【作者】 孟宪鹏；

【导师】 邢建平；

【作者基本信息】 山东大学 ， 电子与通信工程， 2017， 硕士

【摘要】 随着人工智能在2016年的飞速发展,不论是科研研究方面还是工商界中的融资规模,甚至是公众对于人工智能时代的意识,都证明了人工智能元年的开始。在人工智能领域,AlphaGo与李世石的人机大战,谷歌、特斯拉、百度的无人车三年商用五年推广,京东的仓库智能运输车,最强大脑的人机交锋。以上都表明人工智能的发展不仅具有传统意义上的理论研究价值,而且具有巨大的工商界实际应用价值。基于此,如何将人工智能这一概念成熟稳健又适用大众的出现在人们的视野里成为了目前工作的重点之一。同时,机器人平台作为现今人工智能技术的集大成者出现在人们视野之中。面向大众的机器人平台理论与应用研究也在国内外不断发展进行之中。基于面向大众的复杂环境与复杂任务的情况,本文设计了面向任务调度的自主导航机器人平台,并对其关键技术进行相关研究。本文就自主导航机器人平台硬件与软件及交互方式等为主要研究内容。针对自主导航机器人平台在复杂环境与多任务情况下所遇到的多任务处理问题、即时定位与地图构建(SLAM,Simultaneous Localization AndMapping)问题、路径规划问题及人机交互方式问题进行了详细研究。针对上述问题,本文实现了一种面向任务调度的机器人系统(TSRP,Task Schedule Robot Platform),并针对多任务处理提出了分布式自主导航框架(ANDF,Autonomously Navigation Distributed Framework)。实现了一种混合粒子滤波算法(M-PF,Mixture Particle Filter),该算法解决了传统粒子滤波中遇到的粒子衰减问题。对于路径规划问题实现了一种多传感器融合导航算法(MSFN,Multi-Sensor information Fusion Navigation),实现了在满足三维环境导航环境下大幅降低成本。并基于安全考虑提出了一种先验地图的处理方案,该方案考虑到建图时的不足与路径规划导航时可能遇到的危险情况,进行了规避处理。在TSRP系统的基础上基于人机交互方式的考虑,设计了一套可应用与多种情景下的人机交互型面向任务调度的自主导航机器人应用平台(TSRP-HCI,Task Schedule Robot Platform Human Computer Interaction)。本文使用TSRP-HCI机器人平台,针对特定室内巡航场景对本文提出的各种方法进行了测试,测试结果表明本文提出的面向任务调度的自主导航机器人平台在定位精度、导航安全、算法效率与人机交互方式上优势,同时具有可观的应用价值。更多还原

【Abstract】 With the rapid development of artificial intelligence in 2016,whether it is scientific research.or business sector in the financing scale,and even the public awareness of the era of artificial intelligence,have proved the beginning of the first year of artificial intelligence.In the field of artificial intelligence,AlphaGo and Li Shishi man-machine war,Google,Tesla,Baidu’s unmanned car three years of commercial five-year promotion,Jingdong warehouse smart truck,the most powerful brain man-made battle.It shows that the development of artificial intelligence not only has the traditional theoretical value of research,but also has a huge practical value of the business community.Based on this,how to mature the concept of artificial intelligence and the application of the public in people’s vision has become one of the focuses of the current work.At the same time,the robot platform as a master of today’s artificial intelligence technology in the people’s vision.The theory and application of robot platform for the masses are also being carried out at home and abroad.Based on the complex environment and complex tasks for the masses,this paper designs the autonomous navigation robot platform for task scheduling,and carries on the related research to its key technology.In this paper,the hardware and software and interactive mode of autonomous navigation robot platform are the main research contents.Aiming at the problem of multi-task processing in the environment of complex environment and multi-task,the problem of path planning and human-computer interaction is studied in detail,and the problem of SLAM Simultaneous Localization and Mapping is studied in detail.In view of the above problems,this paper proposes a Task Schedule Robot Platform(TSRP),and proposes a distributed Autonomously Navigation Framework(ANDF)for multi-task processing.A Mixture Particle Filter(M-PF)is proposed to solve the problem of particle attenuation in traditional particle filter.A multi-sensor fusion navigation algorithm(MSFN)is proposed for the path planning problem,which greatly reduces the cost in satisfying the three-dimensional environment navigation environment.Based on the security considerations,a scheme of a priori map is proposed,which takes the avoidance process in consideration of the shortage of the map and the dangerous situation that may be encountered in the path planning and navigation.Based on the analysis of human-computer interaction on the basis of TSRP system,this paper designs a set of Task Schedule Robot Platform Human Computer Interaction(TSRP-HCI),which can be applied to a variety of scenarios.In this paper,the TSRP-HCI robot platform is used to test the various methods proposed in this paper.The test results show that the proposed algorithm is based on the mission-based autonomous navigation robot platform,such as positioning accuracy,navigation safety,algorithm efficiency and man-machine Interactive mode advantage,but also has considerable application value.更多还原