Blast furnace iron mouth the location of the outflow of high temperature molten iron,molten iron liquid surface due to periodic out blast furnace iron float,caused by internal factors such as the molten iron circulation flushing hearth,causing hearth iron mouth bring serious erosion,when the ceramic cup is completely after erosion,high temperature molten iron direct contact with the carbon brick,prone to hearth burning through major accidents,etc.Monitoring of blast furnace hearth erosion morphology of blast furnace production safety is important,but in the normal production of blast furnace has a "black box" features,difficult to directly measure hearth erosion thickness,the embedded temperature sensor using refractory brick wall measured temperature data,using the inverse problem of heat transfer simulation of hearth temperature field distribution,on-line monitoring,implementation of blast furnace hearth erosion morphology of blast furnace boiler stop judgment provides the theory basis for overhaul.A two-dimensional polar coordinate heat transfer model was established according to the foot erosion morphology of the hearth.The contact resistance caused by solidification of molten iron and adhesion of slag to the hearth inner wall was considered.The inner node method was used to divide the model grid,the finite volume method was used to discretize the thermal conductivity differential equation,and the discrete equation was solved and calculated by MATLAB.Taking the shrinkage of X42 molten steel in the cooling process of round mould as an example,the correctness of the mathematical model of forward problem is verified.The mathematical model of the furnace hearth inverse problem is established by combining L-M algorithm with the variation part of genetic algorithm.The wavelet analysis method is used to de-noise the temperature data measured by the temperature sensor in the refractory brick wall.Based on noise reduction temperature data,through iterative calculation inverse problem model from inside the cylinder and the contact thermal resistance,heat flow density distribution function problem model of substitution are hearth cross section temperature field distribution,the temperature field temperature measurement points position compared with the measured temperature,through the relative root mean square error of the inverse problem model precision.The influence of different position of temperature sensor in refractory brick wall,noise when measuring temperature data and different Biot number on the calculation result of heat transfer inverse problem is studied,and the data simulation experiment is carried out by using Matlab.In view of the influence of the three factors on the heat transfer inverse problem,the relative mean square root is used to evaluate the accuracy of the heat transfer inverse problem under the influence of different factors. |