有色量测噪声下机器人同步定位与地图构建

针对有色量测噪声模型,提出一种有色量测噪声下的轮式机器人同步定位与地图构建算法。通过重新组合轮式机器人的过程模型和量测模型,将有色量测噪声量测模型转化为虚拟的白噪声量测模型。为使过程噪声和量测噪声不相关,对过程模型进行不相关条件处理。算法按照构造的虚拟过程模型和量测模型进行滤波估计和地图构建。仿真结果验证了算法的一致性和鲁棒性。     

基于不确定性知识的实时道路场景理解

由于室外机器人的工作环境非常复杂，因此机器人的视觉导航必须具有足够的智能和鲁棒性，为此，提出了一种基于不确定性知识的实时道路理解算法，该算法通过不确定性知识推理来融合多种信息和知识，以满足在复杂道路环境下的鲁棒性要求，它即使在有强烈阴影、水迹等干扰下也能给出比较好的结果；通过图象边缘信息的提取可以得到精确的道路边界，以满足视觉导航的精确性要求；同时在算法设计时，兼顾了实时性要求；使得算法得以实时实现，该算法已在实际的机器人上进行了测试，并得到了很好的结果。     

Balance Control of a Biped Robot on a Rotating Platform Based on Efficient Reinforcement Learning

In this work, we combined the model based reinforcement learning (MBRL) and model free reinforcement learning (MFRL) to stabilize a biped robot (NAO robot) on a rotating platform, where the angular velocity of the platform is unknown for the proposed learning algorithm and treated as the external disturbance. Nonparametric Gaussian processes normally require a large number of training data points to deal with the discontinuity of the estimated model. Although some improved method such as probabilistic inference for learning control (PILCO) does not require an explicit global model as the actions are obtained by directly searching the policy space, the overfitting and lack of model complexity may still result in a large deviation between the prediction and the real system. Besides, none of these approaches consider the data error and measurement noise during the training process and test process, respectively. We propose a hierarchical Gaussian processes (GP) models, containing two layers of independent GPs, where the physically continuous probability transition model of the robot is obtained. Due to the physically continuous estimation, the algorithm overcomes the overfitting problem with a guaranteed model complexity, and the number of training data is also reduced. The policy for any given initial state is generated automatically by minimizing the expected cost according to the predefined cost function and the obtained probability distribution of the state. Furthermore, a novel Q(λ) based MFRL method scheme is employed to improve the policy. Simulation results show that the proposed RL algorithm is able to balance NAO robot on a rotating platform, and it is capable of adapting to the platform with varying angular velocity.      

Optimal landmark selection for triangulation of robot position

A mobile robot can identify its own position relative to a global environment model by using triangulation based on three landmarks in the environment. It is shown that this procedure may be very sensitive to noise depending on spatial landmark configuration, and relative position between robot and landmarks. A general analysis is presented which permits prediction of the uncertainty in the triangulated position.

In addition an algorithm is presented for automatic selection of optimal landmarks. This algorithm enables a robot to continuously base its position computation on the set of available landmarks, which provides the least noise sensitive position estimate. It is demonstrated that using this algorithm can result in more than one order of magnitude reduction in uncertainty.     

基于改进UKF算法的移动机器人定位方法研究

针对移动机器人在多传感器融合定位过程中因噪声统计特性未知或不准确引起的定位精度不高的问题,提出了一种基于Sage-Husa滤波改进的无损卡尔曼滤波(UKF)移动机器人定位算法。首先建立了移动机器人定位相关模型;然后根据噪声统计特性时变特点利用Sage-Husa中的噪声估计器,对状态噪声和量测噪声进行自适应地估计,减小扰动噪声给定位解算带来的误差;接着在状态更新时引入收敛因子,加快算法收敛速度;最后将UKF算法和改进的UKF算法应用到实验室移动机器人中进行仿真实验。实验结果表明,所提出的算法对状态扰动具有较强的抵制能力,对机器人定位的准确性与稳定性的提升具有显著效果。     

Dynamic localization with hybrid trilateration for mobile robots in intelligent space

In this paper, we propose a new localization algorithm based on a hybrid trilateration algorithm for obtaining an accurate position of a robot in intelligent space. The proposed algorithm is also able to estimate a position of the moving robot by using the extended Kalman filter, taking into consideration time synchronization and velocity of the robot. For realizing the localization system, we employ several smart sensors as beacons on the ceiling in intelligent space and as a listener attached to the robot. Finally, simulation results show the feasibility and effectiveness of the proposed localization algorithm compared with existing trilateration algorithms.     

