基于Android的智能小车控制及路径规划

发布时间：2018-01-17 22:30

本文关键词：基于Android的智能小车控制及路径规划　出处：《广西师范大学》2017年硕士论文　论文类型：学位论文

【摘要】：目前,市场上热卖的扫地机器人作为智能家居新概念的领跑者已经走进人们的生活,不仅可以自主清扫房间,而且能通过手机设备或是电脑无线控制,实现人机交互,其实用性越来越强,它也越来越受人们的欢迎,然而市面上的扫地机器人有时会出现避障不灵活、不能完全覆盖工作环境或是功能不全面等问题,同样地,关于室内安防问题也是人们一直苦恼又难以解决的问题。结合这些问题和人们的需求,本文搭建了一台智能移动机器人——智能小车,在智能小车前方安装两个摄像头,实现Android客户端的无线控制及摄像监控,并且研究扫地机器人的路径规划,采用栅格地图和改进的往返式”梳”字型相结合的路径规划方法,同时可结合摄像头拍摄到的图像对室内物体进行分类识别,做出相应的避障策略,使得智能小车能更加有效地完成作业,这样将该智能小车应用到扫地机器人上后,扫地机器人既能更加灵活地覆盖清扫区域,又能作为“巡逻兵”,监视室内指定区域,增强室内安防,更好地优化并扩展了扫地机器人的功能。本文主要工作内容如下:(1)智能小车的搭建:此车主要以一个树莓派主控板为核心,再结合一个Arduino板主控板进行搭建。树莓派主要包括自动连接WIFI、传输摄像头拍摄到的视频等功能,Arduino板主要包括返回3个方位超声波传感器数据、电机驱动等功能,关于智能小车的搭建已申请了实用新型专利。(2)智能控制移动摄像监控:本文在Eclipse平台上进行APP开发,采用SurfaceView绘图机制和多线程实现了视频显示,在不同型号的Android手机上进行测试,得出最高平均帧率达到25fps,最高帧率达到30fps;其次可无线控制智能小车方向,本文由闭环PID控制算法和霍尔传感器相结合来控制智能小车左右电机转速,保证左右电机同步,使得智能小车能直线行走不偏移;最后可实时获取小车传感器的数据,作为路径规划的判断依据。(3)路径规划研究:扫地机器人路径规划属全覆盖路径规划,覆盖率是最主要的评估标准,其次是重复率,本文比较常用算法的优缺点和目前全覆盖路径规划难点,采用栅格地图和改进的往返式”梳”字型相结合的路径规划,将本文方法应用在多种不同的环境中,如在十种典型常见室内环境、覆盖率较高文献中的同一室内环境或是构建一个扫地机器人清扫时遇到的常见且难避免的障碍物情形的室内环境中,测得覆盖率最高能达到100.00%,进一步地完善和优化了扫地机器人的全覆盖路径规划。
[Abstract]:At present, the popular floor sweeping robot in the market as a leader in the new concept of smart home has come into people's lives, not only can clean their own rooms, but also through mobile devices or computer wireless control. The realization of man-machine interaction is more and more practical, and it is more and more popular. However, the floor sweeping robot on the market sometimes appears obstacle avoidance inflexible. Cannot completely cover the work environment or the function is not comprehensive and so on the question, similarly, regarding the indoor security question also is the question which the people has been distressed and difficult to solve, unifies these problems and the people's demand. In this paper, an intelligent mobile robot, intelligent car, is built. Two cameras are installed in front of the smart car to realize wireless control and camera monitoring of Android client. And the path planning of the floor sweeping robot is studied, which combines the grid map and the improved round-trip "comb" font. At the same time, it can be used to classify and identify the indoor objects with the images captured by the camera. Make the corresponding obstacle avoidance strategy, so that the intelligent car can complete the operation more effectively, so that after the intelligent car is applied to the floor sweeping robot, the floor sweeping robot can cover the cleaning area more flexibly. It can also act as a "patrol soldier" to monitor designated indoor areas and enhance indoor security. Better optimize and expand the function of floor sweeping robot. The main work of this paper is as follows: 1) the construction of intelligent car: this car mainly takes a raspberry pie main control board as the core. The raspberry pie mainly includes the functions of automatically connecting WiFi, transmitting the video captured by the camera, and so on. Arduino board mainly includes three azimuth ultrasonic sensor data, motor drive and other functions. About the construction of intelligent car has applied for a utility model patent. 2) Intelligent control mobile camera monitoring: this paper on the Eclipse platform for APP development. The video display is realized by using SurfaceView drawing mechanism and multithreading, and tested on different Android mobile phones. The highest average frame rate is 25fps. The maximum frame rate is 30fps. Secondly, the direction of intelligent car can be controlled wirelessly. This paper combines the closed-loop PID control algorithm and Hall sensor to control the speed of the motor around the intelligent car and to ensure the synchronization of the left and right motors. The intelligent car can walk straight line without deviation; Finally, the vehicle sensor data can be obtained in real time, as the judgment basis of path planning. Research on path planning: path planning of sweeping robot belongs to full coverage path planning, coverage is the most important evaluation criteria. Secondly, repetition rate, this paper compared the advantages and disadvantages of common algorithms and the current full coverage path planning difficulties, using raster map and the improved round-trip "comb" combination of path planning. This method is applied to many different environments, such as ten typical indoor environments. Coverage is higher than Gao Wenxian in the same indoor environment or the construction of a sweeping robot cleaning common and unavoidable obstacles in the indoor environment, the highest coverage can be measured to 100.00%. The full coverage path planning of floor sweeping robot is further improved and optimized.
【学位授予单位】：广西师范大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP242

【参考文献】