Research On Steel Slag Thickness Measurement Based On Thermal Model And Image Processing In Tundish

In the process of continuous casting, the steel level of tundish has a direct effect on guaranteeing the quality of billet and the security of product. This research subject is a sub subject of "molten steel level measurement in continuous casting tundish", in the process of measuring the molten steel level based on visual inspection, the actual height of molten steel level is found by the difference of slag layer level and slag layer thickness. So it is of important significance to measure slag layer thickness accurately.The covering slag stays in the measuring rod in the form of incrustation and adhesion, it takes difficulty to collect the temperature information of measuring rod. To solve the problem of measuring the slag layer thickness and collecting the information of measuring rod, by establishing the mechanism model of the measuring rod to analyze the temperature information base on the image processing and the result of model, it can improve the information accuracy with the method of mix remove noise. The main works are as follows:(1) Mechanism of slag thickness measurement. By inserting measuring rod into tundish, through Measuring rod to sense the air layer, protection layer and the molten slag layer temperature, the surface temperature gradient information of the measuring rod can be extracted, to detect the slag thickness. By using ANSYS finite element analysis software to establish the measuring rod model, the changes of measuring rod surface temperature gradient can be simulated, Theoretically the feasibility can be proved based on temperature gradient to measure slay layer thickness. Through more analysis of the maintenance time of the identificated temperature field, the best time of acquisition of temperature information can be obtained.(2) Implementation of slag thickness measuring mechanism model, The measurement in slag thickness is achieved by computer vision measurement system through the analysis and processing of collected image. Because the temperature of molten steel in tundish is about as high as1500℃, slag layer of protection above Molten steel has a certain viscosity caused by high temperature. So it can cover the surface of the measuring rod, what would be the pseudo-source information in the process of measuring. Besides, the noise environment on site disruption also increased the difficulties of information extraction accuracy. Based on the analysis of pseudo-information disturbance, the paper achieved image segmentation, extraction and denoising by using the particle swarm algorithm and hybrid ostu information filtering algorithms, and ultimately measured the slag thickness accurately.Through the mechanism model, this paper proved that there is a significant temperature gradient between the interface of measuring rod, air layer, protection layer and molten slag. The purpose of measuring the slag layer thickness can be achieved by processing the acquisition image and extracting temperature information accurately.