Revolute manipulator workspace optimization: A comparative study

Robotic manipulators with three revolute families of positional configurations are very common in the industrial robots. The capability of a robot largely depends on the workspace of the manipulator apart from other parameters. In this work, an evolutionary optimization algorithm based on foraging behavior of Escherichia coli bacteria present in human intestine is utilized to optimize the workspace volume of a 3R manipulator. The proposed optimization method is subjected to some modifications for faster convergence than the original algorithm. Further, the method is also very useful in optimization problems in a highly constrained environment such as the robot workspace optimization. The test results are compared with standard results available using other optimization algorithms such as Differential Evolution, Genetic Algorithm and Particle Swarm Optimization. In addition, this work extends the application of the proposed algorithm to two different industrial robots. An important implication of this paper is that the present algorithm is found to be superior to other methods in terms of computational efficiency.     

基于深度Q网络的人群疏散机器人运动规划算法

针对公共场合密集人群在紧急情况下疏散的危险性和效果不理想的问题，提出一种基于深度Q网络（DQN）的人群疏散机器人的运动规划算法。首先通过在原始的社会力模型中加入人机作用力构建出人机社会力模型，从而利用机器人对行人的作用力来影响人群的运动状态；然后基于DQN设计机器人运动规划算法，将原始行人运动状态的图像输入该网络并输出机器人的运动行为，在这个过程中将设计的奖励函数反馈给网络使机器人能够在"环境-行为-奖励"的闭环过程中自主学习；最后经过多次迭代，机器人能够学习在不同初始位置下的最优运动策略，最大限度地提高总疏散人数。在构建的仿真环境里对算法进行训练和评估。实验结果表明，与无机器人的人群疏散算法相比，基于DQN的人群疏散机器人运动规划算法使机器人在三种不同初始位置下将人群疏散效率分别增加了16.41%、10.69%和21.76%，说明该算法能够明显提高单位时间内人群疏散的数量，具有灵活性和有效性。     

Path planning for an autonomous mobile robot considering a region with a velocity constraint in a real environment

Recently, many research projects and competitions have attempted to find an autonomous mobile robot that can drive in the real world. In this article, we consider a path-planning method for an autonomous mobile robot that would be safe in a real environment. In such a case, it is very important for the robot to be able to identify its own position and orientation in real time. Therefore, we applied a localization method based on a particle filter. Moreover, in order to improve the safety of such autonomous locomotion, we improved the path-planning algorithm and the generation of the trajectory so that it can consider a region with a limited maximum velocity. In order to demonstrate the validity of the proposed method, we participated in the Real World Robot Challenge 2010. The experimental results are given.     

A Localization Algorithm for Autonomous Mobile Robots via a Fuzzy Tuned Extended Kalman Filter

The capability to acquire the position and orientation of an autonomous mobile robot is an important element for achieving specific tasks requiring autonomous exploration of the workplace. In this paper, we present a localization method that is based on a fuzzy tuned extended Kalman filter (FT-EKF) without a priori knowledge of the state noise model. The proposed algorithm is employed in a mobile robot equipped with 16 Polaroid sonar sensors and tested in a structured indoor environment. The state noise model is estimated and adapted by a fuzzy rule-based scheme. The proposed algorithm is compared with other EKF localization methods through simulations and experiments. The simulation and experimental studies demonstrate the improved performance of the proposed FT-EKF localization method over those using the conventional EKF algorithm.     

