Temperature field reconstruction of hot blast furnace shell based on DE-RBF hybrid algorithm

Abstract This paper explores the radial basis function interpolation algorithm in spatial interpolation techniques to determine the temperature distribution pattern on the furnace shell, serving as a crucial tool for ironmaking. First of all, three types of arrangement schemes for temperature sensors are initially established. The principle of radial basis function interpolation on the outer surface of the hot blast furnace shell is analyzed. Based on this analysis, an objective function linking RBF parameters and the root-mean-square-error(RMSE) of the reconstructed results is formulated. The optimal parameters of RBF are determined using the differential evolution (DE) algorithm. The average absolute error value, RMSE, and coefficient of determination of the reconstructed temperature field are utilized as evaluation metrics. The study was compared to the weighted radial basis function (RBF) interpolation algorithm to determine the temperature distribution pattern on the furnace shell. The method is being compared to the weighted RBF. These findings show that the RBF, when optimized with the Differential Evolution (DE) algorithm, outperforms the weighted radial basis function interpolation algorithm with reductions in average absolute error value and RMSE by 3.37% and 4.07%, and an improvement in the coefficient of determination by 3.99%. This enhancement results in a more effective reconstruction of the temperature field.