基于STM32移动机器人目标动态追踪的研究

发布时间：2018-08-27 08:41

【摘要】：基于视觉的目标动态跟踪是当前智能机器人研究的热点之一,应用机器人视觉模拟人类视觉,具有深远的应用价值。其中,家用扫地机器人就是一个应用方面。对于视觉智能机器人虽有大量研究,就目前而言依然存在许多不足之处。比如在实际应用中,往往希望智能扫地机器人可以通过颜色识别达到目标准确跟踪的目的,但目前市场许多机器人在清扫过程中存在乱撞、迷路、费时等问题。基于上述存在的问题,本文尝试设计一款视觉目标动态跟踪的机器人,为智能扫地机器人的优化设计提供重要参考价值。主要是进行视觉采集跟踪的研究,根据采集到的颜色特征识别目标、P1D调节电机的转速、指令控制转向,实现小车直行或转向等跟踪运动,避免行走过程中乱撞,达到高效率跟踪的效果。论文主要的研究工作如下:1.构建基于stm32控制器的四轮平衡小车目标动态跟踪系统的基本结构框架,按照需求设计相关外围核心硬件电路系统。具体内容主要包括系统的软件算法设计及硬件电路的调试。2.系统硬件设计。对实现机器人动态目标跟踪的相关外围模块的工作参数、稳定性、stm32控制器之间的协调性进行分析,最终拟定最优设计方案,应用Altium Designer开发环境设计系统电路原理图和PCB图。外部电路主要由OV7725目标信息采集模块、TFT-LCD液晶显示模块、E18-D80NK-N红外避障模块等构成。3.软件程序设计。为实现四轮平衡小车系统对不同颜色的目标进行跟踪,四轮机器人主要通过直走、暂停、转弯三种运动方式实现跟踪。设计了合理的图像采集和处理算法对目标准确识别和定位,编写相关PID调电机脉冲宽度程序控制小车平稳快速的追踪目标。4.电路板焊接与软件程序调试。基于设计的PCB原理图,利用电子信息学院提供的资源完成了硬件电路板的制作与焊接,寻找合适的供电系统和下载线,组装小车各模块。最后,在IAR系统平台进行目标跟踪控制程序的调试。设计结果表明,本文设计的小车跟踪系统对某一确切颜色目标可准确识别定位和平稳的跟踪,为家庭扫地机器人的设计提供重要依据。
[Abstract]:Target dynamic tracking based on vision is one of the hotspots in the research of intelligent robot. It is of great value to simulate human vision with robot vision. Among them, household floor sweeping robot is an application aspect. Although there is a lot of research on visual intelligent robot, there are still many shortcomings. For example, in practical applications, it is often hoped that the intelligent floor sweeping robot can achieve the goal of accurate tracking by color recognition. However, at present, many robots in the market have problems such as random collision, lost, time-consuming and so on. Based on the above problems, this paper attempts to design a robot for dynamic tracking of visual targets, which provides an important reference value for the optimization design of intelligent floor sweeping robot. According to the collected color characteristics, the P1D can adjust the speed of the motor, control the steering, realize the tracking motion, such as the car's straight or steering, and avoid the random collision during the walking process, so as to carry out the research of visual acquisition and tracking, and identify the target P1D according to the collected color characteristics. Achieve high efficiency tracking effect. The main research work of this paper is as follows: 1. Based on stm32 controller, the basic frame of dynamic tracking system for four-wheel balancing vehicle is constructed, and the related peripheral core hardware circuit system is designed according to the requirement. The specific content mainly includes the software algorithm design of the system and the debugging of the hardware circuit. 2. System hardware design. Based on the analysis of the working parameters of the relative peripheral modules and the coordination among the stm32 controllers, the optimal design scheme is drawn up, and the circuit schematic diagram and PCB diagram of the system are designed by using the Altium Designer development environment. The external circuit is mainly composed of E18-D80NK-N infrared obstacle avoidance module of OV7725 target information acquisition module and TFT-LCD liquid crystal display module. Software programming. In order to realize the tracking of different color targets by the four-wheel balancing vehicle system, the four-wheel robot can be tracked by three kinds of motion modes: straight walking, stopping and turning. A reasonable image acquisition and processing algorithm is designed to accurately identify and locate the target, and the program of pulse width of the PID motor is written to control the vehicle to track the target steadily and quickly. Circuit board welding and software program debugging. Based on the schematic diagram of PCB, the hardware circuit board is made and welded with the resources provided by the Institute of Electronic Information, the suitable power supply system and download line are found, and the modules of the car are assembled. Finally, the target tracking control program is debugged on the IAR platform. The design results show that the vehicle tracking system designed in this paper can accurately identify and locate a certain specific color target and track it stably, which provides an important basis for the design of a family floor sweeping robot.
【学位授予单位】：内蒙古大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP242

【参考文献】