基于多传感器融合的智能避障系统设计与实现

打开文本图片集

中图分类号：TP242.6 文献标识码：A文章编号：2096-4706（2025）23-0023-07

Abstract：To improvereal-timeobstacle avoidance performance，this paperdesignsand implements anintellgentobstacle avoidance vehicle system based on the STM32F407ZGT6 maincontrolchip 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 planing.Itinnovativelyadoptsadual-loopPIDstructure （incrementalPIforthespeedlopandpositionalPDforthe steringloop）incontrolarchitecture，andintegratesencoderandgyroscopedatatoachievehig-precisionmotioncontrol.By optimizing the visionrecognitionalgorithmthrough ROl-based threshold processng，thedetection effciencyandaccuracyfor blackcylindricalobstaclesaresignifcantlyenhancedExperimentalresutsdemonstratetatthesystemexhibitsexcelentraltime performanceandobstacleavoidancecapability，providingapracticalsolution for thefeldofintellgentmobilerobots.

Keywords： K230; A* algorithm; dual-loop PID; visual ROI; grid map

0 引言

随着人工智能与移动机器人技术的迅速发展，智能避障小车在自动化物流、紧急抢险及环境探测等领域的应用日益广泛。（剩余8821字）