Health Monitoring Of Metallurgical Crane Brakes Based On Machine Vision

The brake is a key transmission component and safety guarantee device in the metallurgical lifting equipment,and its health state directly affects the safe production of the metallurgical industry.Due to the heavy work load,long service period and harsh working environment of the metallurgical crane,it is easy to cause the brake braking efficiency to decrease,and even a safety accident of brake failure.Therefore,accurate and effective brake health monitoring methods have important theoretical value and practical significance.Starting from this practical problem,this dissertation proposes a metallurgical crane brake health monitoring method based on machine vision by analyzing the working principle and failure cause of the metallurgical crane brake.The main research content of this dissertation includes the following parts:(1)According to the commonly used crane brake types of metallurgical enterprises,a theoretical model of the typical braking process of the crane brake is established,the braking performance and working mechanism of the brake are analyzed,and the main failure forms and reasons are studied.(2)The monitoring method of metallurgical crane brake health parameters based on machine vision is proposed.Firstly,the parameter indexes that can reflect the braking performance are constructed;then the sampling method of each parameter is demonstrated,and the mathematical model of machine vision sampling is established;finally,the experimental bench is built based on this to realize image sampling.(3)Taking the sampled image as the research object,an image recognition algorithm for brake torque and push rod displacement scale is proposed.First,through the Retinex algorithm,histogram equalization algorithm,and hough algorithm to deal with the effects of dust,uneven illumination,and tilt of the shooting angle in the sampled image;then enhance the image features and segment the ruler image;finally,use the template matching method to complete the ruler based on the segmentation result Image Identification.This method can identify the scale indication with a small error.(4)In view of the problem that the parameters such as temperature and friction factor are not easy to monitor directly during the braking process,firstly,a thermomechanical coupling model of the block brake braking process is established in the ABAQUS software for parameter simulation analysis;then a BP neural network model The parameter identification method of the coupled thermo-mechanical model is fitted to the mapping relationship between the data,which complements the drawbacks of the large calculation amount and long time consumption of the simulation model.This method realizes the high-efficiency identification of the parameters of the brake health state.(5)Based on the GUI function of MATLAB software,a man-machine interface was developed to complete the design of the monitoring system,which enriched the safe production and equipment maintenance methods in the metallurgical industry.