摘要
为提升实时避障性能,文章设计并实现了一套基于STM32F407ZGT6主控芯片与K230视觉模块的智能避障小车系统。该系统采用栅格地图完成环境建模,运用以曼哈顿距离为启发函数的A*算法实现路径规划;在控制架构上创新性采用双环PID结构(速度环采用增量式PID,转向环采用位置式PID),融合编码器与陀螺仪数据达成高精度运动控制;通过对ROI区域的阈值处理优化视觉识别算法,显著提升了黑色圆柱障碍物的检测效率与准确率。实验验证表明,本系统在实时性、避障性能方面均表现优异,为智能移动机器人领域提供了一套具有实践价值的解决方案。
To improve real-time obstacle avoidance performance,this paper designs and implements an intelligent obstacle avoidance vehicle system based on the STM32F407ZGT6 main control chip and a K230 vision module.The system employs grid maps for environmental modeling and utilizes the A*algorithm with the Manhattan distance as the heuristic function for path planning.It innovatively adopts a dual-loop PID structure(incremental PID for the speed loop and positional PID for the steering loop)in control architecture,and integrates encoder and gyroscope data to achieve high-precision motion control.By optimizing the vision recognition algorithm through ROI-based threshold processing,the detection efficiency and accuracy for black cylindrical obstacles are significantly enhanced.Experimental results demonstrate that the system exhibits excellent realtime performance and obstacle avoidance capability,providing a practical solution for the field of intelligent mobile robots.
作者
何启明
刘鹏来
钟志国
谭展濠
赖顺兴
何鹏晖
HE Qiming;LIU Penglai;ZHONG Zhiguo;TAN Zhanhao;LAI Shunxing;HE Penghui(School of Automotive Engineering,Guangdong Polytechnic of Industry and Commerce,Guangzhou 510510,China;School of Information Engineering,Guangdong University of Technology,Guangzhou 510006,China)
出处
《现代信息科技》
2025年第23期23-28,34,共7页
Modern Information Technology
