共查询到17条相似文献,搜索用时 440 毫秒
1.
2.
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。 相似文献
3.
4.
5.
以集控式足球机器人为研究对象,以双闭环视觉伺服控制结构为主线,提出了基于DSP技术和非线性PI控制算法的机器人小车控制系统设计,保证了控制的准确性,提高了控制的实时性;论证了单参数模糊自校正非线性PI算法应用于机器人位置环控制中的可行性,增强了系统的自适应能力。实验证明了此方案对足球机器人运动控制的有效性,其原理对其它类型移动机器人的控制同样有借鉴作用。 相似文献
6.
机器人足球视觉图像快速识别方法的研究 总被引:1,自引:0,他引:1
机器人足球比赛作为多智能体系统(Multi-agent System, MAS)的理想研究平台已成为复杂人工智能研究的一个新兴领域.机器人足球比赛是使用计算机视觉闭环的反馈控制系统.机器人足球队员的色标设计是直接关系到系统辨识精度、实时性和抗干扰性的一个重要因素.该文通过对4种典型色标设计进行比较和分析,研究出一种比较合理的全新的队员识别算法设计思路,并针对性地提出了一种快速准确的神经网络边缘检测算法,利用足球机器人外形特征取得了良好的辨识结果.该算法对基于视觉的分布式控制式足球机器人系统的研究与开发进行了有益的探索. 相似文献
7.
多智能体足球机器人系统的模型与层次 总被引:2,自引:0,他引:2
机器人足球是集成视觉技术、通讯技术、伺服控制技术、决策对象等多种技术的综合系统,是当前人工智能和机器人领域的研究热点之一。基于机器人足球理论基础的论述,分析了当前足球机器人系统的三种控制方式;按照智能程度的高低,研究了从动作层到决策层的功能,为系统实现构建了完整的层次框架。 相似文献
8.
9.
RoboCup机器人足球比赛作为在世界上最有影响力的机器人赛事之一,备受推崇。同时作为多智能体研究的标准实验任务,已经成为人工智能的研究焦点。本文综述了RoboCup机器人足球赛的发展概况和小型组足球机器人系统,对小型组足球机器人的各子系统和相关关键技术进行较为详细的分析和比较,最后对小型组机器人足球技术的发展趋势进行了分析和展望。 相似文献
10.
11.
12.
本文简要介绍足球机器人系统的四大功能模块即视觉系统、决策系统、通信系统和车体系统。然后对足球机器人的各个功能系统分别提出其实验性设计,以实现每个功能系统的多视角实验平台,该功能系统的实验平台除对足球机器人系统进行功能研究外,还可用于智能控制的教学和研究。 相似文献
13.
随着工业机器人及其应用的不断发展,要求一个强有力的计算机系统来控制它的工作,并具有灵活、方便的机器人编程语言.本文系统地介绍了我们自行设计并实现的一个先进的机器人控制系统——ARCS.该系统主要包括两部分:(1)一个实时多任务的机器人控制软件SVAL系统,该系统支持一种通用性较强的机器人编程语言——SVAL语言.(2)一个支持该软件系统工作的、具有开放式结构的硬件环境.ARCS系统具有良好的实时性、可扩展性及基于外部传感器信号进行控制的能力.由于该系统的开放式结构.使其根据不同要求可方便地增删其功能,并可控制不同类型的机器人.我们已成功地实现对PUMA760机器人的控制,并在其上引入了力觉与接近觉的传感器,采样时间可缩短到16ms.一年多的运行结果证明,该系统稳定可靠,性能良好,现在正向产品转化. 相似文献
14.
Meiying Ou Haibin Sun Shengwei Gu Yangyi Zhang 《International journal of systems science》2017,48(15):3233-3245
This paper investigates the distributed finite-time trajectory tracking control for a group of nonholonomic mobile robots with time-varying unknown parameters and external disturbances. At first, the tracking error system is derived for each mobile robot with the aid of a global invertible transformation, which consists of two subsystems, one is a first-order subsystem and another is a second-order subsystem. Then, the two subsystems are studied respectively, and finite-time disturbance observers are proposed for each robot to estimate the external disturbances. Meanwhile, distributed finite-time tracking controllers are developed for each mobile robot such that all states of each robot can reach the desired value in finite time, where the desired reference value is assumed to be the trajectory of a virtual leader whose information is available to only a subset of the followers, and the followers are assumed to have only local interaction. The effectiveness of the theoretical results is finally illustrated by numerical simulations. 相似文献
15.
Adam Grzech 《控制论与系统》2013,44(5):531-545
The aim of the paper is to present some basic rules allowing comparison of data processing and data communication subsystems in a form applicable in practice. The proposed approach addresses the corporate network communication subsystems design problem. Corporate networks classification, design, evaluation, planning, monitoring, and management as well as tuning, integration, or migration tasks are usually formulated as finding a communication system that is most suitable for a given corporate data processing system. The design procedures are based on the assumption that it is possible to transform data processing subsystem features and requirements into communication subsystem characteristics and vice versa. The features and characteristics when known allow comparison of categorized requirements and available services. 相似文献
16.
《Robotics and Autonomous Systems》2001,34(1):23-38
A self-localization system for autonomous mobile robots is presented. This system estimates the robot position in previously learned environments, using data provided solely by an omnidirectional visual perception subsystem composed of a camera and of a special conical reflecting surface. It performs an optical pre-processing of the environment, allowing a compact representation of the collected data. These data are then fed to a learning subsystem that associates the perceived image to an estimate of the actual robot position. Both neural networks and statistical methods have been tested and compared as learning subsystems. The system has been implemented and tested and results are presented. 相似文献
17.
We propose a biomimetic, two-layered, hierarchical control structure for adaptive locomotion of a hexapod robot. In this structure,
the lower layer consists of six uniform subsystems. Each subsystem interacts locally with its neighboring subsystems, and
autonomously controls its own leg movements according to the weighted sum of three basic vector fields that represent the
three basic motion patterns of the robot body. The upper-layer controller decides the intended body movement, and sends the
lower-layer controllers three variables as the weights of each basic vector field. This approach greatly reduces the communication
between the two layers, and contributes to real-time adaptive locomotion. 3D dynamic simulations, as well as experiments with
a real modularized hexapod robot, show the effectiveness of this hierarchical structure.
This work was presented, in part, at the Sixth International Symposium on Artificial Life and Robotics, Tokyo, Japan, January
15–17, 2001 相似文献