视觉和电磁导航智能车的路径识别及控制研究  

Path Recognition and Control of Intelligent Vehicle by Visual and Navigation

作　　者：焦冰 何雨豪 叶梦君 吴双 

机构地区：[1]南京航空航天大学金城学院机电工程与自动化学院,江苏南京211156

出　　处：《工业控制计算机》2024年第10期46-48,共3页Industrial Control Computer

基　　金：教育部产学合作协同育人项目(202102356013);南京航空航天大学金城学院科研基金项目(XJ202201)。

摘　　要：为实现智能车对多种复杂路径的识别,设计了一种以视觉和电磁传感为主、多种传感器融合协作的智能车路径识别及控制方案。视觉传感器获取赛道灰度图像,通过OTSU算法进行自适应图像二值化,然后提取赛道中心线计算智能车的方向偏差;电磁传感器采集的赛道数据经过放大、滤波、归一化后,采用差比和差算法计算方向偏差,并提出环岛等复杂路径的识别算法。根据方向偏差,综合多种传感器测得的车速、角度、距离等信息,采用PID算法实现智能车的方向和速度控制。实验结果表明,智能车可以正确识别弯道、环岛、路障等多种复杂路径,以2 m/s的速度在赛道上稳定行驶。In order to realize the recognition of various complex paths by intelligent vehicle,a path recognition and control scheme based on visual and electromagnetic sensing and multi-sensor fusion cooperation is designed.The gray image of the track obtained by the vision sensor is transformed into a binary image by the OTSU algorithm,then the center line of the track is extracted to calculate the direction deviation of the smart car.The track data collected by the electromagnetic sensor is amplified,filtered and normalized,and the direction deviation is calculated by the difference ratio and difference algorithm.A recognition algorithm for complex paths such as roundabouts is proposed.According to the direction deviation and the information of speed,angle and distance measured by various sensors,the PID algorithm is used to realize the direction and speed control of the intelligent vehicle.

关 键 词：智能车控制 图像处理 OTSU算法 电磁导航 路径识别 

分 类 号：TP23[自动化与计算机技术—检测技术与自动化装置]