智能家居环境中基于移动机器人的目标搜索

发布时间：2018-05-18 07:56

本文选题：移动机器人 + 机器人操作系统　；参考：《浙江大学》2017年硕士论文

【摘要】：近年来,应用服务型智能机器人到智能家居系统中越来越成为一个热点研究问题。由于在家居环境中,拥有多种传感设备的机器人搜索人与物体的场景时常发生,因此为了解决这个问题,本文着眼于设计适应家居环境的移动机器人目标搜索算法,提出一整套高效的搜索策略。为了完成面向智能家居的目标搜索的任务,首要任务是对移动机器人进行开发。本文在布置好的室内家居环境中,规划和搭建好移动机器人硬件软件平台后,实现了室内移动机器人的各项基础功能,包括机器人运动控制、传感器管理与信息融合、同时定位与地图构建(SLAM)、自主定位与坐标校准、导航与避障等。针对智能家居环境下移动机器人目标搜索的问题,本文提出在栅格地图上基于改进的遗传算法和A*算法进行路径规划的算法,对目标的识别采用基于深度学习方法的YOLO检测系统,机器人在各个单元区域中的搜索需要充分考虑环境因素和传感器条件,结合目标在室内环境中位置概率分布,最终减少搜索过程的总耗时。经过仿真与实际实验验证,本文所提出的算法很好地完成了移动机器人在短时间内搜索到目标的任务,在各种不同的情况下都具有显著的效果,说明了研究内容在智能家居环境中老年人看护、远程医疗等日常任务中的有效应用性。
[Abstract]:In recent years, the application of service intelligent robots to smart home systems has become a hot research issue. In order to solve this problem, a mobile robot target search algorithm adapted to home environment is designed in order to solve the problem, because the scene of robot search for people and objects with multiple sensing devices often occurs in the home environment. A set of efficient search strategies is proposed. In order to accomplish the task of target search for smart home, the most important task is to develop mobile robot. After planning and building the hardware and software platform of mobile robot, this paper realizes the basic functions of indoor mobile robot, including robot motion control, sensor management and information fusion. At the same time, positioning and map building, autonomous positioning and coordinate calibration, navigation and obstacle avoidance, etc. In order to solve the problem of mobile robot target search in smart home environment, this paper proposes a path planning algorithm based on improved genetic algorithm and A * algorithm on grid map. YOLO detection system based on depth learning method is used for target recognition. The search of robot in each unit area needs to take full account of environmental factors and sensor conditions, combined with the probability distribution of target position in indoor environment. Ultimately reducing the total time taken in the search process. After the simulation and actual experiment, the algorithm proposed in this paper has accomplished the task of mobile robot searching target in a short time, and it has remarkable effect in all kinds of different cases. The effective application of the research in the daily tasks of elderly care and telemedicine in intelligent home environment is explained.
【学位授予单位】：浙江大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP242

【参考文献】