a three‐wheeled mobile robot consisting of seven actuators was developed and a control system was created for it. The robot can acquire information about horizontal and vertical forces without using force sensors. By using this robot, detailed information on the forces in the environment can be acquired and the operability of the robot and its ability to adjust to the environment are expected to improve.     

基于改进RBPF的变电站巡检机器人建图方法研究

为解决传统粒子滤波(Rao-Blackwellized Particle Filter, RBPF)巡检机器人建图方法在非结构化环境中计算精度低、计算量大等问题,文章提出一种基于点云匹配的改进RBPF变电站巡检机器人建图方法。基于RBPF方法设计子图构建策略,引入Adaboost学习算法识别相邻子图;针对传统点云匹配方法在拼接仅有部分重合的子图时,拼接精度低的现象,将NDT和ICP算法相结合,设计匹配算法,求解子图间相对位姿。然后,根据子图间相对位姿,通过图优化(General Graph Optimization, G2O)算法对子图全局位姿进行优化求解,得到完整的变电站地图;采用Gazebo软件进行仿真,对不同方法的建图效果进行对比。结果表明：改进RBPF方法在变电站的复杂环境下能够降低硬件成本、提高建图精度,可以为无人值守变电站的设计提供参考。     

一种基于改进人工势场法的局部路径规划算法

针对在存在复杂障碍物的环境中,利用人工势场法进行移动机器人路径规划时出现的局部最小值和目标不可达问题。本文提出了一种基于凹形障碍物补齐的改进人工势场法进行局部路径规划。首先,通过对凹形障碍物补齐,防止机器人进入局部最小值区域。然后,通过新增距离影响因子,改进了斥力场函数,使目标点成为全局势场中的最小点,防止机器人陷入目标不可达区域。最后仿真结果表明,本文所提出的改进人工势场法可以解决存在复杂障碍物的环境中的局部最小值问题和目标不可达问题,并且相对于其他算法,可以有效减少路径补偿,提高规划效率。     

Adaptive control of an underactuated spherical robot with a dynamic stable equilibrium point using hierarchical sliding mode approach

Unlike the traditional underactuated mechanical system, the equilibrium point of the inner swing‐up suspension of a spherical robot possesses stable and dynamic characteristics. The stable dynamic equilibrium position of the underactuated suspension must agree well with the uncertain rolling resistance in practice. To overcome this special underactuated problem with dynamic equilibrium, we present an adaptive control scheme to enhance the ground adaptability using hierarchical sliding mode approach. The control method can drive the spherical robot to a desired position, and at the same time, the suspension subsystem holds on their dynamic equilibrium points rapidly. The proposed control approaches are verified to be effective by numerical simulation for a spherical robot to achieve satisfactory control performances, such as ground adaptability, rapid convergence, and disturbance rejection. Copyright © 2013 John Wiley & Sons, Ltd.     

基于差值矩阵的多端输电线路故障定位研究

针对现有多端输电线路故障行波检测困难、定位精度不高、判定算法复杂等问题,提出一种基于变分模态分解(VMD)算法和Hilbert变换相结合的行波检测法和一种基于差值矩阵的多端输电线路故障行波定位算法。首先通过行波传感器采集故障行波信号,利用VMD分解算法对故障行波进行分解,结合Hilbert变换提取模态分量IMF1的瞬时频率,根据第一个瞬时频率的奇异点位置确定故障行波的达到时刻。然后利用行波到达各端的时间和行波传输原理,得到多端输电线路故障分支判定矩阵。最后根据故障分支判定矩阵确定故障支路,实现故障点的精确定位。ATP/EMTP仿真结果表明,所提检测方法能够准确检测故障初始行波的到达时间,多端输电线路定位算法能够准确判定故障支路,相比于HHT检测方法下的定位算法,进一步提高了定位精度。     

基于终端滑模控制的移动机器人轨迹跟踪

针对移动机器人受到外部不规则扰动时易出现速度和位姿误差跳变的问题,提出了一种基于扰动观测器的动态终端滑模控制方法。结合运动学模型,利用李雅普诺夫法设计虚拟控制器,进一步设计非奇异动态终端滑模轨迹跟踪控制器。为减小外部扰动对系统的影响,设计非线性扰动观测器对控制器进行扰动补偿。最后,将本文所提方法与自适应滑模控制方法进行仿真比较。结果显示,在第15 s扰动发生阶跃变化时,自适应滑模控制方法的线速度和角速度分别发生1.4 m/s和1.24 rad/s的跳变,而本文所提方法速度跳变幅度小于自适应滑模控制方法的1/10。仿真结果表明,本文所提方法可有效地抑制扰动对系统的影响,减小移动机器人速度和位姿误差的跳变幅度。     

视觉与二维激光雷达的目标检测方法

为了改进单一传感器对目标物体的检测范围小、检测特征少以及检测准确率较低的问题,提出一种视觉与二维激光雷达的目标检测方法。在视觉检测方面提出一种改进的GoogLeNet算法实现视觉对目标物体的识别,该方法相比GoogLeNet算法在对6种目标物体的识别准确率上提高了0.7%。在二维激光雷达检测方面采用欧氏聚类算法对二维激光雷达的点云数据聚类,接着使用RANSAC算法对聚类簇中的数据点进行筛选,最后使用卡尔曼滤波算法对目标物体的位置进行预测,实现二维激光雷达在特定平面上360°对目标物体进行跟踪检测和定位。实验结果表明,该方法使得移动机器人扩大了检测范围、增加了检测特征并提高了识别准确率。