【作者】 于虹；

【导师】 崔桂梅；

【作者基本信息】 内蒙古科技大学 ， 控制理论与控制工程， 2009， 硕士

【摘要】 随着汽车的普及,智能汽车及其相关技术的研究己经成为热门。本文以全国大学生智能汽车竞赛为背景,设计开发一辆具有路径检测功能,能实现自主线路识别的智能车。本文采用模块化的设计方法,完成了智能车系统的硬件设计。通过系统分析,将整个系统分为7大模块,包括核心控制模块、电源模块、电机控制模块、舵机控制模块、路径检测模块、车速检测模块及调试模块。本文使用MC9S12DG128开发板作为系统的核心控制模块;路径识别模块采用10对红外光电传感器,一字型等间距布局;本文采用MC33886两片并联驱动电机,增加了驱动能力;车速检测模块采用光电旋转编码器,既满足驱动电机控制的精度要求,又方便传感器的安装。为了系统调试方便,设计了RS—232通信模块及LCD显示模块,并制定了相应的通信协议。本文设计了两层结构的控制系统,底层系统包括“转向控制器”和“车速控制器”。本文在路径识别采用了软件细分算法及基于传感器离散布局的连续路径识别算法,克服了红外光电传感器离散布局所采集到的赛道信息有限的缺点,提高了路径识别的精度;设计了时间最优意义下的全加速、紧急制动和闭环控制等多种模式中平稳切换的“多模式车速控制器”;由于模型车在封闭赛道上连续行驶两圈,在控制策略上采用了基于脉冲点的道路记忆算法,明显的提高了第二圈的成绩;本文给出了车速与转向协调控制规则及模型车调试规则,并经过多次实验,验证了该规则的有效性。更多还原

【Abstract】 As the car becomes widely used, the research of intelligent car and its related technology is becoming more and more popular. On the background of The Freescale Cup National Undergraduate Intelligent Car Contest, this paper studies designing a intelligent car with path detection function, which could identify the guide-line by itself.This paper accomplished the intelligent car system’s hardware design with modularized design method. Based on the system’s analysis, the whole system was divided into 7 different blocks, including core control block, power supply block, motor control block, steering control block, path recognition block, speed detection block, and debug block. The paper used MC9S12DG128 developing board as core control block; path recognition block adopted 10 pairs of infrared photo sensors, which arrayed as a row and equal gap. In order to enhance driving capability, the paper used parallel connection between two MC33886. The speed detection block used Optical Rotary Encoder, not only meet the requirement of motor control, but also convenient for installation. For purpose of facilitating the system debug, the paper designed RS-232 communications block and LCD display block, and developed a corresponding communication protocol.The paper designed two-layer-structure control system, and bottom layer system included direction controller and speed controller. In path recognition block used the software subdivision algorithm and path continuous recognition algorithm based on sensor discrete layout to overcome the shortcomings of the limited information collected because of discrete photoelectric sensor layout and improve the accuracy of the path recognition. Considering the means of time optimal , the paper designed a multimode speed controller, which were stably switched among full acceleration, emergency braking and Closed-Loop Control. Because the model vehicle goes continuously on the seal racecourse twice, the control strategy adopt the path memory algorithm based on path pulse points, greatly improves the speed of the second lap; This article put forward the coordination control rules based on direction an speed and the debugging rules of model vehicle, Moreover, the Validity of these rules confirmed by lots of experiments.更多还原