Noise analysis of robot manipulator using neural networks

Due to a lot of robot manipulators application in industry, low noise degree is very important criteria for robot manipulator's joints. In this paper, joint noise problem of a robot manipulator with five joints is investigated both theoretically and experimentally. The investigation is consisted of two steps. First step is to analyze the noise of joints using a hardware and software. The hardware is a part of noise sensors. The second step; according to experimental results, some neural networks are employed for finding robust neural noise analyzer. Five types of neural networks are used to compare each other. From the results, it is noted that the proposed RBFNN gives the best results for analyzing joint noise of the robot manipulator.     

基于软件人架构的机器人设计与实现

以无人变电站移动巡检机器人为应用对象,运用"软件人"理论方法开展基于"软件人"架构的机器人系统设计:一方面,利用虚拟"软件人"为实体机器人建立软件模型,充分发挥"软件人"的拟人特性,构建"软件人"和机器人协同交互、融合共生的平行系统,通过"软件人"和机器人平行进化的智能算法实现"软件人"和机器人在线、实时、自动平行进化,在运行过程中不断提高机器人的性能;另一方面,利用"软件人"运行平台的高处理能力完成在机器人嵌入式系统中无法完成的一些智能算法,实现对实体机器人的实时作业并行、在线、实时地仿真。     

色标设计与辨识算法研究

在人工智能领域，足球机器人系统是一个非常好的应用与研究平台，它是一个由计算机视觉闭环的反馈控制系统，在足球机器人视觉子系统中，色标的设计是关系到系统辩识精度，实时性和抗干扰性的一个重要因素。通过对3种典型色标设计进行比较和分析，确定在一种比较合理的设计方案，并针对这种方案提出了一种快速准确的补偿逼近算法，取得了良好的辨识结果，该算法为基于视觉的集中控制式足球机器人系统的研究与开发提供了有力的保障。     

基于MLP和SPSO的机器人行为选择与运动控制方法

针对传统行为选择机制(ASM)不能很好地做出控制决策的问题,提出一种基于多层感知(MLP)前馈神经网络的ASM,并将其应用到移动机器人目标跟踪中。首先,根据具体应用场景预定义多个机器人行为。然后,根据机器人配备的图像和红外传感器获得的目标位置和障碍物信息,通过MLP神经网络从预定义行为中选择出所需执行的行为。另外,为了构造最优的MLP模型,采用一种简化粒子群算法(SPSO)来优化网络权值参数。机器人目标跟踪仿真的结果表明,提出的ASM能够准确选择出合适的行为,实现了控制机器人跟踪目标移动且能够避开各种障碍物。     

自主机器人的噪声影响及其消解策略的研究

该文针对典型的觅食任务,以计算机仿真为手段,直观地揭示噪声对机器人系统性能的影响.在此基础上,提出了以过程奖赏（process reward）代替传统的结果奖赏（result reward）,并与优先扫除（prioritized sweeping）的强化学习算法结合作为噪声消解策略.然后与基于结果奖赏的Q学习算法（Q-learning）等其它四种算法进行比较,结果表明基于过程奖赏和优先扫除的强化学习算法能显著降低噪声的影响,提高了系统整体性能.     

Acquisition of Weld Seam Dimensional Position Information for Arc Welding Robot Based on Vision Computing

Recognition and identification of weld environment and seam dimensional position by computer vision is a key technology for developing advanced autonomous welding robot. Aiming at requirements for recognition of weld seam image characteristics, this paper first presents an improved algorithm of subpixel edge detection based on Zernike moments. Comparing with the Ghosal’s original algorithm, the improved algorithm deals with mask effect and first derivative model on edge gradient direction so that it has the strong robust to noise, self-thinning ability and higher locating precision. An algorithm based on ZMs to extract line is also proposed, the comparative results with SHT and RHT show the method has the highest calculation speed and accuracy. The stereovision technology is developed to identify dimensional position of weld seam by computing dimensional coordinates of the weld seam. According to characteristics of weld seam, view field scope model and stereovision model based on baseline are studied and a stereo matching method is presented. In order to evaluate the algorithms and models presented in this paper, a welding robot systems with single camera fixed on the weld torch end-effector has been established for the robot to identify the dimensional position of typical weld seam by one-item and two-position method. The experiment results on S-shape and saddle-shape weld seams show that the vision computing method developed in this paper can be used for acquiring weld seam dimensional position information in welding robot system. Thus the welding path is mapped before the welding operation is executed.     

