Thermal image generation for blast furnace chute based on generative adversarial network

Blast furnaces are the largest source of direct CO 2 emissions in the steel-making process.The chute is one of the main equipment at the top of the blast furnace. Monitoring the conditions of the chute under high temperature has become an urgent problem to be solved. Under the existing equipment image information, a suitable method for the generation of thermal image of chute is proposed. Specifically, the numerical simulation and generative adversarial network (GAN) are introduced to generate images with complete chute. Compared with other typical methods, the proposed method has higher PSNR and SSIM and lower MAE (PSNR = 33.801, SSIM = 0.980, MAE = 0.013). This method can provide a reference for heat condition monitoring of blast furnace chute. Further, it can guide the blast furnace operation and improve gas utilization.