基于视觉导航的两轮自平衡小车系统本科论文 .doc

摘　　要 两轮自平衡小车是移动机器人的一个重要分支，具有体积小、运动灵活、可零半径转弯等诸多优点。并且相对于传统的四轮车，两轮车在直立行走、导航方面对硬件设计、软件开发以及系统调试等提出了更高的要求。 本文设计了一种基于视觉导航的两轮自平衡小车系统，并实现了小车的直立行走、自主寻迹两大功能。系统用飞思卡尔十六位微处理器XS128作为核心控制单元，通过增加各种传感器并编写相应程序以完成平衡控制，速度控制，转向控制和自主寻迹四大任务。 本设计通过对倒立摆动力学模型的分析，得到小车平衡的条件。利用陀螺仪和加速度计获得车体的倾角和角速度，并进行互补滤波，对倾角进行PD控制实现小车的基本直立；通过光电编码器得到车子的速度，通过转向陀螺仪得到转向角速度，对速度进行PI控制，将转速控制信号与平衡控制信号叠加加载到后轮两电机上，实现小车的静止和直立行走；用摄像头获得道路图像信息，进行图像预处理，提取出路径中心线，获得小车的方向偏差控制量，通过对方向控制量和转向角速度进行PD 控制来调节左右两轮的差速实现小车的转向和自主寻迹。本设计利用了串口调试上位机对系统进行实时调试。实践证明，该两轮自平衡车具有很好的静态平衡性能和动态性能，并能在道路上自主寻迹，流畅转弯，且该系统能有效地抵抗外部扰动。 Abstract Two-wheeled self -balancing vehicle is an important branch of the mobile robot, with a small, flexible movement, can zero turning radius, and many other advantages. And relative to the traditional four cars, two-wheelers in the upright walking, navigation hardware design, software development and system debugging and put forward higher requirements. This paper has designed a two wheeled self-balanced intelligent car system based on visual navigation, and finally achieved the goal that the car could walk upright stably and run along the racing track. Freescale sixteen microprocessor systems MC9S12XS128 as the core control unit, by adding a variety of sensors and to prepare appropriate procedures to complete the balance control, speed control, steering control and autonomous tracking of the four tasks. Through this design inverted pendulum dynamics modeling, analog get the car balanced conditions. Then, this system uses gyroscopes and accelerometers to measure the dip angle and angular velocity of the car, and utilizing Complementary Filter to fuse dip angle and angular velocity. PD control inclination is achieved on the basic car upright. Then the design uses photoelectric encoders to obtain the speed of two wheels and steering angular velocity. The speed is controlled by PI controller. The car could keep stationary and upright by combining the speed control signal and dip angle control signal together. The design uses a C