Industrial Restructuring And Layout Optimization Of Quipment Manufacturing Industrial In Hebei

China is the largest steel user,and iron ore as a non-renewable resource can no longer meet my country's demand for use.However,scrap steel,as a resource for recycling,has an increasing trend year by year,and has gradually become the main raw material for steel production in my country.The consumption of scrap steel in my country has reached 200 million tons per year.At present,most domestic steel manufacturers mainly rely on manual measurement and estimation for the quality of scrap steel.However,manual quality inspection has high risks,strong subjectivity,and low accuracy,which affects subsequent production benefits.In order to solve this problem,improve production efficiency.Based on complex steel scrap data,this paper applies deep learning technology and analyzes the weight ratio of different grades of scrap steel to solve the drawbacks of the traditional scrap steel recycling process.A system for grading and testing of scrap steel has been developed,which has met production needs,has produced significant economic and social benefits,and has broad application prospects.This thesis mainly made three aspects of applied research:1)At present,there are relatively few researches on grading and testing of scrap steel,and data sets need to be collected by oneself.And perform image preprocessing on the original data set,and perform data enhancement and feature enhancement.Finally,the labeling tool LabelImg is used to label different types of scrap steel,which lays the foundation for the establishment of a scrap grading system.2)Modify the configuration file in the model,apply the preprocessed data set to the YOLOv5 algorithm model,and perform real-time detection of the scrap category and location in the image.In this paper,the detection effect of the model is evaluated based on the average precision average value,and the parameters are adjusted according to the experimental results,and finally the scrap steel grading detection model is established.3)Through the combination of algorithm models and hardware equipment such as field cameras,a complete scrap grading system is designed and realized.The system can realize scrap data collection,classification detection,data statistics,and information visualization,which can better help steel companies improve their efficiency.