| Belt conveyor is one of the important transportation equipment in coal mine production,but it often transports goods at high speed,heavy load,long distance,and long time in harsh environments,and safety issues often occur.The traditional belt conveyor inspection mainly adopts the manual inspection method.During manual inspection,the harsh frontline work environment is not conducive to the physical and mental health of inspection personnel,even endangering their lives.At the same time,manual inspection is inefficient and has a high leak detection rate.Therefore,intelligent patrol inspection of belt conveyors has attracted widespread attention.At present,there are a small number of rail type inspection robots designed for the inspection task of belt conveyors,but the long distance and harsh working environment of mining belt conveyors lead to limited positioning accuracy of rail type inspection robots.In view of this,this thesis proposes a high-precision positioning method for belt conveyor inspection robots,and designs and develops an intelligent inspection robot system for belt conveyors.The main research work is as follows:(1)The intelligent inspection robot system for belt conveyor is designed.Firstly,based on the inspection tasks of belt conveyors,the main requirements of inspection robots are analyzed.Then,according to the characteristics of the belt conveyor inspection task,a traction type track inspection robot scheme is selected,and a modular functional structure of the inspection robot is constructed based on the demand analysis of the traction type track inspection robot.Finally,the hardware system and software system of the inspection robot are designed respectively.The hardware system is designed from two aspects: controller and peripheral hardware selection.The software system is designed from three parts: electronic control software,monitoring system software,and drive system software.(2)Intelligent inspection strategy for belt conveyor is designed.Firstly,the automatic cruise mode of the drive system is used to achieve automatic robot cruise,obtain on-site environment and belt conveyor operation data,and upload them.Then,intelligent monitoring of environmental parameters is realized based on environmental sensors and response devices collected by the inspection robot.Finally,the main fault types of the belt conveyor are analyzed,the patrol fault types are determined,and the corresponding data sets are collected.Using the training set to train the YOLOv5 fault detection model,the conveyor belt fault detection is implemented in the monitoring system software based on the fault detection model.Combining the intelligent monitoring function of environmental parameters,conveyor belt fault detection function,and the automatic cruise function of robots,the intelligent patrol inspection of belt conveyors is realized.(3)The high-precision positioning method for intelligent inspection robot of belt conveyor is designed.Firstly,the track characteristics are analyzed,and the track segmentation method is proposed to guide the layout of road signs.Then,the encoder recursive positioning method is constructed.After that,based on the historical data of robot operation,a recursive least squares based method for correcting encoder coefficients in segments and directions is proposed.Finally,based on the Kalman filter algorithm,the encoder and NFC data fusion localization is implemented at the segment end of the track segment.Recursive positioning is performed within a segment using encoder correction coefficients and encoder information.Thus,the continuous highprecision positioning of the rail type inspection robot is realized.(4)The intelligent inspection robot prototype of belt conveyor is tested.Firstly,a prototype test experiment is designed using the patrol robot prototype to verify the important functions of the patrol robot system such as network communication,information collection,remote monitoring,and conveyor belt fault detection.Then,in order to verify the adaptability and accuracy of the high-precision positioning method based on encoder and NFC correction fusion on curved tracks,an experimental platform is built and tested separately.The thesis has 53 pictures,13 tables and 105 references. |
PDF Full Text Request