1673-159X

CN 51-1686/N

基于视觉的智能车道路检测与转向控制策略研究

Study of Road Detection and Steering Control Strategies for Vision-based Intelligent Vehicles

  • 摘要: 对以FREESCALE的MC9S12DG128芯片为核心控制器的智能车采用了OV7620数字摄像头进行路径识别, 再利用核心控制芯片对采集到的数据进行分析并提取出赛道黑色中心线, 对赛道上出现的不规则黑色干扰信号进行过滤处理, 从而准确地判别出赛道的形状, 为舵机和电机提供控制依据以使小车平稳快速的行驶。并对智能车的转向控制策略的选择改进进行了分析, 通过对比查表PID控制策略和模糊PID控制策略, 以及具体试验观察, 最终选择了模糊PID控制策略, 实现了对直赛道、弯曲赛道的识别和优化处理, 运行效果得到了改进, 特别是在弯道处, 在保证车体平稳的情况下车速有了极大的提高。

     

    Abstract: The smart car uses FREESCALE's MC9S12DG128 as the core controller chip and OV7620 digital camera for path identification. The core control chip collects data to analyze and extract the black center line, and then filters the irregular black interference signals. Judging the shape of the line accurately is able to control the steering wheel and motor and let the car move stably and fast. The strategy for vehicle speed control system is introduced, and the choice of the control strategies is analyzed and improved by comparing testing look-up table PID control strategy with fuzzy PID control strategy. The final choice is the fuzzy PID control strategy through specific test and observation. As a result, the identification of straight track, curved track is realized, and the effect is improved, especially in corners where the body can keep stable and the speed is greatly enhanced.

     

/

返回文章
返回