摘要
针对当前AGV无法对障碍物做出前瞻性预判的问题,设计一种优化智能物流AGV障碍物激光扫描识别系统。将采集的数据生成图片并进行去噪处理,利用背景减除等方法检测障碍物,并传输信号到AGV系统中;基于多步前瞻算法及TOF测距方法,完成障碍物的弱分界面区分迭代识别及测距。实验结果表明:设计的系统应用到AGV中,障碍物平均正检率高于90%,且障碍物面积检测计算的误差率最大仅为3.94%,设计的系统具有一定的实用价值。
It develops an optimized intelligent logistics AGV obstacle laser scanning recognition system to address the issue of AGV's inability to make forward-looking predictions about obstacles.It generates the images from the collected data and perform denoising processing,uses methods such as background subtraction to detect obstacles and transmits signals to the AGV system;Based on multi-step forward-looking algorithm and TOF ranging method,it completes iterative identification and ranging of weak interface differentiation of obstacles.The experimental results show that the designed system applied to AGV has an average obstacle detection rate of over 90%,and the maximum error rate of obstacle area detection calculation is only 3.94%.This system has certain practical value.
作者
胡跃明
Hu Yueming(Medium CSR Yangtze Co.,Ltd.,Hubei Wuhan,430212,China)
出处
《机械设计与制造工程》
2024年第10期57-61,共5页
Machine Design and Manufacturing Engineering
基金
湖北省科技厅科技支撑项目(2021BAA083)。
