Application of artificial intelligence image recognition technology in blast furnace tuyere monitoring

Blast furnace tuyere is the device that can directly observe the internal thermal state of blast furnace,the monitoring of blast furnace tuyere is also an important reference for blast furnace smelting production. Now the judgment of tuyere state mainly depends on the working experience of operators. It is difficult to monitor and respond in time. The image recognition method of tuyere anomaly was studied,and it was proposed to classify it as fine-grained image classification task,so as to solve the problem of small difference of anomaly categories. The image intelligent diagnosis and early warning system of blast furnace tuyere uses deep convolution neural network model and realizes the diagnosis and alarm functions, such as slag hanging, water leakage, big block falling, burning through,coal breaking,air stopping,fast cutting and brightness. The practice shows that the system runs stably in the No. 2 Ironmaking Plant of Nanjing Iron and Steel Company,and the effect is good. The intelligent level of blast furnace production is improved,and it has the value of popularization and application.