Automatically composing and parameterizing skills by evolving Finite State Automata

We propose a robotics algorithm that is able to simultaneously combine, adapt and create actions to solve a task. The actions are combined in a Finite State Automaton whose structure is determined by a novel evolutionary algorithm. The actions parameters, or new actions, are evolved alongside the FSA topology. Actions can be combined together in a hierarchical fashion. This approach relies on skills that with which the robot is already provided, like grasping or motion planning. Therefore software reuse is an important advantage of our proposed approach. We conducted several experiments both in simulation and on a real mobile manipulator PR2 robot, where skills of increasing complexity are evolved. Our results show that (i) an FSA generated in simulation can be directly applied to a real robot without modifications and (ii) the evolved FSA is robust to the noise and the uncertainty arising from real-world sensors.     

基于非单调共轭梯度算法的声纹识别机器人控制系统设计

传统声纹识别人控制系统识别准确率低,存在语音识别噪声鲁棒性问题。针对上述问题,基于非单调共轭梯度算法设计了一种新的声纹识别机器人控制系统,采用BioVoice 2.0 标准声纹采集器采集数据,提取声纹特征,根据提取的声纹特征建立模型库,同时引用了两个声纹数据采集终端,型号分别是TMC104-B和TMC104,选用型号为AS-MrobotR的机器人配合采集器和采集终端实现工作。在Windows平台下使用C/C+语言研究了一种专用的程序,在程序内部添加mde-api数据库,完成训练程序和识别程序。实验结果表明,基于非单调共轭梯度算法的声纹识别机器人控制系统能够很好地解决语音识别噪声鲁棒性问题,在有噪声环境下识别准确率提高15.24%,在无噪声环境下识别准确率提高21.55%。     

Learning of grasp selection based on shape-templates

The ability to grasp unknown objects still remains an unsolved problem in the robotics community. One of the challenges is to choose an appropriate grasp configuration, i.e., the 6D pose of the hand relative to the object and its finger configuration. In this paper, we introduce an algorithm that is based on the assumption that similarly shaped objects can be grasped in a similar way. It is able to synthesize good grasp poses for unknown objects by finding the best matching object shape templates associated with previously demonstrated grasps. The grasp selection algorithm is able to improve over time by using the information of previous grasp attempts to adapt the ranking of the templates to new situations. We tested our approach on two different platforms, the Willow Garage PR2 and the Barrett WAM robot, which have very different hand kinematics. Furthermore, we compared our algorithm with other grasp planners and demonstrated its superior performance. The results presented in this paper show that the algorithm is able to find good grasp configurations for a large set of unknown objects from a relatively small set of demonstrations, and does improve its performance over time.     

Modelling an arm's spatially correlated biases: an application to camera calibration for teleoperation

Calibration of the extrinsic and intrinsic parameters of cameras viewing the workspace of a teleoperated robot arm is a necessary step in the recovery of metric information for transmission to the operator. In this paper, we develop an algorithm that uses the robot slave manipulator to generate the calibration pattern of points in three dimensions which define the world frame. These points are used, along with their associated image positions, as measurements in a maximum likelihood estimation of the calibration parameters. However, the algorithm does not assume that the robot generates perfect world positions. Instead they are optimised using a model for covariance which respects the physical attributes of the robot arm and its linkages. Errors in world positions are actually correlated biases rather than stochastic noise, and so they are assigned a covariance that is a function of position and that encodes a coherence length over which relative position measurements are accurate. Experiments are presented which compare the calibration results with the camera parameters obtained using a conventional calibration grid, and which validate the proposed covariance model.