Font Size: a A A

Design And Development Of Intelligent Automated Guided Vehicle For Bobbin Doffing And Handling System

Posted on:2024-02-29Degree:MasterType:Thesis
Institution:UniversityCandidate:AYBALA CAKIRFull Text:PDF
GTID:2531306941954319Subject:Pattern Recognition and Intelligent Systems
Abstract/Summary:PDF Full Text Request
Autonomous guided vehicles(AGVs)are capable of actively perceiving their surrounding environment,making autonomous decisions based on environmental data,and executing material handling tasks.They are particularly suitable for improving transportation efficiency,and reliability,and reducing labor intensity in industrial settings,making them a research hotspot in recent years.This article focuses on the background of fiber production equipment and takes a 4-degree-offreedom(4DOF)Cartesian robotic arm-equipped AGV as the research object.It aims to analyze the motion mechanism,design and develop an AGV system and address key issues such as AGV positioning,target alignment,and system control.To address the need for fast and reliable unloading and handling of fiber bobbins or similar materials,a 4DOF Cartesian robotic arm-equipped AGV that can autonomously dock and align the end effector with the center of the bobbin is designed.This design takes into consideration the need for an expanded mobility range,increased rotational angles,and higher degrees of freedom and flexibility.To tackle the key issues involved in autonomous navigation in complex environments,a method is constructed based on the gmapping algorithm and fusion of depth camera information to create an environmental map.The sensor data is first fused using the Extended Kalman Filter(EKF)algorithm,and based on the fused data,the Dijkstra algorithm is employed for autonomous path planning of the AGV,enabling autonomous navigation and real-time obstacle avoidance.For bobbin detection,recognition,and precise positioning,a bobbin center hole detection method based on the YOLO v5 algorithm is proposed.Homography is computed to transform the original 3D model of the bobbin hole into a 2D image that can be matched with the camera image.Bobbin hole positioning is achieved through feature matching.PID control and reinforcement learning algorithms are employed to control the Cartesian robotic arm,enabling fast and accurate unloading and material transfer.Finally,simulation analysis of the 4DOF Cartesian robotic arm-equipped AGV is conducted using a Gazebo,which validates the feasibility of the design.The 4DOF Cartesian robotic arm-equipped AGV designed in this article is mainly aimed at fiber production equipment and serves as a core device for improving precision,efficiency,and reliability,reducing labor intensity,and labor costs in the manufacturing industry.It can enhance the level of automation in enterprises.
Keywords/Search Tags:Autonomous guided vehicle, Robotic arm, Image processing, Material handling and transportation, Machine learning
PDF Full Text Request
Related items