Research On The Technology Of Carriage Residual Coal Cleaning System Based On Machine Vision

The problem of residual coal adhering to coal train carriages is a difficult one that has a significant impact on the production and efficiency of coal-fired power plants.With coal scarcity,high coal prices,stringent energy saving targets and the need for dual-carbon environmental protection,coal-fired power plants need a safe,efficient,cost-effective,realtime and intelligent solution to the problem of residual coal in carriages.At present,domestic and international companies mainly use manual cleaning to solve the problem of coal residuals in coal carriages,but this method is time-consuming and inefficient.In order to solve this problem,this project proposes a system based on machine vision technology to clean the remaining coal in the carriages,which aims to automatically clean the remaining coal in the carriages,reduce the safety risk of the staff,improve the working environment,reduce the labour intensity of the staff and improve the work efficiency.The main research content is as follows:In the overall scheme design of the system,according to the overall design requirements of the system,the functions and technical indicators of the system were determined,and the residual coal cleaning system for the carriage was designed;The system consists of three components: mechanical system,control system,and visual system.Among them,the mechanical system includes two parts: a mobile car and a cleaning device;The control system consists of hardware and software components;The visual system includes carriage number recognition and video surveillance.Finally,explain the workflow of the entire system.The size of the various coal train carriages varies,so it is necessary to identify the coal carriages by number and obtain information about the model.A suitable industrial camera and its installation location are selected,the camera calibration principle is analysed and the calibration parameters are obtained.Firstly,a text detection model is established using PSENet network to pinpoint the coordinates of the character area and text box,and then the carriage location is obtained through the conversion of the coordinate system;then,a CRNN recognition algorithm model is used to recognise the characters in the image.After experiments,the accuracy of the text detection module can reach 96.5% and the accuracy of the text recognition module can reach 98.3%.Therefore,the carriage number recognition and carriage location positioning scheme based on machine vision proposed in this paper is feasible.For the implementation of a machine vision based carriage residual coal cleaning system,first determine the control scheme of the carriage residual coal cleaning system,select the main components in the system,draw the system control circuit diagram,and complete the control program writing and upper computer software development after clarifying the entire system control mode.Finally,the system was tested by Anhui Huadian’s Liuan Power Plant.The carriage residual coal cleaning system is able to automatically identify the carriage number and locate the carriage position,and is highly efficient and intelligent in cleaning the residual coal.During the test,the system took only 3 minutes and 30 seconds to efficiently clean the remaining coal in the carriages,and the test results showed that the whole system was effective.